A search for the epidemic typhus agent in Ethiopian ticks*

Bull. Org. nwnd.SOar 1973, 48, 563-569 Bull. Wld Hlth Org.J A search for the epidemic typhus agent in Ethiopian ticks* W. BURGDORFER,1 R. A. ORMSBEE,1 M. L. SCHMIDT,2 & H. HOOGSTRAAL 3 The presence of antibodies to Rickettsia prowazeki in domestic animals from several parts of Africa, and the isolation of this rickettsia from the blood ofgoats and sheep and from ticks off cattle or camels in Ethiopia, led to the hypothesis that R. prowazeki in nature may occur in an extrahuman cycle involving ticks and domestic animals. This study attempted to recover R. prowazeki from 2 624 ticks (4 genera, 10 species) collected in central and southern Ethiopia. The ticks were examined by the haemolymph test and by the injection of tissues into guineapigs. No strains of typhus rickettsia were received and there was no serologic evidence suggesting the presence of this agent in any of the ticks examined. One Amblyomma cohaerens contained an organism that reacted specifically with fluorescing antibodies against R. prowazeki; attempts to isolate and identify this agent failed. Fifty-seven (2.2 %) Amblyomma ticks (26 A. gemma, 17 A. variegatum, 14 A. cohaerens) were infected with rickettsiae of the spotted fever group, and probably represented R. conori or closely related rickettsial agents. The hypothesis that domestic animals and their ticks are involved in an extrahuman cycle of R. pro- wazeki is based on reports of antibodies to this agent in livestock from various parts of Africa (Giroud et al., 1952; Reiss-Gutfreund, 1956; Imam & Alfy, 1966), and, what is more important, on isola- tions of R. prowazeki in Ethiopia from the blood of 2 goats and 2 sheep, and from various species of ixodid tick: 4 strains from A. variegatum, 4 from Hyalomma marginatum rufipes, and 1 from H. trun- catum, all taken from cattle or camels (Reiss- Gutfreund, 1956, 1961, 1966). Involvement of domestic animals and ticks in the epidemiology of epidemic typhus-which had been considered as a disease maintained in nature in a man-louse-man cycle only-has led to several inves- * From Work Units MRO41.09.01-0037A6HJ and MFl2.524.009-0029BF6I, Bureau of Medicine and Surgery, Department of the Navy, Washington DC, USA. The opi- nions and assertions contained herein are the private ones of the authors and are not to be construed as official or as reflecting the views of the US Department of the Navy or of the naval service at large. I Rocky Mountain Laboratory, US Department of Health, Education, and Welfare, Public Health Service, Hamilton, Montana 59840, USA. 'US Naval Medical Research Unit No. 3, c/o American Embassy, Addis Ababa, Ethiopia. 'US Naval Medical Research Unit No. 3, USIS, c/o Spanish Embassy, Cairo, Egypt. tigations and publications related to the susceptibility of domestic animals to R. prowazeki and to the role of ticks as vectors of this agent. In one of these studies, Philip et al. (1966) attempted to isolate R. prowazeki from 3 723 ticks (5 genera, 11 species) collected from cattle in central and southern Ethiopia. No evidence of typhus rickettsiae was obtained. Additional efforts to recover R. prowazeki from Ethiopian ticks were made from 1969 to 1971. The results of those studies are the subject of this paper. MATERIALS ANX) METHODS Partially or fully engorged ticks, mostly adults, were removed from cattle grazing at various loca- tions in Shoa, Arusi, Sidamo, Bale, and Harar Provinces. In some instances, ticks were removed from cattle at markets in Addis Ababa and Nazareth. The ticks were placed according to location in plastic tubes with moistened plaster of Paris and shipped to the Rocky Mountain Laboratory for testing. Upon arrival, live and dead ticks were separated and iden- tified. Because their external surfaces were often heavily contaminated with microorganisms, the ticks to be examined were immersed for 15 min in aqueous thiomersal (1 : 1 000), then rinsed twice for 15 mi in sucrose-phosphate-glutamate (SPG) solution, and 3050 -563- W. BURGDORFER ET AL. air-dried before they were subjected to the haemo- lymph test (Burgdorfer, 1970). In this test, haemo- lymph obtained from each tick by amputating the distal portion of one or more legs was collected on a slide, heat-fixed, stained by the method of Gim6nez (1964), and examined microscopically for rickettsial agents. For the identification of rickettsiae in ticks posi- tive to the haemolymph test, duplicate smears of haemolymph were treated by the direct fluorescent antibody (FA) technique with conjugates against typhus and spotted fever group rickettsiae. These conjugates were prepared according to the method of Peacock et al. (1971) and