Margolis

Benjamin Aaron Gustafsson AdultCognitiveDisease.com

---

Synaptic Restriction and Proteostatic Collapse

Evaluation of Margolis Models in Alzheimer's

Disease Pathogenesis

Abstract

Alzheimer's disease has been dominated by the amyloid cascade hypothesis, positing that amyloid-beta (Aβ) accumulation drives neurodegeneration¹. However, the limited efficacy of anti-amyloid therapeutics indicates that downstream mechanisms—particularly synaptic degeneration and neuronal proteostasis collapse—warrant direct investigation. This report evaluates the research of Seth S. Margolis and colleagues, focusing on two molecular frameworks: (1) Ephexin5 (ARHGEF15), a RhoA-specific guanine nucleotide exchange factor reactivated by Aβ to drive dendritic spine loss; and (2) the Neuronal Membrane Proteasome (NMP), an ubiquitin-independent proteasome complex that links apolipoprotein E (ApoE) genetic variants to tau aggregation⁵. By examining in vitro, in vivo, and human post-mortem data, this analysis critically evaluates both frameworks. The work decouples some aspects of synaptic failure and tauopathy from obligate amyloid dependency, though significant unresolved questions remain regarding mechanism, translational feasibility, and therapeutic specificity¹.

Introduction

Alzheimer's disease represents the most prevalent neurodegenerative condition in aging populations, characterized clinically by progressive memory loss and histopathologically by amyloid-beta plaques and tau neurofibrillary tangles. The amyloid cascade hypothesis, formulated in the 1990s, posited that Aβ accumulation initiates a linear cascade leading to cognitive decline. Yet clinical evidence challenges this model: the strongest biological correlate of cognitive decline is synapse loss, not overall amyloid load¹². In 1991, Terry and colleagues demonstrated that neocortical synapse density showed correlation coefficients of 0.96¹² with cognitive impairment, whereas plaque density explained only a fraction of that variance¹². This established the "synaptic slaughter" paradigm but did not displace the amyloid-centric framework from pharmaceutical development efforts. Margolis's research program addresses this mechanistic gap by identifying specific molecular pathways active in development that become pathologically reactivated in aging. Two pillars merit evaluation:

1. The Ephexin5 pathway: Aβ-induced EphB2 loss triggers Ephexin5 accumulation, driving RhoA-mediated dendritic spine collapse through reactivation of developmental pruning mechanisms¹.
2. The Neuronal Membrane Proteasome: A non-canonical proteasome complex that modulates activity-induced protein synthesis and whose membrane localization is compromised by the ApoE4 genetic risk variant, creating conditions favorable to tau aggregation independent of Aβ pathology⁵.

Literature Review: The Synaptic Paradigm

The field has oscillated between amyloid-centric and synapse-centric models. In 2000s, research identified soluble Aβ oligomers—rather than insoluble plaques—as the primary synaptotoxic agents⁸. Lennart Mucke's laboratory demonstrated that Aβ oligomers bind directly to the EphB2 receptor, triggering its proteasomal degradation and impairing NMDA receptor function and synaptic plasticity²². Margolis's 2010 Cell publication identified Ephexin5 as the molecular link downstream of EphB2 loss³. Under normal development, EphB2 phosphorylates Ephexin5, leading to its UBE3A-mediated degradation and permitting synaptogenesis³. Loss of EphB2 allows Ephexin5 accumulation, which then hyperactivates RhoA and drives spine collapse. This reframes AD pathology: synaptic degeneration is not a novel toxicity but inappropriate reactivation of a developmental pruning mechanism¹.

Methodology

Margolis's laboratory employs a systematic approach from biochemistry to in vivo behavior:

• In vitro assays: Cultured hippocampal neurons treated with soluble Aβ oligomers, analyzed via Western blotting, immunocytochemistry, GTPase pull-down assays, and live FRET imaging¹⁶.
• In vivo models: hAPP transgenic mice (expressing human amyloid precursor protein); genetic crosses with Ephexin5-knockout mice; lentiviral shRNA knockdown in specific brain regions¹.
• Behavioral testing: Novel Place Preference test (spatial memory); Passive Avoidance test (associative learning)¹.
• Human correlates: Post-mortem hippocampal tissue stratified by Braak stage, validating that molecular alterations observed in mice occur in actual AD brains¹. This multi-scale approach reduces the risk of isolated artifacts from any single model system.

