Investigating Lyssavirus CNS Infection and Control with a Monoclonal Antibody in vivo

Overview

Infections with rabies virus (RABV) and related lyssaviruses are uniformly fatal once virus accesses the central nervous system (CNS). Current immunotherapies are thus focused on the early, pre-symptomatic stage of disease, with the goal of peripheral neutralization of virus to prevent CNS infection. Previous lab research found that a single dose of F11 monoclonal antibody limits Australian Bat Lyssavirus (ABLV) load in the brain and reverses any signs of progression for the disease following lyssavirus infection, even when administered after virus replication in the CNS. However, the mechanism by which F11 controls ABLV infection is not known. Here, I investigated if F11 is capable of entering the CNS during infection and identifying brain resident cells that are infected by ABLV . For testing purposes, either the mice were infected or uninfected, and they were either untreated, treated with F11 three days after infection, or treated with F11 five days after infection. Via fluorescence-based antibody stain analysis, I found that F11 is capable of entering the CNS suggesting that the blood brain barrier is impaired during ABLV infection. Using histology and antibody staining I also found that astrocytes and neurons within the hindbrain and cerebellum were infected with ABLV . These discoveries suggest that immunotherapy may be efficacious in human patients even after ABLV , a lethal neurotropic virus, has entered the CNS. PreVis: Real-time Motion Forecasting Using LiDAR Technology Daniel Mathew Poolesville High School, Poolesville, MD LiDAR is shaping the future as we know it. From its uses in a wide variety of industries such as topographical surveying, medical applications, transportation aids, and law enforcement, it offers accurate and consistent results. T o analyze 3D and 2D environments, however, current solutions have either been power-consuming or very expensive. The proposed solution to this problem, called PreVis, uses a single LiDAR to scan its environment. The output is combined with a novel 2-step spatial localization and motion-forecasting algorithm that allows for the motion of an object to be tracked and predicted, effectively "filling in" for any missing information. The localization algorithm has been pruned with over 40,000 computer-simulated tests, showing nearly 100% success in finding objects. The Motion-Forecasting algorithm is split into Kalman filtering for error suppression and cartesian/parametric polynomial fitting for path recognition of a moving object. Polynomials with up to 30 degrees are tested for multicollinearity and overfitting using MSE and limiting degrees of freedom. In the end, the localization algorithm was able to collect on average 48 data points in a 10-second window for a moving person. These data points were fed into the motion-forecasting algorithm. The best algorithm was the 2-degree parametric polynomial fitting with a mean squared error of 17 but limited overfitting. This concept can be used as a guide for those with disabilities, enhanced collision avoidance systems in automobiles, and much more. PreVis opens the door to inexpensive and highly accurate motion prediction using LiDARs. MICHIGAN

Competition history

• JSHS 2023 Category not listed

Resources

• National JSHS abstract book ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.