Molecular detection of Rickettsia africae in Amblyomma ticks collected in cattle from Southern and Central Mozambique

Introduction : Rickettsia are Gram-negative and obligate intracellular bacteria, which cause typhus and spotted fever-like diseases in humans. In Africa, Rickettsia africae of the Spotted Fever Group Rickettsia (SFGR) is the etiologic agent of the African Tick-Bite Fever. The disease is transmitted by ticks of the genus Amblyomma , which serve as vectors and reservoirs of Rickettsia . In this study, we aimed to detect Rickettsia species in ticks collected from cattle in south and central Mozambique. Methodology: DNA from 412 adult ticks and 22 pools of larvae were extracted and tested for the presence of Rickettsia genes gltA, ompA and ompB by PCR, followed by sequencing and phylogenetic analysis. Results: Our results showed that in adult ticks, 79.5% (n = 330), 66% (n = 274) and 67% (n = 275) samples were positive for gltA , ompA and ompB genes, respectively. Among the 22 pools of larvae analysed, 77.2% (n = 17) were positive for the three genes tested. The infection rates ranged from 43% to 100% for Rickettsia by gltA in all locations studied, with maximum values of 100% observed in the districts of Maputo province namely Changalane, Boane and Matutuine district. The phylogenetic analysis of amplified sequences revealed that samples under study grouped with R. africae for the 3 genes. Conclusion: The study showed that Spotted Fever Group Rickettsia represented by R. africae widely circulate in Amblyomma ticks collected in south and central regions of Mozambique.


Introduction
The genus Rickettsia comprises obligate intracellular bacteria [1], among which 22 species have been proven to be pathogenic to humans [2,3].This genus is divided into three taxonomic groups defined according to the distinctive diseases they cause, namely, the typhus group (TG), spotted fever group Rickettsia (SFGR) and scrub-typhus groups (STG), each comprising several Rickettsia species [3].The three groups differ in terms of the clinical symptoms they cause, on the basis of host specificity, [4,5], in vitro growth conditions, antigenic characteristics, the molecular sequences of conserved genes [3,6,7], clinical features, guanine and cytosine (G + C) content of the genome and intracellular localization [5,8].
Within the SFGR, the species R. africae is the causative agent of the African Tick-Bite Fever (ATBF), a tick-borne disease which was long mistaken for the Mediterranean spotted fever (MSF) caused by R. conorii [14] .Over 100 years ago, MC Naught and Sant'Anna reported what is likely to have been the first description of human cases of ATBF in Mozambique and South Africa [14].This initially led to a debate as to whether these cases were indeed the result of a new type of infection, and not MSF, a disease that had been described in Tunis in 1910 [15].ATBF causes moderately severe clinical features, including fever, headache, myalgia and often an eschar at the site of the tick bite, but can become more serious for elderly and immunocompromised people [16,17].
Recently, the implementation of molecular methods for diagnostics has aided the discovery of several new species of Rickettsiae worldwide [18].However, despite the increasing awareness of the importance of emerging rickettsial diseases, there is still a lack of information on the prevalence of SFGR Rickettsiae, especially in countries such as Mozambique, where diagnostic capacity is limited.
Tick-borne spotted fever is the most frequently reported type of travel-associated rickettsial infection, and the African tick-bite fever is the most commonly reported in Africa [14,19,20].However, even though R. africae is widely distributed in sub-Saharan countries [21], and sero-surveys have shown that infections are extremely common in humans [22,23], reports on ATBF in indigenous people are unexpectedly rare [24].This could be due to the difficulty of proper diagnosis for infections caused by Rickettsia, especially when people are infected at a young age and medical attention is less likely to be sought [24].Furthermore, the lack of sensitive and specific tests contribute to hinder the proper diagnostic capacity in rural areas where the infections are more likely to occur [25].
In Southern Africa, ticks of the species Amblyomma hebraeum are recognized as vectors and reservoirs of R. africae, which has been demonstrated to be transovarially and transtadially transmitted in Amblyomma ticks [24] .ATBF caused by R. africae and transmitted by Amblyomma ticks has been reported in neighboring countries such as Zimbabwe and South Africa [14,19,26,27,28].However, its presence in ticks from Mozambique has not yet been confirmed.Additionally, the extent of zoonotic transmission and the contribution of this agent in cases of non-malarial acute febrile illness in humans in Mozambique is currently unknown, emphasizing the importance of conducting epidemiological studies involving hosts and vectors.
Therefore, this study was designed with the aim of assessing the infection rate of Rickettsia spp, and especially R. africae, in ticks collected from autochthonous cattle from southern and central regions of Mozambique, using a molecular approach.

