Diversity of Ehrlichia ruminantium major antigenic protein 1-2 in field isolates and infected sheep

Abstract

Proteins expressed from the map1 multigene family of Ehrlichia ruminantium are strongly recognized by immune T and B cells from infected animals or from animals that were infected and have recovered from heartwater disease (although still remaining infected carriers). Analogous multigene clusters also encode the immunodominant outer membrane proteins (OMPs) in other ehrlichial species. Recombinant protein analogs of the expressed genes and DNA vaccines based on the multigene clusters have been shown to induce protective immunity, although this was less effective in heterologous challenge situations, where the challenge strain major antigenic protein 1 (MAP1) sequence differed from the vaccine strain MAP1. Recent data for several ehrlichial species show differential expression of the OMPs in mammalian versus tick cell cultures and dominant expression of individual family members in each type of culture system. However, many genes in the clusters appear to be complete and functional and to generate mRNA transcripts. Recent data also suggest that there may be a low level of protein expression from many members of the multigene family, despite primary high-level expression from an individual member. A continuing puzzle, therefore, is the biological roles of the different members of these OMP multigene families. Complete genome sequences are now available for two geographically divergent strains of E. ruminantium (Caribbean and South Africa strains). Comparison of these sequences revealed amino acid sequence diversity in MAP1 (89% identity), which is known to confer protection in a mouse model and to be the multigene family member primarily expressed in mammalian cells. Surprisingly, however, the greatest sequence diversity (79% identity) was in the less-studied map1-2 gene. We investigated here whether this map1-2 diversity was a general feature of E. ruminantium in different cultured African strains and in organisms from infected sheep. Comparison of MAP1-2s revealed amino acid identities of 75 to 100% (mean of 86%), compared to 84 to 100% (mean of 89%) for MAP1s. Interestingly, MAP1-2s varied independently of MAP1s such that E. ruminantium strains with similar MAP1s had diverse MAP1-2s and vice versa. Different MAP1-2s were found in individual infected sheep. Different regions of a protein may be subjected to different evolutionary forces because of recombination and/or selection, which results in those regions not agreeing with a phylogeny deduced from the whole molecule. This appears to be true for both MAP1 and MAP1-2, where statistical likelihood methods detect heterogeneous evolutionary rates for segments of both molecules. Sera from infected cattle recognized a MAP1-2 variable-region peptide in enzyme-linked immunosorbent assay, but less strongly and consistently than a MAP1 peptide (MAP1B). Heterologous protective immunity may depend on recognition of a complex set of varying OMP epitopes.

FIG. 1.

Alignment of MAP1 and MAP1-2 sequences of E. ruminantium strains from Africa and the Caribbean. (A) Alignment of MAP1 sequences. Three major strain-variable regions have been identified previously (37). For purposes of comparison with the MAP1-2 alignment in panel B, variable regions are identified using letters above sequence elements conserved between MAP1 and MAP1-2. Hence, variable region I extends from a to b (residues YIS . . . IGY), variable region II from c to d (FDVK..SPY), and variable region III from e to f (FK . . . FG). The thick line above the sequence shows the location of the MAP1B peptide identified previously (46) as containing a dominant B-cell epitope recognized by infected ruminants and used the ELISAs for Table 2 and Fig. 2. (B) Alignment of MAP1-2 sequences. As in panel A, variable regions are identified with letters above sequence elements conserved between MAP1 and MAP1-2. The thick line above the sequence shows the location of variable region synthetic peptides used for Table 2 and Fig. 2. Sequences were either derived from DNA templates prepared from E. ruminantium strains cultivated in vitro or derived from acutely infected sheep. In the latter case, the animal number is given before the strain designation.

FIG. 2.

ELISA testing reactions of infected cattle with MAP1B or MAP1-2 peptides. Sera were tested at a 1:50 dilution on plates coated with the peptides indicated in Fig. 1, and reactions developed with horseradish peroxidase-labeled protein G and substrate. Responses to variable region peptides of MAP1-2 were detected, but less consistently than those to MAP1B. Gardel and Crystal Springs peptides were recognized most strongly by different animals and at different times postinfection. Animals 130, 131, 132, 136, 137, 138, and 139 were infected by field exposure, and animals 109, 110, and 114 were infected by controlled, repeated exposure to ticks infected with the Plumtree strain.