The blood vessels in the central nervous system (CNS) have a series of unique properties, termed the blood-brain barrier (BBB), which stringently regulate the entry of molecules into the brain, thus maintaining proper brain homeostasis. Despite the dynamic nature of the brain, the BBB has largely been studied in a static context. We sought to understand if the BBB exhibited plasticity and whether neural activity could regulate BBB properties. Using both chemogenetics and a volitional behavior paradigm, we identified a core set of brain endothelial genes whose expression is regulated by neural activity. In particular, neuronal activity regulates BBB efflux transporter expression and function, which is critical for excluding many small lipophilic molecules from the brain parenchyma. Furthermore, we found that neural activity regulates the expression of circadian clock genes within brain endothelial cells, which in turn mediate the activity-dependent control of BBB efflux transport. These results have important clinical implications for efficiency of CNS drug delivery in the day vs. night, the BBB’s role in clearance of CNS waste products including Aβ in health and Alzheimer’s Disease, and understanding how neural activity can control these and other diurnal processes.

Fibrosis is a common pathological response to inflammation in many peripheral tissues and can prevent tissue regeneration and repair. Despite the potential importance, very little is known about the fibrotic response in the central nervous system (CNS), and how this may impact the regenerative process. We identified persistent fibrotic scarring in the brain and spinal cord following neuroinflammation. Using lineage tracing and single-cell sequencing in the EAE mouse model of multiple sclerosis, we determined that this fibrotic scar is derived from proliferative CNS resident fibroblasts, not pericytes or infiltrating bone marrow-derived cells. Ablating proliferating fibroblasts using the fibroblast-specific expression of herpes thymidine kinase led to an increase in oligodendrocyte lineage cells within inflammatory lesions and a reduction in motor disability. We further identified that interferon gamma pathway genes are enriched in CNS fibroblasts, and the fibroblast-specific deletion of Ifngr1 resulted in a reduced fibrotic scar in EAE. Additionally, we identified that platelet derived growth factor receptor signaling affects brain fibroblast development and fibroblast migration. In the retina, fibrosis occurs following retinopathies and can lead to retinal detachment and blindness, and yet little is known about the origin of this fibrosis as it has been difficult to model in mice. We generated a novel mouse model of oxygen-induced retinopathies that exhibits persistent fibrotic pathology, and used histology and single-cell sequencing to determine that the origin of the fibrotic scar in the retina is likely pericytes. These data delineate a framework for understanding CNS fibrosis.

Throughout the body, the vasculature functions as a delivery system, supplying oxygen and nutrients crucial for cell activity and survival. Central nervous system (CNS) vasculature plays a second critical role as a barrier system, limiting brain entry of bloodborne ions, molecules, and cells. Indeed, CNS blood vessels are endowed with a series of unique properties that allow them to tightly regulate the extracellular environment of the brain, and together these properties are termed the “blood-brain barrier” (BBB). The barrier properties of CNS vasculature are critical for proper brain function; BBB dysfunction during disease or injury can lead to a cascade of consequences resulting in neuronal death. Several questions remain regarding how BBB properties are maintained in health and the mechanisms underlying their dysfunction in neuropathologies.

The brain uses an extraordinary amount of the body’s energy, and without constant blood supply from the vasculature, its functions quickly deteriorate. Indeed, neuronal activity leads to a temporary increase in local blood flow to provide neurons with sufficient energy and oxygen, and this phenomenon is known as neurovascular coupling (NVC). NVC has been shown to involve various cell types and signaling molecules, but it is still unclear to what extent endothelial cells, the cells that form the innermost walls of the vasculature, participate in NVC.

This dissertation explores brain vasculature in health and disease. Chapter I focuses on BBB dysfunction in disease, identifying PDLIM1 as a marker for BBB dysfunction and characterizing its role in CNS endothelial cells. Chapter II focuses on the neurovascular unit, specifically whether microglia play a role in maintaining barrier properties in healthy vasculature. Chapter III demonstrates the dynamic nature of neurovascular cholesterol metabolism and how this metabolic pathway influences NVC.

Wound healing is paramount for the survival of multicellular organisms, but this process can lead to scarring if left uncontrolled. Retinal scarring, in particular, occurs in many prevalent retinal diseases and can be devastating as this pathology leads to blindness. Unfortunately, little is known about retinal scarring, such as the origin of cells producing the scar and the cellular mechanisms leading to scar formation. Understanding these elements behind scar production is vital in creating therapeutic targets that can prevent potential blindness in those affected by retinal scarring. Although there are many retinal disease models in mice, a reliable mouse retinal scarring model has not been established. In this study, I develop two scarring models in mice. One is a mouse model for the disease phthisis bulbi, which is characterized as a shrunken, non-functional eye that is caused by injury, inflammation or other infections. We induce these characteristic symptoms by injecting dispase, a neutral protease, into the eyes of mice, causing degradation of retinal tissue, eye shrinkage, and the production of scars. The other scarring model is termed as Long Oxygen Induced Retinopathy (LOIR), which is a more pathologically severe model of the traditional “Smith” OIR mouse model of retinopathy of prematurity. This newly developed scar producing model is induced by prolonging the exposure of highly concentrated atmospheric oxygen from the traditional “Smith” OIR model protocol of 5 days to 14 days. With these models, I answer what cell type(s) are producing scar tissue in the retina.

