Rickettsia africae: identifying gaps in the current knowledge on vector- pathogen-host interactions

Rickettsia africae is a bacterium of zoonotic importance, which causes African tick bite fever (ATBF) in humans. This pathogen is transmitted by ticks of the genus Amblyomma, with Amblyomma hebraeum and Amblyomma variegatum being the major vectors. Tick species other than the above-mentioned have also been reported to carry R. africae DNA. There is scarcity of information on the epidemiology of this pathogen, yet several cases have been recorded in foreign travellers who visited endemic areas, especially southern Africa. The disease has rarely been described in people from endemic regions. The aim of this study was to discuss the information that is currently available on the epidemiology of R. africae, highlighting the gaps in this field. Furthermore, ATBF cases, clinical signs and the locations where the cases occurred are also listed in this review.


Introduction
Rickettsia africae is a bacterium that was first reported as a species 24 years ago [1]. It is mainly transmitted by African Amblyomma tick species, causing African tick bite fever (ATBF) in humans; mostly in tourists visiting southern Africa [2]. This paper reviews R. africae from the epidemiological perspective, looking at the occurrence of ATBF among individuals from presumed naïve populations never exposed to African Amblyomma bites, and those from endemic regions, where the populations are commonly exposed to R. africae challenge, the potential vectors of this disease in Africa and other continents, and the possible role of mammalian hosts as reservoirs.

Review design
Articles with information related to ATBF, R. africae, tick bite fever, South African tick bite fever, and data on the seroprevalence of ATBF were gathered. Articles written in any language other than English were not included, except for those of historical importance. Searches had no restriction on the research period. Data on ATBF cases reported worldwide were also gathered and ATBF cases included in this study (Table 1) were only those reported from 2004 onwards. This review does not include six ATBF reports which were published from 1996 to 2003, however, two references supporting the cases reported during this time interval were included in this paper. The cases reported in this paper are sufficient in terms of geographic representation and individuals involved and period of cases.

Materials
The databases Medline, Science Direct, PubMed, Google scholar and Google.com were used to perform the searches for publications on the topic. Some articles were retrieved from citations and reference lists in papers on the related topic. The first search date was the 10 th of June 2018 and the last search on google.com was conducted on the 6 th of June 2019; the last search date for other databases was the 14 th of June 2019.

Study selection
Duplicates were removed and articles were selected if their titles and abstracts were related to ATBF, R. africae, tick bite fever, South African tick bite fever or ATBF seroprevalence.

Data collection
Articles reporting ATBF cases were included in this review, irrespective of geographical region. In addition, papers focussing on the diagnosis of ATBF and detection of R. africae were also included. Among the articles reporting ATBF cases, only those published from 2004 onwards were considered. Unusual clinical signs of ATBF in humans, which are symptoms not usually reported to be associated with ATBF, and the detection of R. africae in tick species other than Amblyomma were also noted. Additionally, Amblyomma tick hosts are discussed since they are likely to play a major role in the epidemiology of R. africae.

Results
Data on cases of ATBF in presumed naïve populations and those from endemic regions were collected. For all articles reporting case studies, the following data were tabulated: the year of publication, age of affected persons, gender, nationality, country visited (if any), purpose of the visit to the area where the infection was acquired, the clinical signs presented in each case and the diagnostic method used (Table 1). Unusual clinical signs of ATBF in humans and the detection of R. africae in tick species other than Amblyomma were also included ( Table 2). From 36 ATBF reports, 57 recorded ATBF patients were included in this study. This disease was also found to occur as clustered cases since more than one person would be affected. This review includes seven reports of clustered ATBF cases. Out of the 57 patients included in this study, 40 had information on sex and age. Between the age zero to 29 years there were two females and six males, between 30-60 years, there were 11 females and 13 males, between 61 and 90 years, three were females and five were males. Out of the 57 cases mentioned, only three were reported from the endemic areas, two from South Africa and one from Zimbabwe and the rest were from international travellers who had visited African countries for various activities as indicated on Table 1.
The most common ATBF clinical signs were headache, fever, eschars, rash, lymphadenopathy, myalgia, chills, malaise and arthralgia. Some unusual clinical signs were also reported in some patients and these were; myocarditis, pericarditis, conjunctivitis, decreased vision, floaters, panuveitis, and neurological signs such as feacal incontinence, urinary retention, hyperesthesia, depressed and significant irritability. ATBF was also found to have neurological complications in elderly patients.