Chapter 1: The Ephexin5-RhoA Axis

Developmental Physiology

Ephexin5 is a RhoA-specific guanine nucleotide exchange factor (GEF) that acts as a developmental "brake" on excitatory synapse formation. When a synapse is ready to mature, ephrin-B ligands bind EphB2, triggering phosphorylation of Ephexin5 at Tyrosine-361. This phosphorylation recruits UBE3A, an E3 ubiquitin ligase, which tags Ephexin5 for proteasomal degradation. Removal of Ephexin5 relieves RhoA-mediated actin depolymerization, allowing dendritic spine expansion¹. In the healthy adult brain, Ephexin5 expression is minimal, restricted to high-plasticity regions¹.

Pathological Reactivation in Alzheimer's Disease

Margolis's 2017 Journal of Clinical Investigation study demonstrated this pathway's reactivation in AD:

1. In vitro: Soluble Aβ oligomers applied to mature hippocampal neurons caused rapid EphB2 loss and concurrent Ephexin5 elevation, measured by immunocytochemistry¹.
2. In vivo: Aβ oligomer injection into the dentate gyrus of wild-type mice reproduced Ephexin5 elevation. hAPP transgenic mice showed 2to 3-fold Ephexin5 elevation in the DG and CA regions¹.
3. Human tissue: Post-mortem hippocampal samples from advanced Braak stage patients showed 2to 3-fold Ephexin5 elevation versus age-matched controls, confirming this is not an artifact of transgenic overexpression¹.

Causal Evidence

Ephexin5 knockout in hAPP mice (Ephexin5^-/-) completely rescued mature dendritic spine density despite continued Aβ production and plaque formation¹. Behavioral tests showed full restoration of spatial memory and associative learning¹. Lentiviral shRNA-mediated knockdown of Ephexin5 in the dentate gyrus of presymptomatic hAPP mice prevented cognitive decline development¹. These findings establish that Aβ requires Ephexin5 to execute synaptotoxic effects. Ephexin5 is minimally expressed in the healthy adult brain, making it a potentially selective drug target¹.

Chapter 2: Mechanistic Complexity—Ephexin5's Dual Role

A 2025 Science Advances study revealed that the initial linear model requires revision³¹. Ephexin5 is not exclusively a RhoA-specific GEF; it also activates Cdc42, and this substrate selectivity is dynamically regulated by neuronal activity and tyrosine phosphorylation³¹.

Activity-Dependent Substrate Switching

During plasticity-inducing events (chemical long-term potentiation mimicking learning), Ephexin5 undergoes rapid dephosphorylation³¹. This shift in phosphorylation causes preferential activation of Cdc42 rather than RhoA³¹. Consequently, downstream of neuronal activity, Ephexin5 positively regulates synaptic growth and stabilization—the opposite of its resting state role³¹. Condition | Phosphorylation | Primary Target | Effect Basal/resting | High | RhoA | Spine shrinkage Active learning | Rapid dephosphorylation | Cdc42 | Spine growth Alzheimer's disease | Hyperaccumulation | RhoA (overactivation) | Synaptic loss

The Angelman Syndrome Paradox Resolution

Angelman Syndrome, caused by UBE3A loss, would be expected to cause massive spine loss via Ephexin5 accumulation and RhoA hyperactivation². Instead, spine density increases. The 2025 findings explain this: excess Ephexin5 in developing brains preferentially activates Cdc42-mediated stabilization rather than RhoA-mediated pruning³¹.

