Graduation Date

Summer 8-14-2026

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Programs

Integrative Physiology & Molecular Medicine

First Advisor

Howard E Gendelman

Second Advisor

R Lee Mosley

Abstract

The pathobiology of Alzheimer’s disease (AD) and human immunodeficiency virus type 1 (HIV-1) infection overlaps in aged people living with HIV-1 (PLWH). Although antiretroviral therapy has extended the lifespan of PLWH, persistent low-level viral replication, chronic neuroinflammation, immune dysregulation, and the specific depletion of CD4+ T cells contribute to cognitive impairments. As HIV-1 infection has become a chronic age-associated condition, co-morbidities that include Alzheimer ’s-like dementia have become increasingly prevalent. However, studies of HIV-1-AD are hampered by the lack of Alzheimer’s disease mouse models that support productive HIV-1 replication.  In-kind, most HIV-1 models lack AD-relevant genetic mutations that are essential for the development of AD pathology, as mice do not naturally develop Alzheimer’s disease.  Conversely, current AD models are developed on immunocompetent murine backgrounds, which preclude human immune reconstitution required to permit HIV-1 replication.

In humans, AD is largely idiopathic and develops over decades. Only a small fraction of cases are linked to familial AD mutations (FAD). Current transgenic AD models rely on random genomic insertion and overexpression of FAD mutation-bearing transgenes driven by exogenous promoters to accelerate the development of pathology and compress the clinical timeline within the mouse lifespan. Although useful for dissecting AD pathobiology and therapeutic testing, transgenic AD models can demonstrate nonphysiological gene expression, ectopic pathology, and overexpression-associated artifacts that limit studies of early-stage AD pathogenesis and human immune interactions. To overcome this, knock-in mouse models have recently been developed in which FAD mutations are knocked into mouse homologs and gene expression is driven by endogenous promoters. However, the immunocompetent background of these mice does not support human immune reconstitution, limiting their use for studies of HIV-1 replication in the setting of AD pathology.

With the need to develop an HIV-1-AD mouse model, we created a novel knock-in (KI) AD mouse on an immunodeficient NOG background. Using CRISPR-Cas9 gene editing, in an early study, we developed founder lines of single KI NOG mice carrying APPKM670,671NL (Swedish) and PS1M146V FAD mutations. The single APP KI mice (termed NA) were crossed with the single KI PS1 mice to generate a double KI NAPS mouse. The knock-in allows gene expression driven by endogenous promoters, recapitulating early AD pathology while avoiding overexpression artifacts associated with transgenic AD models. Both NA and NAPS mice showed physiological FAD expression, increased APP-CTF-β production, and early intraneuronal human amyloid-β accumulation in the cortex and hippocampus. Importantly, amyloid pathology was predominantly intraneuronal rather than plaque-based and was accompanied by amyloid-associated microgliosis, extensive neuronal loss, dendritic injury, and brain atrophy. This phenotype makes these mice particularly well-suited for studying early AD-like disease stages, where intraneuronal amyloid accumulation, synaptic dysfunction, and neuronal injury may precede overt extracellular plaque deposition. The immunodeficient NOG background allows human immune reconstitution, which facilitates modeling of human-specific comorbidities, including systemic HIV-1 infection, for studies of AD pathologies.

To further improve this model and enable productive HIV-1 infection in the brain, in this study, the APP KI NA line was crossed with human IL-34 (named IL) transgenic NOG mice to generate the NOG/APPKM670,671NL/IL-34 mice (termed hNAIL). Following neonatal CD34+ hematopoietic stem cell (HSC) reconstitution, the human IL-34 transgene supports the development of human microglia-like cells in the brain. These resident human myeloid cells support HIV-1-replication. This allows productive brain HIV-1 infection together with APP knock-in amyloid pathology. Four-month-old humanized hNAIL mice were infected with macrophage-tropic HIV-1ADAstrain and evaluated at eight weeks for viral replication, amyloid pathology, neuroinflammation, neuronal and synaptic integrity, and spatially resolved cell type-specific transcriptional alterations. The hNAIL mice developed productive HIV-1 replication both in the peripheral immune tissues and brain. Compared to uninfected hNAIL controls, HIV-1-infected mice showed increased amyloid-β accumulation, enhanced microglial activation, and synaptic and neuronal marker reduction. Spatial transcriptomic profiling revealed distinct transcriptional signatures associated with Aβ and HIV-1 pathology, while regions containing both Aβ and HIV-1 showed amplified disease-associated responses. Neuronal compartments exhibited the strongest transcriptional alterations, including pathways linked to neuroinflammation, protein trafficking, synaptic dysfunction, and neurodegeneration. Together, these studies establish a first-of-its-kind humanized knock-in AD mouse model that enables productive HIV-1 infection in the brain within an AD-relevant genetic and neuroimmune environment.

In conclusion, the NA and NAPS mouse models represent novel knock-in AD models developed on an immunodeficient NOG background. These mice develop intraneuronal amyloid pathology and extensive neuronal loss, both hallmarks of early AD, which is not commonly recapitulated in current models. The hNAIL AD model represents a further advancement of the NA mice by introducing HIV-1- susceptible human microglia-like cells into the mouse brain. This integrated system provides a novel platform to evaluate how brain viral persistence, microglial activation, amyloid accumulation, synaptic injury, and neuronal loss contribute to accelerating AD-like disease. Together, these studies establish the feasibility and pathological relevance of humanized knock-in AD models for HIV-1 research and open several future directions, including longitudinal studies of disease progression, mechanisms of human immune-amyloid interactions, effects of antiretroviral therapy on CNS pathology, and testing of immune-modulating or disease-modifying interventions for HIV-associated AD and related dementias.

Rights

The author holds the copyright to this work and any reuse or permissions must be obtained from the author directly.

Download

Available for download on Saturday, June 10, 2028

Share

COinS