Historical background
Fever associated to tick bites was first reported in southern Africa in 1911 by Nuttall, who named it "tick bite fever" (TBF) [3,4]. The disease was also termed Boutonneuse fever by Conor and Bruch after its first discovery in Tunisia in 1910 [5].
In the 1930s, Pijper noted differences between the clinical signs of the disease that was termed 'TBF', identified in southern Africa in 1911, and those of Boutonneuse fever, discovered in North Africa in 1910 [6][7][8]. In addition, he noted differences both in the epidemiology and clinical severity of the two diseases. Boutonneuse fever had more severe clinical signs as compared to TBF. Pjiper's peers refuted his findings and attributed the differences in the severity of the two diseases to the age differences between the people who were affected [1,9]. However, Pjiper's observations are currently accepted [10]. In fact Pjiper possibly isolated the causative agent of ATBF in the 1930's and demonstrated that it was different from R. conorii by cross-protection assays (Pjiper 1936, Arch.Inst. Pasteur Tunis 25, 388-401 as cited by Fournier et al. (1998) [8].
The discovery of R. africae Amblyomma hebraeum tick bites were found to be associated with TBF in Africa in the 1990s, and many cases were reported in southern Zimbabwe [1]. This was consistent with Pijper's report, in which he indicated that A. hebraeum ticks were predominant in southern Rhodesia (now Zimbabwe) [7].
In 1992, a 36-year-old Zimbabwean woman presented at a hospital in Chiredzi, a small town in south-east Zimbabwe, with fever, headache, regional lymphadenopathy and inoculation eschar, but no cutaneous rash. A blood sample was collected from the patient on the fifth day after presentation. DNA was extracted and restriction fragment length polymorphism (RFLP) were performed [11]. The isolate was found to be the same as those collected from A. hebraeum from several regions in Zimbabwe [1], as well as a spotted fever group (SFG) rickettsiae isolate from Ethiopia [12]. This new isolate was named Rickettsia africae in  [80] 1996, after it was proved that two different Rickettsiae species were causing two different rickettsial diseases in southern Africa [1]. Tick bite fever, caused by R. africae, was found to be associated with a history of travel to grasslands and game parks, whereas, Boutonneuse fever, caused by R. conorii, was associated with a history of contact with the dog ticks, Rhipicephalus sanguineus, Rhipicephalus simus, and Haemaphysalis leachi [11] in peri-urban or peridomestic settings. Humans infected with this pathogen showed severe clinical signs, which were associated with high mortality rates [1].

Biology and characteristics of R. africae
Rickettsia africae is an obligate intracellular, Gramnegative coccobacillus [13]. It was proved by electron microscopy that this bacterium can be found within the cytoplasm of host cells and has an outer slime layer as well as a tri-laminar cell wall. The cell wall of R. africae contains lipopolysaccharide antigens. These are highly immunogenic and responsible for extensive crossreactivity with other species of SFG rickettsiae [13]. Species-specific protein antigens are found in the highmolecular-weight rickettsial outer membrane protein A (rOmpA) and B (rOmpB). The rOmpA protein seems to be specific to SFG rickettsiae [14]. Most species of the SFG rickettsiae have been characterized by SDSpolyacrylamide gel electrophoresis (PAGE), Western blot and PCR-RFLP analysis [14]. The bacterium cannot be cultured in cell-free media. However, it can grow in the yolk sacs of developing chicken embryos, and in cell cultures [1].

Detection of R. africae
The methods commonly used to detect and confirm the presence of R. africae in tissue are PCR and sequencing respectively. A quantitative PCR (qPCR) targeting the citrate synthase gene (gltA) is the most frequently used Rickettsiae genus-screening assay [15]. After screening, gltA positive samples are usually tested in a conventional PCR (cPCR) targeting the ompA gene of SFG rickettsiae [15] and more recently, a qPCR targeting the ITS gene was developed for the same purpose [16]. Sequencing is the only method currently available to identify SFG rickettsiae at species level.