Therapeutic Implications and Complications

The discovery of dual functionality substantially complicates drug development³¹. If Ephexin5-mediated Cdc42 activation is physiologically necessary for activity-dependent spine stabilization and memory formation, then global Ephexin5 inhibition could impair learning and memory consolidation in treated patients³¹. Simple small-molecule inhibitors (aggressively pursued after the 2017 study) lack the necessary pharmacological sophistication. Safe translation requires allosteric modulators or biased ligands that selectively prevent Ephexin5-RhoA interaction while preserving Ephexin5-Cdc42 activation during normal learning⁷.

Chapter 3: The Neuronal Membrane Proteasome

Protein homeostasis disruption is a hallmark of neurodegenerative disease¹⁶. Historically, attention focused on the ubiquitin-proteasome system (26S proteasome) and autophagy¹⁶. Margolis's laboratory discovered an alternative: the Neuronal Membrane Proteasome (NMP)¹⁶.

Architecture and Function

The NMP is an uncapped 20S proteasome core particle embedded in the neuronal plasma membrane (characterized 2017-2023)⁴⁵. Unlike the canonical 26S proteasome (cytosolic/nuclear, ATP and ubiquitin-dependent), the NMP operates independently of ubiquitin and ATP¹⁶. Its primary function is to degrade the "nascentome"—proteins newly synthesized at the synapse immediately following neuronal activity⁴. When a neuron fires, local translation rapidly produces plasticity-related proteins (CaMKIIα, PSMD2) necessary for synapse strength modification. However, unchecked synthesis can cause hyperexcitability and destabilizing positive feedback loops⁴. The NMP degrades approximately one-third of nascent polypeptides directly at the membrane during active translation, providing rapid homeostatic modulation⁴.

Evidence for NMP Function

In Xenopus laevis tadpole optic tectum (where transparency permits real-time calcium imaging), acute NMP inhibition (biotin-epoxomicin) induced neural circuit hypersynchronization and abolished learning-induced behavioral plasticity, demonstrating that NMP function is required for adaptive learning⁴.

ApoE4, NMP Localization, and Tau Aggregation

Margolis's recent work (e.g., Paradise et al., 2022) connected NMP function to Apolipoprotein E (ApoE), whose E4 allele is the strongest genetic risk factor for late-onset AD⁵. Key finding: ApoE4 drastically reduces NMP localization to the neuronal plasma membrane compared to the neutral ApoE3 or protective ApoE2 isoforms⁵. This creates localized proteostatic failure at the synapse⁵. Critically, newly synthesized tau is exceptionally vulnerable to aggregation when NMP function is compromised⁵. NMP selective inhibition alone was sufficient to induce spontaneous aggregation of endogenous, newly synthesized tau into sarkosyl-insoluble, high-molecular-weight species—the hallmark of AD tauopathy⁵. The quantitative effect is substantial: ApoE4 lowers the tau aggregation threshold approximately 50-fold relative to ApoE3 (and 200-fold relative to ApoE2), entirely through NMP depletion⁵. This provides an ApoE4-mediated, Aβ-independent mechanism for tau pathology in genetic risk carriers⁵.

Chapter 4: Genetic Architecture and Translational Challenges

ARHGEF15 in Human Genetics

Transcriptomic analysis of AD brains shows ARHGEF15 (Ephexin5) upregulation in both incipient and advanced disease⁴³. Genome-wide epigenetic profiling reveals significant DNA methylation alterations at ARHGEF15 in AD, consistent with transcriptional dysregulation⁴⁴. Notably, rare missense mutations in ARHGEF15 are identified as causal for autosomal dominant hereditary Cerebral Small Vessel Disease (CSVD)⁴⁵. This indicates ARHGEF15 has distinct functions in cerebral endothelial cells, where it mediates VEGF-induced Cdc42 activation to promote angiogenesis and vascular growth⁴⁸.

The Pleiotropy Problem

ARHGEF15 is not brain-specific. The vascular involvement noted above creates a critical challenge: systemic Ephexin5 inhibitors designed to halt AD-related synaptic pruning risk compromising blood-brain barrier integrity or exacerbating cerebral small vessel disease through anti-angiogenic effects. Safe translation requires targeted delivery: neuronal-specific proteolysis-targeting chimeras (PROTACs) that degrade Ephexin5 only in neurons, or highly specific allosteric modulators that avoid off-target effects in vascular tissue⁷.