included: (1) antiserum against R. prowazeki from a goat immunized with the Breinl strain; (2) antiserum against R. rickettsi from guineapigs immunized with the " R " strain; and (3) antiserum against R. conori from guineapigs immunized with the Simko strain of this agent. For additional identification and isolation of rick- ettsiae, ticks were dissected individually in cold SPG, and duplicate smears for conventional andFA micro- scopy were prepared from Malpighian tubule and hypodermal tissues. The remaining tissues were then triturated in 4.0 ml of cold SPG, and 0.25 ml of the suspension was injected intraperitoneally into each of 4 male meadow voles (Microtus pennsyl- vanicus). The remaining suspension was stored at - 65'C for future reference. Two voles, 1 on day 5 and the other on day 6, were killed for the prepara- tion of at least 2 sets of tissue smears from the peritoneal lining and tunica vaginalis. One set was stained by the Gimenez method and the other with the above-mentioned conjugates. All ticks negative to the haemolymph test were pooled in lots containing up to 20 specimens depend- ing on the stage of engorgement. Each lot was triturated in 4.0 ml of cold SPG and centrifuged at 1 000 g for 5 min. The supernatant fluid was then injected into 2 male guineapigs, each receiving 0.5 ml intraperitoneally; the remaining suspensions were stored at - 65°C. The guineapigs, whose rectal temperatures were recorded for 14 days, were held until day 28, when they were bled by heart puncture. Their sera were then tested for antibodies to typhus and other rickettsial agents by microagglutination (Fiset et al., 1969) and complement fixation (Fiset, 1964) tests. RESULTS Ticks were collected in 15 lots from May to December, inclusive, in 1969, 1970, and 1971. Unfor- tunately, some shipments contained many dead ticks, predominantly males. Mortality was particularly high among ticks that had been removed from cattle more than 7 days previously. A total of 2 624 living ticks (4 genera, 10 species) was tested (Table 1). This included 4 engorged nymphs whose specific identity could not be established. A. cohaerens and A. varie- gatum were the most abundant species, representing 48% and 22 %, respectively, of all ticks examined. Ticks positive to the haemolymph test Fifty-eight Amblyomma ticks (2.2%.) contained haemocytes with rickettsia-like organisms (Fig. 1) that, according to FA staining reactions, appeared to belong to the following rickettsial groups. Typhus group. One partially engorged A. cohaerens female taken in Nazareth market off an animal from Harar Province was infected with an agent that reacted specifically against anti-R. prowazeki con- jugate. In Gimenez-stained smears this organism resembled a typical rickettsia and was distributed throughout the tick, with hypodermal and ovarian tissues exhibiting a rather intense infection (Fig. 2). In FA-treated smears this organism appeared highly pleomorphic, varying from rod-shaped to coccal- shaped. Typical rickettsia-like forms (Fig. 3) were detected in moderately infected tissues, such as muscles and Malpighian tubules, whereas large masses of morphologically ill-defined antigen were present in hypodermal and ovarian tissues. All attempts to isolate this agent by the injection of tick tissues into voles or guineapigs failed-i.e., none of the test animals developed detectable rickettsial infections or produced antibodies to antigens of either the typhus or the spotted fever group. Subsequent cultivation of the remaining tick tissues on blood agar plates and in brain-heart-infusion broth yielded an organism identified as Pseudo- monas sp. However, smears of these cultures did not react when treated with anti-R. prowazeki or anti-spotted fever group conjugates. Spotted fever group. As demonstrated by specific FA staining of duplicate haemolymph smears, the remaining 57 ticks positive to the haemolymph test (26 A. gemma, 17 A. variegatum, 14 A. cohaerens) were infected with rickettsiae of the spotted fever group (Fig. 4). Study of 29 infected Amblyomma (14 A. varie- gatum, 8 A. gemma, 7 A. cohaerens) under the micro- scope showed rickettsial infections similar to those produced in ticks by R. rickettsi and other members 564 Fig. 1. Haemolymph-test-positive smear from Amblyomma gemma (Gim6nez stain). Fig. 2. Unidentified rickettsia-like microorganism in hypodermal tis- sue smear from Amblyomma co- haerens (Gim6nez stain). Fig. 3. Unidentified rickettsia-like microorganism reacting with anti- Rickettsia prowazeki conjugate in tissue smear from Amblyomma cohaerens (fluorescent antibody staining). 9 P.KV 5 pm 9* w Lr 'l [.. OF... _ . 