Studies on the Mislocalized Exonuclease TREX1 that Causes a Systemic Vasculopathy

Cerebral hereditary angiopathy with vascular retinopathy and impaired other organs caused by TREX1 mutations (CHARIOT) is a rare, autosomal dominant disease of middle-age onset with 100% penetrance that manifests primarily with visual disturbances and neurological symptoms due to small vessel occlusions in the retina and white matter of the brain. Neuroimaging reveals frontoparietal mass-occupying lesions often with substantial edema and scattered white matter hyperintensities. The mean age at diagnosis is 42.9 ± 5.3 years and the disease leads to premature death from relentless neurological decline (mean age of death 53.1 ± 9.6 years). CHARIOT is caused by frameshift mutations in three prime repair exonuclease 1 (TREX1). TREX1, also known as DNase III, is the major mammalian intracellular DNA 3′-5´ exonuclease and the frameshift mutations result in a protein that retains nuclease activity but is diffusely distributed throughout the nucleus and cytoplasm instead of being associated with the endoplasmic reticulum. This thesis work will focus on the clinical phenotype of CHARIOT and on the role of TREX1 mutations in the disease process. CHARIOT was initially described as three distinct diseases, Cerebroretinal Vasculopathy (CRV), Hereditary Vascular Retinopathy (HVR) and Hereditary Endotheliopathy with Retinopathy, Nephropathy and Stroke (HERNS). Although all share the vascular retinopathy, there were some apparent distinctions in the original descriptions. In the initial two HVR papers white matter lesions with were not reported as they were in CRV and HERNS. Furthermore, HVR emphasized migraines and Raynaud\u27s phenomenon while HERNS emphasized other systemic findings such as renal dysfunction. In 2002 linkage mapping confirmed that all 3 mapped to chromosome 3p.21 and in 2007 they were shown to be caused by frameshift mutations in TREX1. The first part of this thesis work focuses on characterizing the natural history of CHARIOT in the families originally reported in the CRV, HVR and HERNS papers as well as in more recently ascertained families. We point out that the disease has a remarkably similar course in all eleven families studied with respect to the vascular retinopathy and white matter lesions as well as to the more systemic findings such as hepatic and renal dysfunction. The histopathology demonstrates a systemic vasculopathy with luminal narrowing and multi-laminated basement membranes in all subjects. We also propose diagnostic criteria to aid in the recognition of CHARIOT as many cases are initially misdiagnosed. The second part of this thesis focuses on evaluating the expression of TREX1 in human brain tissue from normal controls and subjects with CHARIOT or ischemic stroke. Using a rabbit polyclonal antibody raised against recombinant TREX1 by our laboratory we confirmed the expression of the wild-type and V235fs mutant TREX1 in EBV transformed B-cell lines by Western blot. The antibody was also used for immunofluorescent staining of normal human brain tissue which demonstrated for the first time that TREX1 is expressed exclusively by some, but not all, microglia with the relative abundance varying across brain regions. Further immunohistochemical staining showed that TREX1+ microglia are often found adjacent to the microvasculature. Thus, the TREX1+ cells identify a subset of microglia that may have a role in vascular homeostasis. In the cerebral cortex of normal controls, TREX1+ cells are more prevalent in the gray matter (21.6 ± 5.5% of TREX1+ cells in white matter) which is in contrast to patients with CHARIOT where the majority are found in the white matter (83.5 ± 2.3% of TREX1+ cells in white matter), commonly within clusters in histologically normal appearing tissue. As lesions in CHARIOT are exclusively found in the white matter, this distribution may reflect damage initiated by the TREX1+ microglia or a response to early injury. In histologically damaged white matter TREX1 is expressed along the periphery of the lesion by activated microglia and/or macrophages. Moving away from the lesion towards normal tissue, the cell morphology progressively changes to that of ramified, resting microglia. TREX1 staining in various neuropathologies showed that the protein is upregulated in ischemic stroke in a pattern similar to that observed around white matter lesions in CHARIOT. Thus, in CHARIOT and ischemic stroke, TREX1 may play a role in the phagocytosis and clearance of necrotic debris. The third part of this thesis work focuses on developing a knock-in mouse model of CHARIOT. The first was a knock-in introducing a missense mutation in murine Trex1 that mimics the most common mutation found in CHARIOT, V235fs. Although two chimeras from two different embryonic stem cell clones were obtained, both died before they could breed. The second approach was a conditional knock-in of human TREX1 into the mouse. In addition to the V235fs associated with CHARIOT, mice carrying the wild-type human allele and the D272fs associated with systemic lupus erythematosus were also created. Mice expressing the wild-type human allele have not been generated thus far. Mice heterozygous for the D272fs and V235fs mutations and homozygous for the D272fs mutation have been aged from 6 to 20 months but no phenotype has yet been observed. Overall, the work presented in this thesis furthers our understanding of CHARIOT. The first part provides a comprehensive overview of the clinical, pathological and radiological findings in CHARIOT in all known families diagnosed with the disease. This report will increase awareness of the disease and facilitate making an accurate diagnosis. The second part defines the expression of endogenous TREX1 in human brain to a subset of microglia. As TREX1+ microglia are commonly found in close association with the vasculature in normal tissue, further work clarifying its exact function may lead to an understanding of how a functional but mislocalized exonuclease leads to the microvasculopathy observed in CHARIOT. In areas of damaged brain tissue we propose that TREX1 in activated microglia and infiltrating macrophages aids in the removal of necrotic debris. In the third part of this work we attempt to develop a mouse model to elucidate the role of the frameshift mutations in TREX1 but no definitive phenotype has so far been detected