NMP Therapeutic Development

Because AD pathology in this model stems from loss of NMP membrane localization (driven by ApoE4), classical enzyme inhibitors are ineffective⁵. Therapeutic approaches must involve:

• Small molecules or gene therapies that rescue NMP trafficking to the plasma membrane
• Strategies enhancing 20S proteasome assembly at the membrane
• Synthetic compounds mimicking the yet-unidentified extracellular signaling peptides generated by normal NMP cleavage Current pharmacology lacks established precedent for this class of intervention⁴.

Critical Evaluation: Strengths and Limitations

Strengths

1. Multi-scale validation: Molecular findings tracked from biochemistry through in vivo behavior and human post-mortem tissue¹.
2. Causal establishment: Genetic deletion and lentiviral knockdown experiments, not merely correlational observations¹.
3. Mechanism-specific: Both frameworks identify specific molecular nodes rather than invoking "amyloid toxicity" generally¹.
4. Genetic integration: Findings align with transcriptomic and epigenetic AD data, and the ApoE4 connection addresses a major genetic risk factor directly¹.
5. Honest recent updates: The 2025 Science Advances findings on Ephexin5 dual functionality are incorporated, complicating the initial model rather than glossing over contradictions³¹.

Significant Limitations and Uncertainties

1. Ephexin5 as a complete explanation: The model accounts for spine loss downstream of Aβ and ApoE4, but does not explain:

• Why Aβ accumulation initiates in the first place
• The contribution of neuroinflammation
• Non-synaptic neuronal dysfunction
• The timeline and sequence of tau emergence relative to Aβ The framework is mechanistically sophisticated but still incomplete regarding AD initiation¹.

2. NMP-tau causality remains partial: NMP inhibition induces tau aggregation in vitro, and ApoE4 reduces membrane localization¹. However:

• Whether this mechanism is sufficient to explain primary tauopathies (tau pathology without significant Aβ pathology) is unclear
• The identity and role of the signal peptides generated by NMP cleavage remain unknown
• Human evidence remains limited to post-mortem correlation and animal models

3. Therapeutic complexity: The discovery of Ephexin5's dual role (RhoA shrinkage vs. Cdc42 stabilization)³¹ fundamentally complicates the drug development trajectory outlined in the original framework. Blunt inhibitors appear unviable; the field must develop novel biased ligands or allosteric modulators with no established precedent⁷.
4. Pleiotropy and off-target risk: Both ARHGEF15's expression in vasculature and the NMP's broad distribution mean that systemic interventions carry genuine risk of unintended effects⁴⁵¹.
5. Transgenic model limitations: While hAPP transgenic mice show robust synaptic phenotypes, they may not fully recapitulate the temporal dynamics or cellular context of human AD¹. The dependency on APP overexpression systems is a recurring concern¹.
6. Reversibility unclear: The experimental rescues (Ephexin5-/- mice) prevent pathology development; less is known about whether the same interventions can reverse established synaptic loss or cognitive deficits¹.

What This Analysis Cannot Determine

1. Sufficiency of each mechanism: While Margolis's frameworks account for certain aspects of AD pathology (spine loss, tau aggregation), they do not establish that targeting these pathways alone will robustly reverse cognitive decline in humans. Combination therapies may be necessary.
2. Temporal and causal sequencing in human disease: Which comes first—Aβ accumulation, NMP dysfunction, or tau emergence—remains unclear in human brain. Animal models cannot fully resolve this ambiguity.
3. Cell-type specificity: Both Ephexin5 and the NMP function in multiple cell types (neurons, glia, vasculature). Whether the pathology in AD is neuron-intrinsic or involves glial contribution is not definitively answered.
4. Clinical translation timeline: No human clinical trials have tested Ephexin5 inhibitors or NMP-targeting therapeutics. The feasibility of achieving the required pharmacological selectivity in patients remains speculative.
5. Disease heterogeneity: AD is likely heterogeneous. These mechanisms may dominate in some patients while being secondary in others. How to identify responders to Ephexin5 or NMP-based interventions is unaddressed.
6. Long-term safety: The off-target risks noted (vascular effects, learning impairment) are theoretical but untested. Chronic systemic pharmacological intervention carries unknown risks.
7. Integration with other pathways: Neuroinflammation, metabolic dysfunction, and mitochondrial pathology are documented in AD but not mechanistically integrated into the Ephexin5-NMP framework.

