Ticks were kept frozen at -20°C (in Niger) or at -80°C (in other countries) before being tested. DNA of each tick was extracted as described (7). Rickettsial and ehrlichial DNA was detected by PCR as described, using specific primers (Table).

The sequences of PCR products were obtained and analyzed with the corresponding sequences of rickettsial or ehrlichial species as described (7). Multiple alignment analysis was performed by using the ClustalW program version 1.8 in the DNA Data Bank of Japan (DDBJ; Mishima, Japan [http://www.ddbj.nig.ac.jp/htmls/E-mail/clustalw-e.html]). All sequences used in the study are available in GenBank; the accession numbers of the new genotypes detected in this work are shown in the Table footnotes.

Rickettsial DNA was detected in 24 (7.2%) of the 332 ticks examined (Table). R. africae was detected from A. variegatum from Mali (2/6), Niger (2/6), and Burundi (1/13), and from 1 of 16 A. lepidum from the Sudan. R. aeschlimmanii was detected in Hyalomma marginatum rufipes from Niger and Mali (8/24 and 3/20, respectively) and R. massiliae in 2/37 Rh. muhsamae from Mali. Further, three new ompA sequences (590 bp) were obtained from A. variegatum from Mali and Niger (Table). These were 99.3%-99.5% identical to those of R. africae. In the phylogenetic tree based on these ompA sequences, the three rickettsiae (named RAv1, RAv3, RAv9) were closely related to one another (95.7% bootstrap value) and branched with R. africae (86.1% bootstrap value) (data not shown). Partial sequences (316 bp) of the gltA gene of RAv1, RAv3, and RAv9 were also found to be closely related to those of R. africae (99% of similarity). Two new 16S rRNA ehrlichial genotypes were detected, including ERm58 (1,380 bp) in 7/37 Rh. muhsamae from Mali and EHt224 (1366 bp) in 1/5 H. truncatum from Niger. Both sequences were very similar (99.34% similarity), but different from those described for all the known ehrlichiae (i.e., 98.55% similarity with E. chaffeensis, 98.26% with E. canis and E. ewingii, and 97.75% with E. muris and Cowdria ruminantium). In a phylogenetic tree based on 16S rRNA gene sequences, ERm58 and EHt224 were found to be closely related and to belong to the E. canis group (data not shown). Enlarged gltA sequences of ERm58 (1,140 bp) and EHt224 (1,189 bp) were also obtained from the above ticks. Phylogenetic analyses of these sequences confirmed that ERm58 and EHt224 belonged to the E. canis group (data not shown).

Figure

Figure. Geographic distribution of spotted fever group rickettsiae occurring in Africa. R. = rickettsia.

This study has shown for the first time that R. africae, the agent of African tick bite fever, is present in West Africa (Mali and Niger), the Sudan, and Burundi. It also indicates a potential role for A. variegatum and A. lepidum as vectors of R. africae in these areas. Recently, we also documented cases in tourists returning from numerous countries, including those in West and East Africa (9) (Figure). Our results support the hypothesis that the geographic distribution of African tick bite fever parallels that of the distribution of Amblyomma spp., as ticks are known to be vectors and also reservoirs of tick-borne rickettsiae (2). In Africa, although the principal vector of R. africae appeared to be A. hebraeum, which is prevalent in southern Africa, A. variegatum (which is widely distributed throughout sub-Saharan Africa) appears as a potential vector. Amblyomma are known to readily feed on people in Africa and are commonly infected with rickettsiae (up to 100%). Thus, African tick bite fever may have a high prevalence throughout the continent. Studies have shown seroprevalences of 30%-80% for spotted fever group rickettsiae in sub-Saharan Africa (2), although it is unclear what proportion of those infections might be due to R. africae infection.

In this study, we report for the first time the presence of "R. mongolotimonae" in Africa (Niger). This pathogen was first isolated from an H. asiaticum collected in Inner Mongolia, China. Later, the same agent was isolated from the blood and skin of a febrile woman from Marseille in 1996 (2), which demonstrated its pathogenicity for humans. Subsequently, we have recognized four more cases in southern France (10, and unpub. data). Results of this study suggest that R. mongolotimonae may be associated with Hyalomma sp. ticks throughout the world. In this work, we also detected two rickettsiae of unknown pathogenicity, namely R. aeschlimannii and R. massiliae. Although this is the first recognition of these rickettsiae in Mali and Niger, the epidemiologic importance of this finding has yet to be determined. Finally, we detected three new spotted fever group genotypes closely related to R. africae. Until further studies clarify the position of these organisms, we suggest they may be considered variant strains of R. africae.

Although previous reports, based on the results of serosurveys, have indicated that human ehrlichioses occur in Africa, firm evidence is still absent. Because of the serologic cross-reactivity between ehrlichiae, serosurvey results have to be interpreted carefully. We detected two new ehrlichial genotypes, that is, Erm58 in Rh. muhsamae from Mali and Eht224 in H. truncatum from Niger. Both belong to the E. canis group, which includes E. chaffeensis, C. ruminantium, E. muris, E. ewingii, and a new isolate detected in Japanese ticks (11). In this group, as within each group of ehrlichiae, members share homologous surface antigens and thus cross-react extensively in serologic assays (12). Erm58 and Eht224 may also be organisms responsible for such serologic cross-reactions, including in serosurveys and case reports of human ehrlichioses in Africa. In 1997, new ehrlichial genotypes were also detected in Namibia and Zimbabwe, and a number of ehrlichiae in Africa may be responsible for serologic cross-reactions in serosurveys of humans and animals for currently recognized pathogenic ehrlichiae (13,14). The pathogenicity of the Erm58 and Eht224 recognized in our study has yet to be determined, and further studies to characterize the human ehrlichioses in Africa are indicated. Moreover, it remains to be demonstrated whether H. truncatum and Rh. muhsamae ticks act as vectors or reservoirs of the new ehrlichiae, since ticks also could have been infected while feeding on bacteremic mammals.

Although this study detected for the first time certain richettsiae and ehrlichiae in African countries, systematic sampling was not done, and results cannot address their prevalence and distribution. However, this work provides a starting point for epidemiologic studies there.