Investigating the role of pericytes in multiple sclerosis by inducing experimental autoimmune encephalomyelitis in genetically modified mouse models

Background: Pericytes are mural cells positioned on the abluminal side of endothelial cells. They have crucial roles in blood-brain-barrier (BBB) maturation and maintaining immune homeostasis in the central nervous system (CNS). Recent studies of pericyte cell cultures and well-characterized pericyte deficiency mouse models suggest that pericytes also have immunomodulatory functions. However, the animals used in these studies are inborn pericyte-deficient mouse models that already have a deficient BBB formation and a disrupted microvasculature morphology. Furthermore, partial ablation of pericytes in developing mice leads to increased infiltration of leukocytes to the CNS in later life, both in steady-state and pathological conditions. These abnormalities might confound the actual findings related to the pericyte function in neuroinflammatory diseases. Therefore, studying the neuroinflammatory roles of pericytes by using the existing pericyte deficiency models is incompetent in making confident conclusions about the findings. The most ubiquitous neuroinflammatory CNS disease is Multiple Sclerosis (MS). It is a disease of autoimmunity in which the peripheral inflammatory cells infiltrate the brain after the opening of BBB, resulting in the degeneration of myelin sheaths, oligodendrocytes, and underlying nerve fibers. The players initially causing the preternatural passage of these immune cells to the CNS parenchyma are not clearly defined yet. Besides, the line of events happening during the pathogenesis of the disease is also not well-known. Aim: In this study, we first aimed to generate and characterize a tamoxifen-inducible pericyte ablation model in adult mice that have a properly developed vasculature and mature BBB. Secondly, we sought to investigate the effect of pericyte loss in MS pathogenesis through the most frequently used model of MS-like neuroinflammation, the experimental allergic encephalomyelitis (EAE) model. Methods: To create our inducible pericyte ablation model, PDGFRβ-PA2-CreER+/- mice (JAX Mice, 030201) were crossed with Rosa26-DTA176+/+ mice (JAX Mice, 010527). After genotyping, 9-10 weeks old progeny was intraperitoneally injected with 100 mg/kg tamoxifen for two (2X), three (3X), or five (5X) consecutive days to induce the expression of diphtheria toxin in Pdgfrβ-positive pericytes. PDGFRβ-Cre+/DTA176+ (Cre-positive) mice were characterized as pericyte ablation group, and PDGFRβ-Cre-/DTA176+ (Cre-negative) mice were used as controls. To characterize our model, the level of pericyte coverage, Pdgfrβ gene expression, and vasculature changes were examined in the brain and spinal cord tissues of both groups. To do so, immunohistochemistry, quantitative PCR, and AngioTool software were used, respectively. The analyses were separately done for the short-term (15 days) and long-term (60 days). Additionally, behavioral tests of spatial recognition memory (Y-maze performance) and locomotor activity (Grip Strength Test) were run. According to the results of the characterization experiments, the pericyte ablation group that was administered with two consecutive days of TAM injections (2X TAM) was selected for further EAE experiments. EAE was induced by MOG35-55 EAE kit (Hooke Labs, EK2110) 15 days after TAM injections. Cre-negative and wild-type C57BL/6 mice were used as pericyte ablation controls for EAE experiments, and they together were considered as "wild-type mice" when discussing the results. After EAE induction, weight and clinical symptom variations were monitored and scored for 28 days. Statistical analysis was done by repeated-measures 2-way ANOVA multiple comparison test. Results: There was a 75% (+-SEM=0.9192) decrease in pericyte coverage of the cortex (p<0.01) and a 40% (+-SEM=2.385) decrease in the pericyte coverage of the spinal cord (p<0.01) in the 2X TAM pericyte ablation group mice (n=8) compared to the Cre-negative controls (n=7). Similarly, Pdgfrβ gene expression was decreased to 10 % (+-SEM=2.230) in the cortex (p=0.036) and 60% (+-SEM=4.511) in the spinal cord (p=0.036). The vessel analysis and the behavioral analysis revealed no significant differences between the 2X TAM pericyte ablated group and Cre-negative controls. Based on these results, pericyte ablated mice created with 2X TAM were chosen to be used in EAE experiments. During EAE, the pericyte ablated mice (n=7) and the wild-type mice (n=23) had similar clinical progression up to score 1.5 until the 15th day of the EAE induction. However, after that time point, the symptoms of the pericyte ablated mice started to worsen with a higher exacerbation rate. This difference got statistically significant after Day 20 and continued till the end of the experiment (p= 0.0141, 0.0056, 0.008, 0.0276, and 0.0188 at Day20-28). Concurrently, starting with the first symptoms of EAE, pericyte ablated mice also lost more weight than the wild-type mice (p= 0.0411, 0.0067, 0.0236, 0.0093, 0.01, 0.0194, and 0.0081 at Day8 and Day16-28). On the other hand, the immunohistochemical analysis of the spinal cord tissues revealed that the number and the size of inflammatory lesions were more and larger in the pericyte ablated mice compared to the lesions of the wild-type mice. Additionally, the periventricular region of the brain presented unprecedented inflammatory foci in pericyte ablated mice, while the wild-type EAE mice did not show any cell accumulations in the brain. Conclusion and Comment: This project has generated and characterized a TAM inducible pericyte ablation model in adult mice. The advantages of this new generation pericyte ablation model over the other pericyte deficiency models are that pericyte ablation amount and ablation induction time could be controlled, and while so, the model does not require the injection of any toxin. In this project, by using our pericyte ablation model in combination with the EAE model, pericytes were shown to play a protective role during the resolution of neuroinflammation. Key Words: Blood-Brain Barrier, Pericytes, Pericyte Ablation, Pdgfrβ, Multiple Sclerosis, Experimental Allergic Encephalomyelitis, Transgenic Mouse Models