Most infections don’t usually cause prolonged illness in humans because the body’s immune system recognizes the presence of a molecular fragment made by the pathogen, termed an antigen, as alien, and triggers a defence response that eliminates the pathogen. However, pathogens use a range of strategies to evade such destruction. One approach is called antigenic variation, whereby a pathogen population keeps changing the antigens that are expressed. If antigenic variation occurs more rapidly than the host can respond to a newly expressed antigen, infection can persist. Writing in Nature, Müller et al.1 report that in the parasite Trypanosoma brucei, the structure of the DNA–protein complex known as chromatin has a role in how antigenic variation occurs in this organism.
The process of antigenic variation has evolved independently in many organisms2–5. It has certain common features, such as the presence of a reservoir of many versions of a particular gene, and hence the possibility that many different antigens can be expressed that correspond to that gene or gene family. Another aspect central to infection persistence is the presence of mechanisms to ensure that only one version of such a gene is expressed at a time, with all the other versions existing in a silenced state that might later be reversed6.
Antigenic variation has been studied intensively in T. brucei, which causes African trypanosomiasis, historically known as sleeping sickness, in humans, and a range of diseases in livestock. The disease can be fatal if trypanosomes enter the brain, causing a range of neurological symptoms that including the disturbance of sleep patterns7. Although the incidence of the human disease is in decline8, the animal illness remains a major cause of poverty among farmers in sub-Saharan Africa9.
The surface of a T. brucei trypanosome is covered with closely packed molecules of a glycoprotein termed VSG (Fig. 1). During an infection, switching events occur that result in a different version of the VSG being expressed from a reservoir of thousands of VSG genes, most of which are substantially different from each other10,11. This switching process enables the parasite to evade immune-mediated destruction, and the infection can thus persist for decades12.
Most of the parasite’s VSG-encoding gene repertoire occurs in tandem arrays close to DNA sequences called telomeres, which are found at the ends of chromosomes13; these arrays are known as subtelomeric arrays. In addition, at any given time, approximately 15 other VSG-encoding genes — including the one being expressed6 — are present in expression sites. These are regions of chromosomes next to telomeres that are specialized for the expression of VSG-encoding genes. Only one expression site is active, and it is located in a nuclear structure termed the expression-site body14. The other expression sites are inactive, and all the genes are said to be silent. Antigenic variation can occur either by a change in the sequence of the VSG gene in the active expression site through a DNA-mediated process called recombination5, or by the replacement of one expression site with another in the expression-site body5.
The processes involved in gene silencing must operate on all copies of the VSG-encoding gene apart from the one being expressed. Müller and colleagues’ study addressed three questions about this process. How are the subtelomeric arrays of VSG-encoding genes kept silent? Is the same mechanism used for all the silenced expression sites? And how is this silencing reversed?
Müller and co-workers report a newly generated assembly of the T. brucei genome that adds substantially to the one previously reported15. The authors reconstructed 33 subtelomeric arrays of VSG-encoding genes, and determined on which of the chromosomes 27 of these were located. This advance in our understanding of the trypanosome genome reveals that approximately half of the parasite’s DNA is devoted to VSG-encoding genes.
Their genome assembly allowed the authors to investigate the silencing of VSG-encoding genes. They first confirmed by RNA sequencing that the subtelomeric arrays of VSG-encoding genes are not expressed. Second, using a DNA-crosslinking technique called Hi-C to monitor the physical proximity of DNA sequences to each other in the nucleus, the authors report that there is a greater compaction of subtelomeric arrays than of other regions of the chromosomes. Such compaction is characteristic of silenced chromatin in which genes are not expressed. The telomeres of T. brucei are located near the outermost region of the nucleus16, termed the nuclear periphery, and it is probable that the silent VSG-encoding arrays are located there, too.
How might gene silencing and localization of silent arrays of VSG genes to the nuclear periphery be maintained? The answer is probably complex. Several factors influence the silencing of VSG-encoding genes17,18. Müller and co-workers report that two variant versions of histone proteins, called H3.V and H4.V, also have a role in this silencing process. These histones are a component of chromatin, and mark sites in the genome of T. brucei at which the synthesis of RNA transcripts by the enzyme RNA polymerase II is terminated19.
The authors engineered T. brucei to lack either H3.V or H4.V, or both, and investigated how this affected the structure of the nucleus and chromatin and the silencing of VSG-encoding genes. Experiments using Hi-C and assessing the localization of telomeres in the nucleus indicated that the absence of H3.V, but not of H4.V, altered nuclear organization and resulted in increased clustering of telomeres at the nuclear periphery.
The authors next analysed chromatin structure using a technique called ATAC-seq, which assesses the ability of an enzyme to access specific sequences of DNA. If the enzyme can access a particular sequence, the DNA is probably in an uncompacted chromatin structure that might facilitate access for the components needed to drive gene expression. Müller et al. found that, if both H3.V and H4.V were absent, chromatin accessibility of VSG-encoding sequences in expression sites was increased compared with accessibility in the wild-type situation. To address how these changes affected the expression of VSG-encoding genes, the authors used single-cell RNA sequencing to determine the number of expression sites being expressed in individual cells. They found that, although most cells still expressed just a single VSG in the absence of H3.V and H4.V, some cells in the population expressed up to four different VSGs. It is not known why all the expression sites were not activated when the chromatin accessibility for these genes increased more than usual. But there are probably many levels of control to limit VSG expression to just one at a time, given that this capacity is a key element of pathogen survival.
Many questions remain to be answered. What happens to the expression-site body in the cells that lack H3.V and H4.V? Does its location change, and might the number of these structures increase? Perhaps the cell’s ability to construct multiple expression-site bodies is restricted, which might therefore limit the number of active expression sites in parasites that lack H3.V and H4.V. Another interesting issue is whether changes in chromatin accessibility alter the ease with which expression-site sequences can move into an expression-site body. Answers to these questions might help to illuminate the intimate relationship between genome architecture and the mechanism of antigenic variation in one of the world’s most puzzling, problematic and pugnacious pathogens.
Nature 563, 40-42 (2018)