Author Contributions: P.K.S, E.N.S.P., S.S., and E.H.N., designed research; P.K.S., E.N.S.P., Z.L.C., D.T., M.C., and A.S. performed experiments; P.K.S., S.S., and E.H.N., wrote the paper.

Competing Interests: The authors have no competing interests.

ARIA Main

Alzheimer's disease (AD) is a neurodegenerative dementia characterized by the accumulation of amyloid beta (Aβ) aggregates in the brain parenchyma and in/around blood vessels (cerebral amyloid angiopathy,

CAA)(1, 2). An early feature in AD is the disruption of the blood-brain barrier (BBB), which leads to the extravasation and accumulation of blood proteins within the brain, worsening AD pathology(1, 3, 4). Fibrinogen, an abundant blood protein and major component of blood clots, can form a complex with Aβ upon binding to Aβ’s N-terminus (residues 8-20)(1, 5). Aβ may contribute to vascular abnormalities in AD by binding to fibrinogen and altering fibrin clot degradation (1). Consistent with this idea, reducing ifbrinogen levels , or inhibiting Aβ/fibrinogen binding in AD mice leads to decreased BBB permeability, reduced neuroinflammation , decreased CAA, and less cognivtie decline (6, 7).

The FDA-approved immunotherapy for AD, lecanemab, directed against Aβ protobfirils, reduces Aβ burden and slows cognivtie decline in AD patients (8). However, lilte is known regarding its mechanisms-of-action . Lecanemab targets Aβ’s N-terminus (residues 1-16), overlapping with fibrinogen ’s binding site on Aβ(2, 5). We investigate d the interaction between Aβ42 and fibrinogen in the absence and presence of lecanemab (Fig 1A-C). Lecanemab blocked Aβ42 binding to fibrinogen in a dose-dependent manner, while human IgG had no eefct. The Aβ42 preparation used was comprised of curvy linear aggregates (small protofibrils) 3090 nm in length (Fig 1D).

The interaction between A β42 and fibrinogen leads to an abnormal clot structure that is resistant to plasmin-induced fibrinolysis (1). Since lecanemab blocked the Aβ42/fibrinogen interaction , we analyzed the eefct of lecanemab on Aβ42/fibrinogen -mediated impaired fibrinolysis in a purified protein system. As reported, Aβ42 protobfirils delayed plasmin-induced fibrinolysis . However, lecanemab blocked this eefct of Aβ42, whereas human IgG did not (Fig 1E&F).

Aβ42 aggregates also alter fibrin clot turbidity, an indicator of altered fibrin assembly (7). Lecanemab, but not human IgG, rescued the defect in fibrin assembly caused by A β42 protofibrils (Fig 1E&G). We also analyzed clot morphology using scanning electron microscopy (SEM). As previously reported, Aβ42 disrupts normal clot morphology, causing thinning of the fibrin bundles, abnormal clustering, and entangled clumps (Fig 1H-J). However, in the presence of Aβ42 protofibrils and lecanemab, these structural clot abnormalities were significantly corrected (Fig 1H-J). Human IgG control had no eefct on Aβ42induced clot abnormalities (Fig 1H -J).

To determine if lecanemab inhibits the Aβ42/fi brinogen complex ex vivo, we incubated biotinylated Aβ42 protofibrils (B-Aβ42) with buffer, lecanemab, or human IgG and added them to normal human plasma (NHP). Immunoprecipitation was performed to pull down Aβ42 and any bound proteins. Fibrinogen immunoprecipitated with Aβ42 (Fig 2A&B). However, in the presence of lecanemab, Aβ42 did not pulldown fibrinogen , indicating that lecanemab blocked A β42/fibrinogen complex formation in NHP (Fig 2A&B). Also, consistent with the in vitro results (Fig 1E-J), lecanemab significantly corrected Aβ42-induced clot abnormalities in human plasma (Fig 2C-H).