The Validity Ledger

The argument above is only as strong as its weakest load-bearing joint, and the reader is owed an explicit accounting of where it stands on the ground and where it stands on inference.

Conclusion

Margolis's research program provides specific molecular frameworks addressing a genuine gap in AD mechanism: how does Aβ-driven EphB2 loss translate into structural synaptic collapse, and how does genetic risk (ApoE4) predispose to tau without obligate Aβ dependency? The Ephexin5 pathway offers a plausible mechanistic link between EphB2 depletion and dendritic spine loss through RhoA hyperactivation, reframing AD synaptic failure as inappropriate reactivation of developmental pruning. The NMP framework provides an Aβ-independent mechanism connecting ApoE4 genetic risk to tau aggregation through impaired membrane-localized protein homeostasis. Both frameworks are grounded in careful experimental work spanning biochemistry, in vivo imaging, behavior, and human tissue correlation. However, the 2025 discovery of Ephexin5's dual functionality substantially complicates therapeutic development, the NMP-tau mechanism remains incomplete (with critical signaling peptides unidentified), and neither framework fully explains AD initiation or accounts for other documented pathological processes. The work represents a significant advance in understanding downstream mechanisms of synaptic failure and specific aspects of genetic risk. Translation to safe, effective human therapeutics will require solutions to pleiotropy, development of unprecedented pharmacology (biased Ephexin5 ligands, NMP trafficking modulators), and resolution of unanswered questions regarding causality and mechanism. This is a strong but partial framework, not a complete solution to AD pathogenesis.

Works Cited

1. Seth Margolis Paper.pdf

2. Talan J. Researchers Identify Novel Potential New Pathway and Target for Alzheimer's Disease. Neurology Today 2017;17(13). DOI 10.1097/01.nt.0000521708.80399.66.

https://neurologytoday.aan.com/doi/10.1097/01.NT.0000521708.80399.66

3. Margolis SS, Salogiannis J, Lipton DM, Mandel-Brehm C, Wills ZP, Mardinly AR, et al.. EphB-mediated degradation of the RhoA GEF Ephexin5 relieves a developmental brake on excitatory synapse formation. Cell 2010;143(3):442-55. DOI 10.1016/j.cell.2010.09.038.

4. He H, Ahsan A, Bera R, McLain N, Faulkner R, Ramachandran KV, et al.. Neuronal membrane proteasomes regulate neuronal circuit activity in vivo and are required for learning-induced behavioral plasticity. Proceedings of the National Academy of Sciences 2023;120(3). DOI 10.1073/pnas.2216537120.

5. ApoE isoforms differentially regulate neuronal membrane proteasomes to shift the threshold for pathological aggregation of endogenous Tau | bioRxiv, accessed March 23, 2026, https://www.biorxiv.org/content/ 6. (PDF) Activity-dependent regulation of Cdc42 by Ephexin5 drives synapse growth and stabilization - ResearchGate, accessed March 23, 2026, https://www.researchgate.net/publication/390249669_Activity-dependent_regulatio n_of_Cdc42_by_Ephexin5_drives_synapse_growth_and_stabilization

6. Cook EK, Sell GL, Schaffer TB, Margolis SS. The emergence of Ephexin5 as a therapeutic target in Alzheimer's disease. Expert Opin Ther Targets 2019;23(4):263-265. DOI 10.1080/14728222.2019.1586884.