1 ii-itd ._w A; Fig. 4. Spotted-fever-group rickett- siae in fluorescent-antibody- stained haemolymph smear from Amblyomma cohaerens. Fig. 5. Spotted-fever-group rickett- siae in fluorescent-antibody- stained smear of ovarian tissue of Amblyomma variegatum. Fig. 6. Rickettsial infection in smear of tunica vaginalis of Micro- tus pennsylvanicus 5 days after inoculation of infected Amblyom- ma gemma tissues (Gim6nez stain). EPIDEMIC TYPHUS Table 1. Survey of rickettsial infections in Ethiopian cattle ticks, 1969-71 Geographic distribution according to provinces Tick species -Totals Shoa Arusi Bale Sidamo Harar Amblyomma variegatum 114 (9)a 53 (1) 159 (5) 234 23 (2) 583 (17) coheerens 268 (1) 63 (4) 76 1 850 (10) 1 258 (15) gemma 4 (1) 21 (4) 4 (1) 1(1) 41 (19) 71 (26) spp. 2 2 4 Hyalomma excavatum 14 14 rufipes 3 3 truncatum 2 2 Rhipicephalus evertsi 60 3 2 17 82 simus 1 1 pulchellus 39 18 1 261 55 374 Boophilus decoloratus 19 13 10 34 151 227 Totals 523 (11) 171 (9) 250 (6) 537 (1) 1 143 (31) 2624 (58) a ( ) = Number of ticks with rickettsia-like organisms in haemocytes, as revealed by haemolymph test of the spotted fever group (Fig. 5). All tissues were heavily infected and intranuclear proliferation of rickettsiae was usually evident. Tissue suspensions from each of 27 infected Amblyomma inoculated into meadow voles pro- duced rickettsial infections detectable by conven- tional microscopy of smears from the peritoneum and tunica vaginalis (Fig. 6). In each case, rickettsiae were stained specifically by fluorescent antiserum against R. rickettsi or R. conori. Tissue suspensions prepared from one A. variegatum and from one A. cohaerens killed voles within 48 hours of inocula- tion; the cause of death could not be determined. A more extensive characterization of the rickett- siae found in these ticks will be reported in the future. Suffice it to state here that guineapigs and meadow voles inoculated with tick tissues judged positive by the haemolymph test responded with significant levels (>1: 8) of complement-fixing antibodies, as mea- sured with group antigens prepared from yolk sacs infected with R. rickettsi or R. conori. Serologic cross-reactions with antigens made from R. pro- wazeki or R. typhi were not detected. Ticks negative to the haemolymph test A total of 196 pools containing 2 566 ticks nega- tive to the haemolymph test was injected into guinea- pigs. None of these pools produced typhus infections in these animals. However, suspensions of 8 pools elicited antibodies to spotted fever group antigens. DISCUSSION Ethiopia has long been recognized as an important endemic focus of louse-borne typhus (World Health Organization, 1970). The disease occurs most fre- quently at the peak of the rainy season in June and again during the winter months of December and January. Since 1950, approximately 3 000 cases have been reported annually and the disease continues to occur in all the provinces. The involvement of 567 W. BURGDORFER ET AL. domestic animals and their ticks in the natural cycle of R. prowazeki as stated by Reiss-Gutfreund (1956, 1961, 1966) remains unconfirmed. In the present search for typhus rickettsiae in Ethiopian ticks collected from May to December in 1969, 1970, and 1971, R. prowazeki was not recovered and there was no serologic evidence sug- gesting the presence of this agent in any of the 2 624 ticks examined. It has long been suggested that antibodies to typhus group rickettsiae in domestic animals may be caused by microorganisms that share certain antigenic components with R. prowazeki and are maintained and distributed by haematophagous ar- thropods, possibly ticks. The finding of such an agent in a single A. cohaerens might support this hypothesis. Unfortunately, we were unable to isolate and identify this organism, which produced no detectable infection and elicited no development of antibodies to typhus group antigens when injected into meadow voles and guineapigs. Although the agent was of rickettsial morphology and reacted spe- cifically in the FA test, it is highly doubtful that it was R. prowazeki. Had it been, its occurrence in such large numbers throughout the tick tissues would readily have yielded isolates from meadow voles and guineapigs. The haemolymph test was used on the assumption that R. prowazeki-like tick-borne rickettsiae-may infect all tick tissues, including haemocytes. How- ever, during this study it was learned from separately conducted experiments with ixodid ticks including certain species examined here that, when ingested in rather large concentrations, R. prowazeki survives for only short periods, during which it may produce localized infections in the midgut but not in other tissues (Burgdorfer et al., 1972). If such