Background:

According to the Centers for Disease Control and Prevention, there are approximately 30,000 new cases of Lyme Disease (LD) in the United States every year³. In its early stages, LD presents with flu-like symptoms, including headache, fatigue, and achiness⁶. If left untreated or otherwise allowed to progress, symptoms of infection can worsen, including arthritis and neurological damage. Consequently, early and accurate diagnosis of LD is critical.

Among the most widely used diagnostic tests for LD is the Western blot. When B. burgdorferi - the causative agent of LD - establishes an infection in a host organism, the host organism produces antibodies that bind proteins on the surface of the B. burgdorferi organism. Western blotting essentially detects the presence of these antibodies in the host organism’s blood. Current diagnostic standards for detection of early-stage LD infection (IgM) look for antibodies that bind to 3 specific proteins, with the presence of antibodies against any 2 of them being sufficient for LD diagnosis⁵.

One of these 3 proteins is outer surface protein C (ospC), which is highly expressed by B. burgdorferi in the early stages of human infection. OspC is also a highly variable protein, with many different versions (called alleles) known to exist. Current Western blot tests typically only look for antibodies against OspC type A in patients’ serum. Antibody cross-reactivity between various OspC types can be as low as 38%¹This presents a clear problem: a patient that has been infected by B. burdorferi with OspC type K may receive a negative result for a test which only looks for antibodies against type A, resulting in delayed treatment.

Approach:

Developing a test that captures a greater proportion of OspC type diversity requires a better understanding of which specific epitopes (antibody-binding regions) are responsible for the observed differential antibody reactivity for different OspC types; these epitopes will herein be referred to as critical epitopes. Baum et al.¹ describes an experimental approach in which correlations between local amino acid sequence identity (Ms) and antibody cross-reactivity (MD) for many different OspC types (considered pairwise) were used to identify potential critical epitopes, which are those loci with the highest r(Ms, MD) scores; the full set of such scores will herein be referred to as a critical epitope profile.

This study will attempt to glean further potentially-useful information from the Baum dataset to help solve the diagnostic test problem. Specifically, this study will assess whether or not the subset of 4 specific OspC types which are known to be dominant in invasive strains of B. burgdorferi (namely types A, B, I, and K) have a different critical epitope profile than the remaining types, many of which may not be clinically-relevant. Given that clinically-relevant OspC alleles are under a set of different (and likely more intense) evolutionary pressures from the human host immune system, it is reasonable to infer that the critical epitope profile obtained by Baum may not be accurate for the subset of invasive strains which are relevant to our goal of optimizing LD diagnostic testing.