7. Marsh J, Alifragis P. Synaptic dysfunction in Alzheimer's disease: the effects of amyloid beta on synaptic vesicle dynamics as a novel target for therapeutic intervention. Neural Regen Res 2018;13(4):616-623. DOI 10.4103/1673-5374.230276.

8. The Future Of Alzheimer's Disease: Some Hope, But Daunting Numbers, accessed March 23, 2026, https://margolisbloom.com/planning-for-life/the-future-of-alzheimers-disease-some -hope-but-daunting-numbers/

9. Pimplikar SW. Reassessing the amyloid cascade hypothesis of Alzheimer's disease. Int J Biochem Cell Biol 2009;41(6):1261-8. DOI 10.1016/j.biocel.2008.12.015.

10. Crary J. Primary age-related tauopathy and the amyloid cascade hypothesis: the exception that proves the rule?. Journal of Neurology and Neuromedicine 2016;1(6):53-57. DOI 10.29245/2572.942x/2016/6.1059.

11. Terry RD, Masliah E, Salmon DP, Butters N, DeTeresa R, Hill R, et al.. Physical basis of cognitive alterations in Alzheimer's disease: synapse loss is the major correlate of cognitive impairment. Ann Neurol 1991;30(4):572-80. DOI 10.1002/ana.410300410.

12. Correcting Synapse Loss in the Alzheimer's Disease Brain - BrightFocus, accessed March 23, 2026, https://www.brightfocus.org/grant/correcting-synapse-loss-in-the-alzheimers-disea se-brain/

13. Sell GL, Schaffer TB, Margolis SS. Reducing expression of synapse-restricting protein Ephexin5 ameliorates Alzheimer's-like impairment in mice. J Clin Invest 2017;127(5):1646-1650. DOI 10.1172/JCI85504.

14. Sell GL, Schaffer TB, Margolis SS. Reducing expression of synapse-restricting protein Ephexin5 ameliorates Alzheimer's-like impairment in mice. J Clin Invest 2017;127(5):1646-1650. DOI 10.1172/JCI85504.

15. Mechanisms of ubiquitin-independent proteasomal degradation and their roles in

Church TR, Margolis SS. Mechanisms of ubiquitin-independent proteasomal degradation and their roles in age-related neurodegenerative disease. Front Cell Dev Biol 2024;12:1531797. DOI 10.3389/fcell.2024.1531797.

17. Synaptic Loss in Alzheimer's Disease: Mechanistic Insights Provided by Two-Photon in vivo Imaging of Transgenic Mouse Models - Frontiers, accessed March 23, 2026, https://www.frontiersin.org/journals/cellular-neuroscience/articles/0 20.592607/full

18. Terry N, Masliah A, Overk C, Masliah E. Remembering Robert D. Terry at a Time of Change in the World of Alzheimer's Disease. J Alzheimers Dis 2019;70(3):621-628. DOI 10.3233/JAD-190518.

19. John A, Reddy PH. Synaptic basis of Alzheimer's disease: Focus on synaptic amyloid beta, P-tau and mitochondria. Ageing Res Rev 2021;65:101208. DOI 10.1016/j.arr.2020.101208.

20. Alzheimer's disease as a synaptopathy: Evidence for dysfunction of synapses during disease progression - Frontiers, accessed March 23, 2026, https://www.frontiersin.org/journals/synaptic-neuroscience/articles/ 023.1129036/full

21. Koffie RM, Meyer-Luehmann M, Hashimoto T, Adams KW, Mielke ML, Garcia-Alloza M, et al.. Oligomeric amyloid β associates with postsynaptic densities and correlates with excitatory synapse loss near senile plaques. Proceedings of the National Academy of Sciences 2009;106(10):4012-4017. DOI 10.1073/pnas.0811698106.

22. van Dijken I, van der Vlag M, Flores Hernández R, Ross A. Perspectives on Treatment of Alzheimer's Disease: A Closer Look into EphB2 Depletion. J Neurosci 2017;37(47):11296-11297. DOI 10.1523/JNEUROSCI.0214-17.2017.