Synapse loss in AD is associated with memory impairment(4, 9). For example, the reduction of presynaptic protein synaptophysin (SYP) and post-synaptic density protein -95 (PSD-95) in the hippocampus corresponds to cognivtie deficits in AD (10, 11). Extravasated fibrinogen can contribute to synapcti dysfunction (1, 3, 4, 12). Therefore, we explored if lecanemab could alter fibrinogen’s eefct on A β42mediated synaptotoxicity by examining the levels of SYP and PSD-95 in mouse organotypic hippocampal culture (OHC). Treatment of OHCs with a mixture of Aβ42 protofibrils and fibrinogen reduced SYP and PSD95 (Fig 2I-K). Lecanemab, but not human IgG, inhibited Aβ42/fibrinogen -mediated synaptic changes (Fig 2I-K).

Lecanemab also dissociated preformed Aβ42/fibrinogen complex es in human plasma (Fig 2L&M). Moreover, lecanemab migtiate d synaptotoxicity induced by preformed Aβ42/fibrinogen complex es in mouse OHCs (Fig 2N-P). Therefore, the ability of lecanemab to block Aβ/fibrinogen complex formaotin or dissociate the complex may be a component of its beneficial eefcts.

Amyloid-related imaging abnormaliti es (ARIA), a side eefct of A β immunotherapy, are common in CAA(2, 13). Blocking the Aβ/fibrinogen interacotin reduces CAA pathology and improves memory in AD mice(6). Our results show that lecanemab also targets the Aβ/fibrinogen complex. Although in some AD patients lecanemab treatment induces serious ARIA, it causes less ARIA than other anti-Aβ antibodies(2, 8, 14). Future studies are necessary to understand the connection between the Aβ/fibrinogen complex, lecanemab, and ARIA.

Our findings suggest that further invesgtiation s into lecanemab’s mechanisms-of-action are necessary in AD mouse models and AD patients. Studies include assessing lecanemab's ecfiacy in dissociating and method of clearing Aβ/fibrinogen complexes in vivo and understanding its mechanism in migtiating Aβ/fibrinogen -induced synaptotoxicity. Moreover, given the neurodegenerative impact of extravasated ifbrin (ogen) into the brain parenchyma independent of Aβ(1, 3, 4), exploring a combinatorial therapeucti strategy using lecanemab alongside a fibrin -specific antibody or another relevant target could be a promising treatment plan to improve upon the current AD immunotherapies (12).

Details of reagents and methods (Aβ42 preparation , binding, immunoprecipitation , in vitro and ex vivo cloting and fibrinolysis, EM, and mouse OHCs) are included in the supporntig information. This work was supported by NIH grant NS106668, Samuel Newhouse Foundation, John A. Herrmann, Zina Stern Fellowship, Mr. and Mrs. Pamela and Bill Michaelcheck, and May and Samuel Rudin Family Foundation. constitutes PBS+DMSO. Comparisons among mulptile groups were performed using one-way ANOVA followed by Newman-Keuls mulptile comparison test. Data are presented as mean ± SEM. ****p<0.0001, ***p<0.001; ns, not significant. 10. 11. 12. 13.

C. I. Sze et al., Loss of the presynaptic vesicle protein synaptophysin in hippocampus correlates with cognivtie decline in Alzheimer disease. J Neuropathol Exp Neurol 56, 933-944 (1997). R. Sultana, W. A. Banks, D. A. Buetrefild, Decreased levels of PSD95 and two associated proteins and increased levels of BCl2 and caspase 3 in hippocampus from subjects with amnestic mild cognivtie impairment: Insights into their potential roles for l oss of synapses and memory, accumulation of Abeta, and neurodegeneration in a prodromal stage of Alzheimer's disease. J Neurosci Res 88, 469-477 (2010).

J. K. Ryu et al., Fibrin-targeting immunotherapy protects against neuroinflammation and neurodegeneration. Nat Immunol 19, 1212-1223 (2018).

L. Sveikata, A. Charidimou, A. Viswanathan, Vessels Sing Their ARIAs: The Role of Vascular Amyloid in the Age of Aducanumab. Stroke 53, 298-302 (2022).

M. Thambiseyt, R. Howard, Lecanemab and APOE Genotyping in Clinical Practice -Navigating Uncharted Terrain. JAMA Neurol 80, 431-432 (2023).