Dnmt3a and Tet2-Driven Clonal Hematopoiesis Have Opposing Effects on the Pathogenesis of Alzheimer’s Disease

The clonal expansion of mutant hematopoietic stem cells (HSCs) known as clonal hematopoiesis is associated with an increased likelihood of developing many age-associated diseases. However, the connection to neurodegenerative diseases including Alzheimer’s Disease (AD) is less clear. Prior studies have established that inflammation plays a central role in AD pathogenesis. Mounting evidence suggests systemic inflammatory signals can be transmitted to the central nervous system (CNS), where they may play a direct role in microglia activation and beta-amyloid plaque clearance. Here we investigated the role of the two most commonly mutated clonal hematopoiesis-associated genes, DNMT3A and TET2, in the pathogenesis of AD using a mouse model.

We utilized 5xFAD (Familial Alzheimer’s Disease) mice which express human disease variants of amyloid precursor protein and presenilin and typically develop beta-amyloid plaques and cognitive decline by 6-9 months of age. At 6-8 weeks of age, 5xFAD mice were non-competitively transplanted with Dnmt3a^-/-, Tet2^-/- or wildtype (WT) bone marrow (BM). Mice were then challenged weekly with LPS to mimic systemic chronic inflammation seen in aging. At 6 months of age, transplanted mice were assessed for signs of AD pathogenesis as well as the presence of infiltrating peripheral immune cells within the brain.

5xFAD mice transplanted with Dnmt3a^-/- BM displayed exacerbated AD, including worsened cognitive impairment and decreased microglia activation compared to those transplanted with WT BM. Mice transplanted with Dnmt3a^-/- BM also had fewer peripheral immune cells infiltrating the brain compared to WT-transplanted recipients. In contrast, 5xFAD mice transplanted with Tet2^-/- BM showed improved cognitive status, decreased amyloid plaques, and increased microglia activation. Unlike mice transplanted with Dnmt3a^-/- BM, Tet2^-/- transplanted mice had a higher percentage of activated infiltrating myeloid cells within the brain compared to WT controls. Taken together, these data suggest that loss of Dnmt3a and Tet2 exert opposite effects on AD pathogenesis, with divergent impact on peripheral immune cell infiltration to the brain. We hypothesize that Tet2^-/- peripheral immune cells can more effectively infiltrate the CNS, leading to enhanced microglia activation, better amyloid plaque clearance and preserved cognitive function.

To assess whether peripherally derived immune cells directly replace microglia in the CNS, we performed flow cytometry to identify donor-derived peripheral immune cells that were positive for microglia marker Tmem119 within the brain. Interestingly, mice transplanted with Tet2^-/- BM, had a higher proportion of donor-derived mutant microglia-like cells when compared to mice transplanted with WT BM. While the absolute number of peripherally derived microglia-like cells was low, these data indicate that microglial replacement from the periphery can occur.

To evaluate the impact of DNMT3A or TET2 mutation on human microglial function, we utilized microglia derived from isogenic human iPSC lines harboring DNMT3A (DNMT3A^R882H/WT) or TET2 mutations (TET2^DelE3-E11/WT). Cytokine bead array of culture supernatant showed that TET2-mutant induced microglia (iMGs) released significantly more inflammatory cytokines following LPS treatment than either WT or DNMT3A-mutant iMGs. TET2-mutant iMGs also demonstrated a greater capacity to phagocytose both myelin and beta-amyloid upon LPS activation compared to WT or DNMT3A-mutant iMGs. As myelin damage likely precedes and contributes to amyloid plaque buildup in early AD, TET2-mutant microglia may prevent disease through clearance of both damaged myelin and beta-amyloid plaques.

In summary, we found that loss of Tet2 provided protection against AD through increased myeloid cell infiltration and microglia activation, leading to more efficient amyloid plaque clearance and improved cognitive performance. In contrast, the opposite was true for loss of Dnmt3a. Our study of clonal hematopoiesis and AD marks the first report in which DNMT3A and TET2, which have opposite roles in DNA methylation, induced opposing effects on disease progression. This work has implications for our understanding of clonal hematopoiesis and AD pathogenesis and confirms that peripheral immune cells can have direct effects on neuroinflammation and immunologic control of neurodegenerative disease.