23. Cissé M, Halabisky B, Harris J, Devidze N, Dubal DB, Sun B, et al.. Reversing EphB2 depletion rescues cognitive functions in Alzheimer model. Nature 2011;469(7328):47-52. DOI 10.1038/nature09635.

24. Wood H. Microglial senescence is a potential therapeutic target for Alzheimer disease. Nature Reviews Neurology 2024;20(7):379-379. DOI 10.1038/s41582-024-00979-3.

25. Shi XD, Sun K, Hu R, Liu XY, Hu QM, Sun XY, et al.. Blocking the Interaction between EphB2 and ADDLs by a Small Peptide Rescues Impaired Synaptic Plasticity and Memory Deficits in a Mouse Model of Alzheimer's Disease. J Neurosci 2016;36(47):11959-11973. DOI 10.1523/JNEUROSCI.1327-16.2016.

26. Eph receptors at synapses - Ip Lab - Molecular Neuroscience Laboratory, accessed March 23, 2026, https://iplab.hkust.edu.hk/pdf/Reviews%20pdf/2012_03%20CellSig.pdf

27. Seth Margolis - The Solomon H Snyder Department of Neuroscience, accessed March 23, 2026, https://neuroscience.jhu.edu/research/faculty/109

    Find this paper

28. Azar B. QnAs with Hollis T. Cline. Proceedings of the National Academy of Sciences 2023;120(3). DOI 10.1073/pnas.2221189120.

29. Schaffer TB, Smith JE, Cook EK, Phan T, Margolis SS. PKCε Inhibits Neuronal Dendritic Spine Development through Dual Phosphorylation of Ephexin5. Cell Rep 2018;25(9):2470-2483.e8.2018.11.005.

    Find this paper

30. Schmidt SI, Blaabjerg M, Freude K, Meyer M. RhoA Signaling in Neurodegenerative Diseases. Cells 2022;11(9). DOI 10.3390/cells11091520.

31. Petshow S, Coblentz A, Hamilton AM, Sarkar D, Anisimova M, Flores JC, et al.. Activity-dependent regulation of Cdc42 by Ephexin5 drives synapse growth and stabilization. Sci Adv 2025;11(13):eadp5782. DOI 10.1126/sciadv.adp5782.

32. Sheffler-Collins SI, Dalva MB. EphBs: an integral link between synaptic function and synaptopathies. Trends Neurosci 2012;35(5):293-304.2012.03.003.

    Find this paper

33. Sell GL, Schaffer TB, Margolis SS. Reducing expression of synapse-restricting protein Ephexin5 ameliorates Alzheimer's-like impairment in mice. J Clin Invest 2017;127(5):1646-1650. DOI 10.1172/JCI85504.

34. Sell GL, Schaffer TB, Margolis SS. Reducing expression of synapse-restricting protein Ephexin5 ameliorates Alzheimer's-like impairment in mice. J Clin Invest 2017;127(5):1646-1650. DOI 10.1172/JCI85504.

35. Petshow S, Coblentz A, Hamilton AM, Sarkar D, Anisimova M, Flores JC, et al.. Activity-dependent regulation of Cdc42 by Ephexin5 drives synapse growth and stabilization. Sci Adv 2025;11(13):eadp5782. DOI 10.1126/sciadv.adp5782.

36. Duman JG, Blanco FA, Cronkite CA, Ru Q, Erikson KC, Mulherkar S, et al.. Rac-maninoff and Rho-vel: The symphony of Rho-GTPase signaling at excitatory synapses. Small GTPases 2022;13(1):14-47. DOI 10.1080/21541248.2021.1885264.

37. Duman JG, Blanco FA, Cronkite CA, Ru Q, Erikson KC, Mulherkar S, et al.. Rac-maninoff and Rho-vel: The symphony of Rho-GTPase signaling at excitatory synapses. Small GTPases 2021;13(1):14-47. DOI 10.1080/21541248.2021.1885264.