Tick collection
From March to August 2013, adult ticks and larvae of Amblyomma spp.were collected from cattle.Adult ticks were collected in 12 districts of the southern and central regions of Mozambique, in Maputo, Inhambane, Sofala and Manica Province.
For convenience, the cattle from which the adult ticks were collected (n = 30) were selected based on the criterion of good tick visibility on the animals.Ticks were removed manually and stored into tubes containing 70% ethanol.Each tube was labeled according to the place of collection (district), species, sex and collection date.
Larvae were collected from pastures in the South of the country by dragging, pooled into tubes with 10-20 larvae and kept in 70% ethanol.Subsequently, adult ticks and larvae were submitted to taxonomic identification at the Biotechnology Center of Eduardo Mondlane University (CB-UEM) laboratory, based on the entomological keys of Walker et al. [29].

DNA extraction
Adult ticks were weighed, and those within the 25-40 mg range were included in the study.Ticks were subsequently washed five times using phosphate buffer saline (PBS) and cut in small pieces to be processed for extraction.Tick and larvae DNA extraction was performed using the QIAamp DNA Blood Mini kit 250 Cat.No 51106 (Qiagen, Hilden, Germany), following the manufacturer's instructions.The DNA was stored at -20° C for further molecular analysis.Samples were identified according to each locality with a three-digit code referring to the three initial letters of the collection location as follows: Cha for Changalane; Manh for Manhiça; Mat for Matutuíne; Boa for Boane; Mab for Mabote; Mbn for Mambone; Mass for Massinga; Msr for Mossurize; Mgd for Magude; Mch for Machaze; Ins for Inhassoro and Chi for Chibabava.

PCR detection
Screening of Rickettsial DNA was performed using Rickettsia specific PCR assay for gltA, ompA and ompB genes, chosen according to previously published studies on Rickettsia detection [2].Two primer sets, CS2-F/R and CS239/CS1069 for the citrate synthase gene (gltA) present in all Rickettsia spp (SFGR and TG), were used to amplify fragments of 401 bp and 830 bp, respectively [2].
The reaction mixture consisted of 2 µL of extracted gDNA, 5 µL of PCR master mix (5x Green Go Taq Buffer, Promega, Madison, USA), 1.25 µL each of forward and reverse primers, 14.75 µL of sterile water and 0.25 µL of Go Taq.Negative and positive controls for the PCR were sterile water and R. conorii DNA from cell culture.The reaction was carried out in a PCR Applied BioSystem machine (Gene Amp® PCR System 9700, Foster City, California, USA) with the following conditions: initial denaturation at 95°C for three minutes, followed by 40 cycles of 95°C for 15 seconds, 48°C for 30 seconds, 72°C for 30 seconds, with an additional extension period of 72°C for seven minutes on the final cycle.

