Soluble brain homogenates from diverse human and mouse sources preferentially seed diffuse Aβ plaque pathology when injected into newborn mouse hosts

A C1qTNF3 collagen domain fusion chaperones diverse secreted proteins and anti-Aβ scFvs: Applications for gene therapies

Enhancing production of protein cargoes delivered by gene therapies can improve efficacy by reducing the amount of vector or simply increasing transgene expression levels. We explored the utility of a 126-amino acid collagen domain (CD) derived from the C1qTNF3 protein as a fusion partner to chaperone secreted proteins, extracellular "decoy receptor" domains, and single-chain variable fragments (scFvs). Fusions to the CD domain result in multimerization and enhanced levels of secretion of numerous fusion proteins while maintaining functionality. Efficient creation of bifunctional proteins using the CD domain is also demonstrated. Recombinant adeno-associated viral vector delivery of the CD with a signal peptide resulted in high-level expression with minimal biological impact as assessed by whole-brain transcriptomics. As a proof-of-concept in vivo study, we evaluated three different anti-amyloid Aβ scFvs (anti-Aβ scFvs), alone or expressed as CD fusions, following viral delivery to neonatal CRND8 mice. The CD fusion increased half-life, expression levels, and improved efficacy for amyloid lowering of a weaker binding anti-Aβ scFv. These studies validate the potential utility of this small CD as a fusion partner for secretory cargoes delivered by gene therapy and demonstrate that it is feasible to use this CD fusion to create biotherapeutic molecules with enhanced avidity or bifunctionality.

Aβ Amyloid Scaffolds the Accumulation of Matrisome and Additional Proteins in Alzheimer's Disease

We report a highly significant correlation in brain proteome changes between Alzheimers disease (AD) and CRND8 APP695NL/F transgenic mice. However, integrating protein changes observed in the CRND8 mice with co-expression networks derived from human AD, reveals both conserved and divergent module changes. For the most highly conserved module (M42, matrisome) we find many proteins accumulate in plaques, cerebrovascular amyloid (CAA), dystrophic processes, or a combination thereof. Overexpression of two M42 proteins, midkine (Mdk) and pleiotrophin (PTN), in CRND8 mice brains leads to increased accumulation of A β ; in plaques and in CAA; further, recombinant MDK and PTN enhance A β ; aggregation into amyloid. Multiple M42 proteins, annotated as heparan sulfate binding proteins, bind to fibrillar A β 42 and a non-human amyloid fibril in vitro. Supporting this binding data, MDK and PTN co-accumulate with transthyretin (TTR) amyloid in the heart and islet amyloid polypeptide (IAPP) amyloid in the pancreas. Our findings establish several critical insights. Proteomic changes in modules observed in human AD brains define an A β ; amyloid responsome that is well conserved from mouse model to human. Further, distinct amyloid structures may serve as scaffolds, facilitating the co-accumulation of proteins with signaling functions. We hypothesize that this co-accumulation may contribute to downstream pathological sequalae. Overall, this contextualized understanding of proteomic changes and their interplay with amyloid deposition provides valuable insights into the complexity of AD pathogenesis and potential biomarkers and therapeutic targets.

Modeling the Competition between Misfolded Aβ Conformers That Produce Distinct Types of Amyloid Pathology in Alzheimer's Disease

The amyloid pathology characteristic of Alzheimer's disease (AD) can be broadly classified as either fibrillary amyloid or diffuse amyloid. Fibrillary amyloid is found in cored-neuritic deposits, fibrillar deposits, and vascular deposits, and binds strongly to the amyloid revealing dyes Thioflavin-S or Congo Red. Diffuse amyloid can appear as wispy dispersed deposits or compact tufted deposits dispersed in neuropil, and binds amyloid dyes weakly if at all. In AD brains, both types of pathology are detected. Homogenates from AD brains, or the brains of transgenic mice modeling AD-amyloidosis, have been used to seed pathology in vulnerable host transgenic models. These studies suggest that pathologies may arise from distinct conformers or strains of misfolded Aβ, similar to propagating prions. Using Aβ strains sourced from four different AD-amyloidosis models, we injected pathological seeds into the brains of newborn mice from three different transgenic hosts with distinctive Aβ pathologies. Two of the seeding sources were from mice that primarily develop cored-neuritic Aβ deposits (cored strain) while the other two seeding sources were from mice that develop diffuse Aβ deposits (diffuse strain). These seeds were injected into host APP mice in which the resident strain was either diffuse or cored-neuritic pathology. Seeding-homogenates were injected into the brains of newborn mice to initiate propagation as early as possible. Depending upon the level of transgene expression in the host, we show that the injected strains of misfolded Aβ from the seeding homogenate were able to outcompete the resident strain of the APP host model. In serial passaging experiments, it appeared that the diffuse strain was more easily propagated than the cored strain. Collectively, our studies align with the idea that different types of Aβ pathology in AD brains arise from different populations of Aβ conformers that compete to populate the brain.