localized infections occur under natural conditions, the injec- tion of tick suspensions' into susceptible test animals, rather than the haemolymph test, should be used for detecting R. prowazeki in ticks. According to Philip et al. (1966), R. conori, the etiologic agent of boutonneuse fever, is widely dis- tributed among Ethiopian cattle ticks. This conclu- sion was based on 2 isolations (1 from a pool of 30 Rhipicephalus simus and the other from a pool of 42 A. variegatwn) and on the appearance of CF antibodies in guineapigs injected with pools of R. evertsi, R. pulchellus, A. cohaerens, A. variegatum, H. marginatum rufipes, and Boophilus decoloratus. In the present study, only Amblyomma spp. were found infected. Of the 58 ticks positive by haemo- lymph test, 57 (26 A. gemma, 17 A. variegatum, 14 A. cohaerens) contained rickettsiae that were identified by FA staining as members of the spotted fever group. The precise relationship of these rickett- siae to R. conori is unknown. Of 196 pools containing 2 566 ticks negative to the haemolymph test, 8 (4°/) produced guineapig infections characterized by the production of CF antibodies that reacted with spotted fever group antigens. Failure to detect infection by the haemo- lymph test in these instances may have been due to mild infections involving only a few haemocytes, or to infections in an initial phase of development and therefore still limited to the midgut. Such situa- tions certainly may occur if cattle, as postulated by some investigators (Giroud et al., 1963), act as sources for infecting ticks with R. conori. Isolation of this agent (Sampaio & Faia, 1952) from the one- host tick, B. decoloratus, and serologic evidence (Philip et al., 1966) in guineapigs injected with tissue suspensions of this tick taken from Ethiopian cattle strongly support this contention. None of the 227 B. decoloratus tested here revealed rickettsial infec- tions. ACKNOWLEDGEMENTS The authors are indebted to Dr G. M. Kohls, Dr C. M. Clifford, and Dr J. E. Keirans for assistance in identifying ticks; to Dr C. R. Owen for the bacteriological work; and to Mr M. G. Peacock and Mr A. J. Mavros for technical assistance. RItSUMIt RECHERCHE DE L'AGENT DU TYPHUS EPIDEMIQUE CHEZ DES TIQUES D'WTHIOPIE On a tente de mettre en evidence Rickettsia prowazeki, agent responsable du typhus epidemique (transmis par les poux), chez des tiques du betail en Ethiopie. A cet effet, 2624 specimens de tiques, appartenant it 4 genres et a 10 especes, ont ete recoltes; l'hemolymphe a ete examin6e afin de deceler les rickettsies et des suspensions de tissus ont ete injectees a des cobayes. Cette recherche n'a pas permis d'isoler des souches de rickettsies du 568 EPIDEMIC TYPHUS 569 typhus et n'a apporte aucune preuve serologique d'une infection par ces agents chez les tiques examinees. Cependant une Amblyomma cohaerens femelle 6tait infect6e par un micro-organisme de type Rickettsia donnant une reaction positive a I'6gard de Rickettsia prowazeki ai I'6preuve des anticorps fluorescents. Les essais d'isolement et d'identification de cet agent ont echoue. L'injection de tissus de cette tique a des campa- gnols (Microtus pennsylvanicus) et a des cobayes n'a pro- voque aucune infection decelable ni suscite la production d'anticorps actifs contre les rickettsies du typhus. Cinquante-sept (2,2%) tiques du genre Amblyomma (26 A. gemma, 17 A. variegatum, 14 A. cohaerens) etaient porteuses d'une infection a rickettsies decelee par 1'exa- men de l'hemolymphe. L'epreuve des anticorps fluores- cents a montr6 que les micro-organismes en cause appar- tenaient au groupe de la fievre tachet6e. L'inoculation au cobaye et au campagnol de tissus de certaines de ces tiques a provoqu6 la formation d'anticorps actifs contre R. conori et R. rickettsi. REFERENCES Burgdorfer, W. (1970) Amer. J. trop. Med. Hyg., 19, 1010 Burgdorfer, W. et al. (197.2) Amer. J. trop. Med. Hyg., 21, 989 Fiset, P. (1964) Techniques in experimental virology, London, Academic Press, p. 225 Fiset, P. et al. (1969) Acta Virol., 13, 60 Gim6nez, D. F. (1964) Stain Technol., 39, 135 Giroud, P. et al. (1952) Bull. Soc. Path. exot., 45, 313 Giroud, P. et al. (1963) Bull. Soc. Path. exot., 56, 629 Imam, I. Z. E. & Alfy, L. (1966) Bull. Wld Hlth Org., 35, 123 Peacock, M. et al. (1971) Infect. Immun., 3, 355 Philip, C. B. et al. (1966) Bull. Wld Hith Org., 35, 127 Reiss-Gutfreund, R. J. (1956) Bull. Soc. Path. exot., 49, 946 Reiss-Gutfreund, R. J. (1961) Bull. Soc. Path. exot., 54, 284 Reiss-Gutfreund, R. J. (1966) Amer. J. trop. Med. Hyg., 15, 943 Sampaio, A. & Faia, M. de M. (1952) Bol. Inst. Sup. Hig. Dr Ricardo Jorge, 7, 118 World Health Organization (1970) Wkly epidem. Rec., 45, 297