Antigenic variation in the African trypanosomes involves the sequential expression of genes coding for different variant surface glycoproteins (VSGs) (reviewed in refs 1-3). When expression of some VSG genes is switched on, a newly duplicated copy of the expressed gene has been observed within the trypanosome genome, which is not found after the gene's expression is switched off again. The duplicated copy has therefore been called an expression-linked copy (ELC). The expression of the gene appears to be strictly coupled to the presence of the ELC. This has led to the hypothesis that the duplicative transposition generating the ELC may itself be responsible for the control of VSG expression. With other VSG genes, expression-linked duplication has not been observed, and expression is clearly not controlled in this way. Data are presented here which demonstrate that either of these observations may be obtained with a single VSG gene, depending on the chance selection of particular clones from antigenically switched populations. Thus, the different observations do not imply the existence of two distinct classes of VSG gene controlled by different mechanisms, but different aspects of processes common to all VSG genes.

Restriction endonuclease maps of some trypanosome variant surface glycoprotein (VSG) genes have a site 3′ to the gene where many enzymes appear to cut. This site is sensitive to Bal31 exonuclease, indicating a natural double-strand break in the DNA. One VSG gene exists in several cell clones as a non-integrated minichromosome.

The cDNA sequence for the variable surface glycoprotein (VSG) expressed in Trypanosoma brucei clone ILtat 1.4 (called clone D for brevity) hybridizes strongly to three regions in trypanosome genomic DNA. These three regions were extensively characterized by Southern hybridization analyses, genomic DNA cloning and DNA sequence determinations. All three regions occur in the genomes of all trypanosome clones of the ILTAR 1 repertoire regardless of whether or not VSG D was being expressed. Extensive (clone dependent) DNA rearrangements and a (clone independent) double strand DNA break were found distal to the 3'-end of the VSG D coding sequence of one of the regions. VSG D mRNA is most likely synthesized from this region, but a recombinant DNA clone of the VSG coding sequence could not be obtained for confirmation. Recombinant clones of the other two regions were obtained. DNA sequence analyses revealed that their coding sequences differ from each other by 17%. They differ from the ILtat 1.4 cDNA sequence by 4% in one case, and 13% in the other. By analogy with another VSG gene system, one of these two regions may have originally given rise to the third region from which the mRNA is probably transcribed.

Some variable surface glycoprotein (VSG) genes of Trypanosoma brucei undergo duplication and transposition when they are expressed. We report here the cloning of cDNAs coding for two VSGs from the ILtar 1 repertoire. Analysis of the genomes of trypanosomes expressing these and other antigens shows that there is no additional copy of the sequences coding for eight VSG in expressing clones of trypanosomes, and reveals rearrangements analogous to those previously described for the gene for another VSG from this antigen repertoire. The data indicate that duplication does not accompany the expression of these VSG genes. Transposition to a specific expression site cannot be excluded, but would have to involve either a much larger segment of DNA, or movement to a region of much greater homology with the previous flanking sequences, than is observed for VSG genes that are duplicated when expressed. It is reasoned that the control of expression by coupled duplication and transposition is not sufficient to account for the selection of a single VSG gene for expression.

African trypanosomes express different surface antigens sequentially in their mammalian hosts1. Expression of different variable surface glycoprotein (VSG) antigens is associated with rearrangements of genomic DNA2–5. Several laboratories have reported that expression of some VSG genes is accompanied by the appearance of a duplicated copy of the genes at a new location in the genome3,4. However, in a series of sequentially related Trypanosoma brucei clones from a different stock we were unable to detect duplication of one VSG gene (ILTAT 1.2, B) in expressing clones. Two copies of this gene were found in expressing and non-expressing clones. Apparent insertions and deletions occurred near the 3′ ends of both gene copies, but could not be correlated with expression2,5,6. We report here the observation of both types of rearrangement for different VSG genes in the same series of trypanosome clones. Thus the different types of rearrangement are not a function of different trypanosome stocks, but of different VSG genes within a stock.

A simplified procedure for isolation of specific mRNA using Staphylococcus aureus protein A immunoadsorbent chromatography has been developed. This procedure has been applied to the mRNA encoding Trypanosoma brucei variable surface antigen. Trypanosome polyribosomes were reacted with antibodies isolated from an anti-variable antigen-specific serum by protein A-Sepharose column chromatography. Antibody-bound variable antigen-synthesizing polyribosomes were then separated from unbound polyribosomes also by protein A-Sepharose column chromatography. This simple protocol gave a greater than 50% yield of variable antigen-specific mRNA which appeared to be very highly purified as determined by translation in an mRNA-dependent reticulocyte lysate assay. The mRNAs encoding three different T. brucei surface antigens have been purified. The procedure described here should be useful in purification of other mRNAs.

For the purpose of investigating the genetic basis of antigenic variation in Trypanosoma brucei, we have analyzed the structure of the genome surrounding the gene coding for one T. brucei variable antigen (ILTat 1.2) in several T. brucei clones expressing this and other variable antigens. In each case there are two copies of the gene. We found no evidence of an extra copy associated with the expression of this gene. Differences were found between the two copies in a single clone, and between the copies in different clones. The differences could be explained by insertion and deletion of various lengths of DNA in a region beyond the C-terminal end of the gene. Differences in genomic structure were found even between clones expressing the same antigen, whether ILTat 1.2 or another. Thus, no feature of the rearrangements observed can be correlated with the expression of a particular antigen.

We have described some rearrangements of a variable-antigen gene in T. brucei. We suggest that there are two copies of the ILTat 1.2 variable-antigen gene in each of a number of trypanosome clones closely related by sequential relapses. Both copies of the gene undergo rearrangements, apparently the result of insertion and deletion of lengths of DNA in a region at or beyond the 3' end of the coding sequence, giving rise to different-size restriction enzyme fragments in different clones of trypanosomes. No feature of these rearrangements can be correlated with expression of the gene. Our data differ from those of Hoeijmakers et al. (1980) in two important respects: (1) Neither copy of the gene remains in a constant genomic context in all trypanosome clones. (2) We do not see an extra copy associated with the expression of the gene. These observations do not suggest any obvious mechanism for the phenomenon of antigenic variation.

The capacity of African trypanosomes to express sequentially a large repertoire of different surface antigens during an infection enables the parasite to evade the immune response of its host, and makes attempts to produce a vaccine against the disease difficult. It is evident that point mutations cannot account for antigen diversity. Variable antigens like immunoglobulins are derived from an extensive family of genes of which only one is expressed in a given cell. As somatic tic recombination is involved in the immunoglobulin gene system, this similarity prompted us to search for somatic rearrangements in trypanosome variable antigen genes. We have constructed a recombinant plasmid containing approximately half the DNA sequence coding for a Trypanosoma brucei variable antigen and hybridised the inserted sequences to various restriction enzyme digests of nuclear DNA from different trypanosome clones. Differences in the sizes of restriction tion fragments hybridising to the inserted variable antigen coding sequence show altered positions of enzyme sites relative to this sequence, indicating different arrangements of DNA sequences around this gene in different trypanosome clones.