Amblyomma vectors of R. africae
It is generally accepted that the ticks that transmit R. africae in Africa belong to the genus Amblyomma of the family Ixodidae (hard ticks), with A. variegatum and A. hebraeum being the main vectors [13]. A. hebraeum is mainly distributed in southern Africa and A. variegatum in West, Central and eastern Africa, as well as in the eastern Caribbean [17]. A. variegatum is also present in some parts of southern Africa where it extends into Zambia, north eastern Botswana, the Caprivi Strip of Namibia, Angola, north western Zimbabwe and central and northern Mozambique. It also occurs in Madagascar and several Indian Ocean islands [17].
Amblyomma hebraeum tick species are considered to be the main vector of R. africae in South Africa [13]. The infestation of humans by this species and infection with R. africae is relatively common due to the wide distribution of the tick vector in rural areas of the country. Rickettsia africae infection rates of up to 100% have been detected by PCR and sequencing in Amblyomma ticks from endemic areas [18,19]. The A hebraeum and A. variegatum tick species are three-host ticks and all active developmental stages (larvae, nymphs and adults) have been proved to be potential vectors of Rickettsiae [20,21]. Studies performed by  and Mason and , reported that these tick vectors can maintain R. africae through transovarial and trans-stadial transmission through two generations [22,23].
Rickettsia africae DNA was also detected in Amblyomma lepidum, Amblyomma gemma, Amblyomma cohaerens, and Amblyomma compressum in Sudan, Djibouti [23], and in the Somali region of Ethiopia [16,19]. Furthermore, R. africae was also detected in Amblyomma loculosum, a tick species that is usually known for infesting marine birds in tropical islands [24,25]. R. africae DNA was found in Amblyomma ovale ticks collected from dogs in Nicaragua, Central America, in 2013. This was the first report on R. africae in the American continent [26]. Given the documented distribution of R. africae among African Amblyomma species, it is reasonable to infer that all African Amblyomma species could be competent vectors of this pathogen. However, this assumption should be confirmed. In this context, it is worth mentioning the recent finding of the intergration of R. africae chromosome in the nuclear genome of A. variegatum, which can have major implications on detection specificity of R. africae in Amblyomma species [27].
Different studies conducted between 2003 and 2016 in the African continent (Table 2) report other tick species in which R. africae DNA was found such tick species belong to Haemaphysalis, Hyalomma and Rhiphicephalus genera. However, the available data does not provide direct evidence of vector competence for any of these vectors. All the studies that had such reports indicated that the ticks were collected from vertebrate hosts hence the R. africae DNA detected could have been from the blood meals they had on the hosts and not due to infection of the tick tissues with the Rickettsia. Therefore, there is a need for further investigation on R. africae vector competence. In addition to the above mentioned ticks, R. africae was also detected in fleas collected from migratory birds [28].

Mammalian hosts as reservoirs of R. africae
On the African continent, A. hebraeum and A. variegatum, the main vectors for R. africae, have a wide host range that includes domestic and wild species [29]. These vectors show a marked preference for large animal species and thus prefer cattle to other domestic species such as goats, sheep and donkeys. Among wild species, buffalo, eland, giraffe and kudus are preferred [13]. These wild ungulates are of major importance to the ecology of the Rickettsia species in areas where domestic animals are dipped intensively or where these animals are absent [13]. The adult stages of these ticks feed on wild ungulates. The hosts for larvae and nymphs are the same as those for adult ticks, however, they can also feed on lizards, small mammals and ground-feeding birds [30]. Humans are accidental hosts for these ticks and legs are the usual attachment sites for them. The ticks can also crawl on the skin and may be found attaching to the groin or axilla, where there is moisture [13]. These ticks respond to stimuli like carbon dioxide, ammonia, humidity, aromatic chemicals, airborne vibration and body temperature, all of which are strongly associated with their predilection sites on their hosts [31].
Ticks require blood meals for their continued development, reproduction and survival. Cattle play an important role in the ecology of R. africae by maintaining tick populations [18]. Serological surveys in cattle, conducted in Zimbabwe, using the immunofluorescence antibody (IFA) assay, showed that 80-100% of animals have antibodies against SFG rickettsiae [32,33]. In spite of the serological evidence indicating exposure to R. africae, no clinical signs associated to infection with R. africae have been reported in animals.
Experimental studies on the pathogenesis of SFG rickettsiae in Zimbabwe suggest the maintenance of the pathogen in cattle [32]. All sero-negative cattle (n = 8), experimentally infected with rickettsia organisms isolated from A. hebraeum ticks and cultured in Vero cells, were found to be positive on IFAT after three days post-infection. To determine ricketsiemia in these cattle, sero-negative guinea pigs were inoculated with blood from the experimentally infected cattle. All guinea pigs sero-converted, indicating that these cattle were rickettsemic for at least 32 days post-infection [32]. This constituted the first experimental evidence of the possible role of cattle as reservoirs for R. africae. However, it is worth pointing out that this suggestion is solely based on the sero-conversion in cattle and guinea pigs, which should be regarded with caution considering the low specificity of Rickettsiae serological assays. To confirm bovine hosts as reservoirs of R. africae, experiments using DNA-based methods should be performed. The uncertainty of the role of cattle as R. africae reservoirs is further corroborated by a study conducted in Kenya, where no rickettsemia was detected in cattle, sheep and goats, while 92.6% of A. variegatum recovered from the same animals tested positive R. africae DNA [18]. The scarcity of studies on R. africae diversity and the role of cattle or any other mammalian hosts as R. africae reservoirs has been recognised as a major gap in the understanding of R. africae epidemiology in the African continent.
Transovarial transmission of R. africae is well documented [22]. However, there are no studies on the efficiency of transovarial transmission for several generations in Amblyomma ticks. This is of great importance since it can provide conclusive evidence on whether R. africae can be maintained in its vector without the need for mammalian hosts as reservoirs.