Limit of detection
To analyze the limit of detection and robustness of the primers, primers CS2-F/CS2-R Rr190k71p/Rr190k70n and RICI/RICII were tested in triplicate with a range of 10 ng to 1 fg (1 fg = 1×10 -6 ng) of purified genomic DNA, (1 genomic copy assuming the size of Rickettsia genome as of ~1Mbp = 1fg).Rickettsia DNA with known concentration (18.6 ng/μL) was serially diluted to obtain quantities ranging from ~10 ng to ~1 fg.Dilutions were visualized on agarose gel for quality and integrity control.
Furthermore, a sample containing only tick DNA was included in the reaction to test the potential inhibitory effect of tick DNA on Rickettsia detection.After mixing, total DNA (tick DNA and Rickettsia DNA) was diluted 10X and followed the same procedure as the positive control in the previous step.Finally, each pair of primers was tested using 3 positive samples for Rickettsia, namely Cha2, Cha30 and Cha29 with known concentrations of 24.9 ng/μL, 131.1 ng/μL and 162.1 ng/μL, respectively.The PCR was conducted in 3 replicates following the procedure mentioned above in a 1:10 serial dilution.The PCR product was visualized on 1.5% agarose gel for DNA integrity control and to assess the limit of detection.

Sequencing and phylogenetic analysis
Selected PCR products of the genes gltA, ompA and ompB were purified using the NEB cleanup kit (New England Biolabs, Ipswich, Massachusetts, USA).Five microliter of purified PCR product were mixed to 10 pmol of the appropriate primer and sent to Macrogen Europe for sequencing (https://dna.macrogeneurope.com/eng/).The sequences obtained were used to determine the phylogeny through a comparative analysis with sequences of Rickettsia sp available in Genbank retrieved by BLASTn [31].Sequences were aligned using Clustal X version 2.0 (Conway Institute UCD Dublin, Ireland) [32].Phylogenetic trees were produced using Bayesian Inference (BY).BY phylogenetic reconstructions were performed using the HKY85 substitution model by running 1×10 6  generations, with Markov chains sampled every 1000 generations.A burn-in of 10% was applied and the remaining trees were used to compute a 50% majority rule consensus tree and posterior probabilities.
Larvae were only collected in Matutuíne and Changalane districts, and in the Veterinary Faculty campus, Eduardo Mondlane University.Specifically, 22 pools of larvae were collected in the study: 12 from the Veterinary Faculty campus, 7 from Changalane and 3 from Matutuíne districts in Maputo province (Figure 1).
When looking at the Rickettsia infection rates in adult tick per district, A. hebraeum ticks ranged from 43% to 100%, with the highest values (100%) detected in ticks collected in districts of the Maputo province, namely Changalane, Boane and Matutuine.The lower infection rate (43%) was observed in ticks collected in Inhambane Province.Adult A. variegatum ticks collected in Inhambane showed an infection rate of 67%.
Larvae collected in the districts of Maputo Province showed Rickettsia infection rates ranging between 71 % and 75% in Changalane district and the Veterinary Faculty campus, while a 100% of infection rate was observed in Matutuíne district (Figure 1).
The limit of detection of Rickettsia DNA for the three primers, remained the same in the presence of tick DNA (10 -2 equivalent to 10 5 copies of genomic DNA) (Table 1).

Confirmation of Rickettsiae species by sequencing
A total of 32 samples were sequenced for the gltA, ompA and ompB genes.After trimming and BLASTn comparison with publicly available Rickettsia sequences, low quality sequences (short, weak signal) were eliminated, resulting in 10, 10 and 7 good quality sequences for gltA, ompB and ompA, respectively.
Unexpectedly, one sample isolated from A. variegatum matched with Ehrlichia ruminantium (Accession number CR925677) with 87% of sequence similarity.

Phylogenetic analyses
The phylogenetic analysis generated by separate alignments of the three amplified genes confirmed that Rickettsia detected in this study belonged to the R. africae species (Figure 2 and Figure 3).The phylogenetic tree originated by the alignment of 10 gltA gene amplicons and 23 reference sequences showed that all the samples from this study clustered within the R. africae group (Figure 2).These results were also confirmed by the alignment of the ompA gene with reference strains (data not shown).Furthermore, an in silico restriction map (RFLP in silico) for the ompA gene was made using the sample under study and reference sequences (Figure not shown).The restriction analysis demonstrated that our samples had the same restriction profile as the reference sequences.This indicates that we are working with the same group of samples, which confirms the clustering observed in the phylogenetic trees.

