The dynamics and nature of the hierarchy of VSG expression during Trypanosoma brucei infection

Trypanosoma brucei is a tsetse fly-transmitted kinetoplastid protozoan that parasitises a wide range of mammals in sub-Saharan Africa. The chronic infections typical of trypanosomes have a profile characterised by intermittent parasitaemic peaks, that become smaller and further apart. A key component that facilitates the survival of the parasite within the bloodstream of the host, and the generation of chronic infections, is antigenic variation. This process involves the inherent switching of the surface-expressed variant surface glycoprotein (VSG), and the new variable antigen types (VATs) in order to preempt the host immune response. There are approximately 1,000 genes in the VSG repertoire, and the observation that certain VATs appear at similar times within infections, has led to the conclusion that the expression of VSGs is semi-predictable and semiordered. This hierarchical system allows optimal use of the VSG repertoire, and is highly likely to be a significant factor in generating chronicity. It has been established that VSGs encoded by subtelomeric genes tend to appear early in infection, whilst those encoded by chromosomal internal genes tend to appear later. The first aim of this thesis was to examine the timing of expression of a subset of VSGs within replicate infections in mice and cattle, and to link the timing of expression to the genetic locus of the silent, donor, VSG gene. The VSGs examined represented all VSG locus types, incorporating 1 bloodstream expression site (BBS) gene (the transcription site for VSG genes in the bloodstream stage), 2 metacyclic expression site (MBS) genes, 2 minichromosomal genes, and 2 chromosomal internal genes. The infections were initiated with pleomorphic trypanosomes that switch at a high rate, and are a close-to-field, 'wild type' strain. The infections in mice confirmed that there was a statistically significant difference in the timing of onset of VAT-specific immune responses across replicate batches of infections with two separate analyses; one by comparing the average time of onset of the immune response (General Linear Model [GLM]; F6,106=7.49, p < 0.0001), and the second by ranking the onset of the immune response by sequence of appearance (GLM; Fe,112=8.03, p