ATBF presentation in different populations
Cases of ATBF in humans usually occur in clusters. This is because of the feeding habits of Amblyomma ticks; they hide in their microhabitats and attack hosts as they appear. This is especially noticed in tourists visiting endemic areas [34]. ATBF presents with flulike symptoms with fever, nausea, fatigue, headache and myalgia. Most of the cases in Table 1 were associated with these common ATBF clinical signs. The disease is usually self-limiting. However, in the elderly and immunocompromised individuals, it can be more severe. Some ATBF cases in Table 1 were associated with complications such as chronic fatigue, reactive arthritis, encephalitis, myocarditis and cellulitis have also been reported, but mostly in the elderly [35]. Complicated ATBF in a 40-year-old Italian traveller returning from Zimbabwe was reported to have painful sacral syndrome characterised by severe pain on the leg, urinary retention and faecal incontinence and rectal tenesmus and these were attributed to be due to immune mediated mechanisms [36]. Duval and Merrill (2016), also reported a complicated case of ATBF where retinitis was the main symptom in a 67 year old lady from Canada [37].
Tick bite sites appear as either single or multiple inoculation eschars [35]. The typical inoculation eschar associated with ATBF consists of a central black crust surrounded by a red halo, occurring as a result of inflammation [9]. Acute cases have been reported in travellers from Europe and America after they visited southern African countries [38]. A report by Frean et al. (1998), indicated an estimated infection rate of 4-5% in foreign travellers visiting South Africa [9]. Clinical signs usually appear after they return to their country of origin since the incubation period of ATBF is five to ten days [24]. ATBF is a self-limiting disease hence many people may be affected and they do not visit hospitals for treatment. The disease can also be misdiagnosed for other diseases which present with fever like malaria and typhoid. This could be the case in many African countries where proper diagnostic laboratories and facilities are lacking. Difficulty in diagnosing the disease in the indigenous population could be also attributed to pigmented skin since it could be very difficult to notice the inoculation eschars hence such pathognomonic features of the disease are easily missed [21].
Although underreporting and misdiagnosis of ATBF can contribute to the underestimation of the disease in populations from endemic regions, the epidemiology of this infection in African rural areas strongly suggests early exposure leading to the establishment of endemic stability in these populations [22]. Furthermore, there are limited publications on the seroprevalence of R. africae in the rural population in areas where the tick vector exists [13], which makes it difficult to determine the status of immunity to this pathogen at population level.

Conclusions
Rickettsiae africae, transmitted by A. variegatum and A. hebreaum, was definitively associated with ATBF in 1996. Since then, the organism's DNA has been detected in other Amblyomma species, both in the African and American continents. Furthermore, R. africae DNA has also been detected in tick species other than Amblyomma. Further studies on the vector competence of other tick genera for this pathogen should be performed in order to fully clarify the dynamics of R. africae infection in different ecological niches. Literature on the role of mammalian hosts is scarce and contradictory. Moreover, in spite of the confirmation of transovarial transmission, the capacity for vertical transmission for several generations has yet to be fully elucidated.
A striking feature of the clinical presentation of ATBF is the marked difference between humans from presumed naïve populations and those from endemic regions since almost all of the reported ATBF cases reported worldwide were from international travellers after trips to ATBF endemic areas. In order to confirm whether endemicity is the cause of the sporadic occurrence of clinical signs from humans in rural areas of Africa, structured serological surveys including different age cohorts should be conducted.
This review highlights significant gaps in R. africae research, which, if addressed, will result in the better comprehension of ATBF epidemiology.