Late‑onset Alzheimer’s disease (LOAD) accounts for more than 95% of Alzheimer disease cases and presents heterogeneous genetic and environmental risk factors. Traditional transgenic overexpression mouse models have off‑target effects and fail to reproduce the clinical heterogeneity of LOAD, limiting their translational value for preclinical therapeutic development. The Model Organism Development and Evaluation for Late‑onset Alzheimer’s Disease (MODEL‑AD) Consortium aims to develop novel mouse strains that better model human‑relevant risk architecture for LOAD.
Methylenetetrahydrofolate reductase (MTHFR) is an enzyme in the folate/methionine pathway. Human variants in MTHFR, notably the 677C>T allele, have been associated with Alzheimer’s disease and related dementias (ADRD). Prior work showed that an Mthfr677C>T mouse model phenocopies some features observed in humans carrying the variant, including cerebrovascular deficits. The present study tests how homozygosity for the Mthfr677C>T variant affects disease‑relevant signatures when combined with humanized LOAD risk alleles.
MODEL‑AD researchers created a novel strain on the C57BL/6J (B6) background that is homozygous for the Mthfr 677C>T allele and carries humanized versions of key Alzheimer’s risk genes: humanized Abeta, APOEε4, and Trem2*R47H. The combined genotype is referred to as LOAD2.Mthfr677C>T. This approach embeds the Mthfr variant in a genetic context intended to reflect multiple human LOAD risk factors without relying on overexpression of amyloid precursor constructs.
Mice were assessed across multiple ages with repeated behavioral measurements and biometric sampling performed longitudinally up to 24 months of age. At defined time points, blood and brain tissue were collected for multiomic analyses, including transcriptomics and proteomics, as well as for correlation with human disease signatures and neuropathological evaluation. The study design emphasizes capturing prodromal and age‑dependent molecular changes rather than end‑stage amyloid pathology alone.
Comparative molecular profiling of brain tissue from LOAD2.Mthfr677C>T mice versus LOAD2 control mice identified transcriptional and proteomic signatures that map to biological processes relevant to Alzheimer’s disease. Specifically, alterations were observed in pathways and molecular markers associated with the cerebrovasculature, myelination, and synaptic biology. These molecular changes emerged despite the absence of overt amyloid deposition or marked neuroinflammatory changes in the model, indicating that the Mthfr677C>T variant modulates early disease‑related biology in vascular and white matter compartments as well as synaptic systems.
Neuropathological assessment reported that LOAD2.Mthfr677C>T mice lacked hallmark pathologies commonly emphasized in many AD models, such as overt amyloid plaque deposition and significant neuroinflammation, when compared to LOAD2 controls. The study therefore highlights prodromal molecular signatures rather than classical histopathological endpoints. Details on quantitative neuropathology measures or regional distributions were not reported in the abstract of the source document.
The transcriptional and proteomic changes in LOAD2.Mthfr677C>T brains were reported to resemble patterns seen in human LOAD patients, particularly those linked to cerebrovascular compromise, myelination deficits, and synaptic alterations. This correspondence supports the relevance of the model for studying aspects of ADRD that are not directly tied to amyloid aggregation but may contribute to early disease processes and clinical heterogeneity.
The authors conclude that the LOAD2.Mthfr677C>T mouse strain provides a useful platform to study prodromal disease signatures relevant to late‑onset Alzheimer’s disease, especially cerebrovascular compromise, myelination changes, and synaptic biology. Because the model combines the Mthfr 677C>T variant with humanized Abeta, APOEε4, and Trem2*R47H alleles on a B6 background, it aims to represent a more human‑relevant genetic context without the confounds of transgenic overexpression.
The findings emphasize molecular pathways outside of classical amyloid pathology that may be important in LOAD pathogenesis and could inform preclinical investigations targeting vascular and white matter contributors to cognitive decline. The study is presented as a preprint and has not been peer reviewed. Funding reported includes NIH Common Fund support under U54 AG054345. The authors declared no competing interests.
(Notes: The source abstract did not report detailed numerical results, statistical measures, regional neuropathology quantification, or specific gene/protein lists; those details are therefore not included here.)