38. Behl C. In 2024, the amyloid-cascade-hypothesis still remains a working hypothesis, no less but certainly no more. Front Aging Neurosci 2024;16:1459224.24.1459224.

    Find this paper

39. A?? Immunotherapy Leads to Clearance of Early, but Not Late, Hyperphosphorylated Tau Aggregates via the Proteasome | Request PDF - ResearchGate, accessed March 23, 2026, https://www.researchgate.net/publication/253126561_A_Immunotherapy_Leads_t o_Clearance_of_Early_but_Not_Late_Hyperphosphorylated_Tau_Aggregates_via _the_Proteasome

    Find this paper

40. Neuronal membrane proteasomes homeostatically regulate neural circuit activity in vivo and are required for learning-induced behavioral plasticity | bioRxiv, accessed March 23, 2026, DOI 10.1101/2022.01.29.478314">https://www.biorxiv.org/content/10.1101/2022.01.29.478314

41. He HY, Ahsan A, Bera R, McLain N, Faulkner R, Ramachandran KV, et al.. Neuronal membrane proteasomes regulate neuronal circuit activity in vivo and are required for learning-induced behavioral plasticity. Proc Natl Acad Sci U S A 2023;120(3):e2216537120. DOI 10.1073/pnas.2216537120.

42. Davidson K, Pickering AM. The proteasome: A key modulator of nervous system function, brain aging, and neurodegenerative disease. Front Cell Dev Biol 2023;11:1124907. DOI 10.3389/fcell.2023.1124907.

43. Human Gene Set: BLALOCK_ALZHEIMERS_DISEASE_UP, accessed March 23, 2026,

https://www.gsea-msigdb.org/gsea/msigdb/human/geneset/BLALOCK_ALZHEIME RS_DISEASE_UP.html

44. C Silva T, Zhang W, Young JI, Gomez L, Schmidt MA, Varma A, et al.. Distinct sex-specific DNA methylation differences in Alzheimer's disease. Alzheimers Res Ther 2022;14(1):133. DOI 10.1186/s13195-022-01070-z.

45. Ding X, Chen Y, Guo C, Fu Y, Qin C, Zhu Q, et al.. Mutations in ARHGEF15 cause autosomal dominant hereditary cerebral small vessel disease and osteoporotic fracture. Acta Neuropathologica 2023;145(5):681-705. DOI 10.1007/s00401-023-02560-6.

46. Ogunwale AN, Schulz PE, des Bordes JK, Elefteriou F, Rianon NJ. Potential Biological and Genetic Links Between Dementia and Osteoporosis: A Scoping Review. Geriatrics (Basel) 2025;10(4). DOI 10.3390/geriatrics10040096.

47. Lydon E, Harris M, Mudar R, Rogers W. Understanding Everyday Needs of Older Adults with Cognitive Impairments: Insights from Subject Matter Experts. Archives of Physical Medicine and Rehabilitation 2020;101(11):e10. DOI 10.1016/j.apmr.2020.09.027.

48. Kim K, Lee SA, Park D. Emerging Roles of Ephexins in Physiology and Disease. Cells 2019;8(2). DOI 10.3390/cells8020087.

49. Kim KW, Kusuhara S, Katsuyama-Yoshikawa A, Nobuyoshi S, Kitamura M, Mori S, et al.. Changes in Gene Expression Profiling and Phenotype in Aged Multidrug Resistance Protein 4-Deficient Mouse Retinas. Antioxidants (Basel) 2021;10(3). DOI 10.3390/antiox10030455.

50. ARHGEF15 Gene - Ma'ayan Lab – Computational Systems Biology, accessed March 23, 2026, https://maayanlab.cloud/Harmonizome/gene/ARHGEF15

51. Türker F, Brennan A, Margolis SS. Neuronal membrane proteasome-derived peptides modulate NMDAR-dependent neuronal signaling to promote changes in gene expression. Mol Biol Cell 2024;35(1):ar6. DOI 10.1091/mbc.E23-06-0218.