The network of blood vessels in the central nervous system is critically important for nutrient delivery and waste removal in higher organisms. A special set of properties distinguishes blood vessels in the central nervous system from those in other organs, because these vessels must provide an extra layer of protection and responsiveness to the real-time demands of neurons. Much is known about way the normal physiology of neurons induces changes in local blood vessels to re-direct nutrient flow and relieve the buildup of waste products. However, there are many potential interactions between neurons and blood vessels about which we know very little. Whether vessels respond to different types of neuronal dysfunction, and if neural activity plays an important role in development of the vascular network are two such unexplored questions. This dissertation is comprised of several projects which all address the underlying mechanisms, consequences, and applications of, neural-vascular interactions in retinal development and disease.

The aim of the first study (Chapter 2) was to determine whether early spontaneous neural activity is the postnatal mouse retina is related to contemporaneous angiogenesis. We found that activity of the cholinergic retinal circuitry was necessary to establish proper layer-specific vascular development and to turn on blood-retinal barrier properties, and elucidated the signaling pathways involved in this process.

The aim of the second study (Chapter 3) was to determine whether specific injury to neurons in the retina would alter transcription in endothelial and microglial cells. We found a number of transcriptomic changes in these cells that may represent RNA originating from non-microvascular cells.

These studies describe new findings concerning the basis for intercellular communication between neurons and blood vessels in the retina, both during development and disease. A better understanding of neurovascular interactions may promote novel therapeutic strategies and biomarker discovery for blinding diseases of development and aging.

The blood-brain barrier (BBB) is a set of properties belonging to endothelial cells that make up blood vessels in the central nervous system (CNS). The BBB’s function is to reduce the transport of nonspecific ions and molecules between the blood and brain and import specific molecules to maintain brain homeostasis. The cerebral vasculature is also known to be tightly coupled to neural activity and dynamically changes blood flow to satisfy the energy demand of active brain regions. Although the cerebral vascular blood flow is tightly coupled to neural activity, the question remains whether BBB properties can dynamically respond to neural activity, as well. Through manipulating glutamatergic neurons in mice, we have shown that neural activity can dynamically regulate both the expression and function of BBB genes in the cerebral vasculature, namely, ATP- binding cassette (ABC) efflux transporters. Additionally, we found that neural activity dynamically regulates circadian clock-related genes in endothelial cells. This led us to ask the question of what is the functional role of the circadian clock in endothelial cells of the cerebral vasculature? Through knocking out a core clock gene, Bmal1, specifically in endothelial cells, we found that the vascular clock regulates ABC efflux transport and animal behavior. We propose the idea that neural activity entrains the vascular-specific circadian clock, which regulates ABC transporter activity in a circadian manner, and deletion of the clock in endothelial cells causes dysregulation of ABC transport, thus disrupting the chemical microenvironment of the brain, resulting in complex behavioral phenotypes.

Human retinal vasculopathies commonly result in breakdown of the blood retinal barrier and formation of scars on the retinal surface. The formation of rigid scar tissue creates traction on the retinal layers that can result in tearing, detachment, and complete blindness. Currently, the only treatment option for retinal fibrosis is surgical removal, an invasive procedure that lacks consistent success at treating the condition. Furthermore, an optimized animal model for retinal fibrosis has yet to be developed. The absence of an animal model severely limits the improvement of patient standard of care. This creates a strong need for a model capable of emulating the clinical progression of retinal fibrosis in a laboratory setting. Mouse models have been proven effective at modeling the angiogenic features of retinal vasculopathies. However, current models have yet to produce the long-term scarring that is also present in a variety of ischemic retinal vasculopathies. Therefore, this study focuses on the scarring and leakage induced by the long oxygen induced retinopathy (l-OIR) as it compares to the classical oxygen induced retinopathy model, referred to as the standard-oxygen induced retinopathy (s-OIR) model. This study primarily focuses on how breakdown of the blood retinal barrier contributes to disease pathophysiology. This was accomplished through an extensive set of leakage assays that follow the disease progression in the mice across an array of time points. It was observed that leakage occurs immediately upon completion of the high oxygen treatment in the l-OIR model. This leakage persists for 7 days and was no longer observed at 14 days. While the s-OIR model leads to a greater acute peak in leakage, the l-OIR model produces leakage over a longer duration as well as a defined fibrotic plaque. This difference may be due to the difference in induced disease severity across the models, with l-OIR producing retinas with more damaged vascular structures, often lacking in central veins and arteries. Furthering scientific and clinical understanding of the role that leakage and scarring play in ischemic retinal vasculopathies will allow future research into bettering the standard of care for patients suffering from such diseases.