Discussion
In this study, R. africae DNA was detected in A. variegatum and A. heabreum ticks collected from southern and central regions of Mozambique.Previous studies reported that A. hebraeum and A. variegatum were vectors and reservoirs of R. africae in some regions of Africa where there is a close interaction between humans and cattle parasitized by ticks [14,25,33,34].The results of this study confirm, for the first time, the circulation of R. africae, the causative agent of ATBF in A. hebraeum and A. variegatum, in Mozambique.
Southern African countries such as Zimbabwe, South Africa, Lesotho, and Swaziland, have also reported infection in ticks and transmission of R. africae to humans [22,23].Specifically, in Zimbabwe, R. africae infection rates above 80% were reported in A. hebraeum [35,36].Similarly, in our study 80% of adult A. hebraeum ticks were infected with R. africae.The high infection rate observed is probably due to transovarial transmission in ticks [34,36], which is consistently supported by an infection rate of 77%, found in pools of non-feeding A. hebraeum larvae recovered from pastures in Maputo Province.Both these results confirm the role of A. hebraeum ticks as reservoirs of R. africae in Mozambique.
Moreover, even though A. hebraeum is the main vector of R. africae in southern Africa, A. variegatum also plays an important role as a vector and is widely distributed in sub-Saharan Africa [24,[37][38][39].Our results show that A. variegatum is also a potential vector of R. africae in Mozambique, with an infection rate of 73%.In Mozambique, A. variegatum is present in all north and central regions of the country [40], while A. hebraeum is widespread in all the provinces south of Save River [41].
Although our results confirm the presence of the pathogen and the vector role of Amblyomma species using molecular approaches, these do not allow for a complete assessment of vector capacity.However, studies conducted by Socolovschi et al, [35] and Kelly & Manson [42] have experimentally proven the transmission of R. conorii and R. africae by A. hebraeum and A. variegatum and its role as a vector of Rickettsia.
The high rate of R. africae infection in ticks of the species A. hebraeum and A. variegatum reported in the present study suggests that the transmission to humans by these species should be considered a significant public health issue in the southern and central regions of Mozambique.
The presence of Rickettsia species of the SFG in ticks, and the significant degree of habitat sharing between domestic and wild animal hosts increase the likelihood of the pathogen's circulation in this habitat.The abundance of infected ticks at the sites included in this study, combined with the proximity between humans and domestic animals, suggest a high probability of infections occurring within rural human populations of Mozambique, where the problem is generally underestimated due to a lack of proper diagnosis.This scenario points to a possible occurrence of silent outbreaks of the disease, or to the existence of an endemic stability within the rural communities.

Conclusions
Through the use of molecular detection methods, this study confirmed the presence of SFGR R. africae in A. hebraeum and A. variegatum ticks, circulating in the south and central regions of Mozambique.Their geographic distribution in the studied areas and observed infection rates point to a potential underestimated threat to human health.It is, therefore, necessary to conduct further epidemiological studies involving hosts and vectors in domestic animals and human population interfaces to characterize the transmission of Rickettsia between these groups.
designed and supervised the study.All authors edited and approved the final manuscript.

Figure 1 .
Figure 1.Map of the study area, showing each district sampled.Adults Amblyomma ticks captured per district.B: Amblyomma larvae captured.

Figure 2 .
Figure 2. Bayesian based phylogenetic tree on partial gltA sequences showing the phylogenetic placement of R. africae strains among Rickettsia species.Numbers at nodes indicate values of posterior probabilities.Strains from this study are indicated in orange.

Figure 3 .
Figure 3. Bayesian based phylogenetic tree on partial ompA sequences showing the phylogenetic placement of R. africae strains among Rickettsia species.Numbers at nodes indicate values of posterior probabilities.Strains from this study are indicated in orange.

Table 1 .
Limit of detection for each dilution per primer and gene tested in the study.

Table 2 .
Percentage of identity of amplified genes to reference strains of R. africae.Strains under study are designed by: Cha-Changalane; Manh-Manhiça; Mass-Massinga; Msr-Mussorise; Mbn-Mambone.Dash symbols indicate that the gene was not sequenced for that specific strain.