Introduction

The protozoan Toxoplasma gondii is an obligate intracellular parasite responsible for toxoplasmosis. Its life cycle is complex and involves multiple hosts. Felids, as definitive hosts, excrete the parasite oocysts produced after sexual reproduction in their digestive tract [1,2]. Intermediate hosts (all warm-blooded animals) may then be infected by ingesting oocysts present in the environment [3]. Key reservoir species include rodents, which are felids preferred preys. Other ground-foraging animals, such as poultry, are good bioindicators of T. gondii circulation in the environment [4].

In humans, the infection source varies depending on hygiene and eating habits [2]. With an estimated 25% of the world population being infected asymptomatically [5], Toxoplasma infection does not appear to be a public health priority, especially on tropical continents such as Africa where much more morbide infections like malaria or tuberculosis are widespread [6]. Yet, toxoplasmosis may be lethal in some large at-risk groups like fetuses or immunocompromised patients [3], but also in immunocompetent patients with severe forms of toxoplasmosis mainly described in South America [7,8]. Some investigations have shown that a link exists between the parasite genotype and the virulence in both mouse [9] and human [5,6,10]. Though the underlying processes remain unknown, such a relationship enhances the importance of documenting T. gondii genetic diversity and structure. In sub-Saharan Africa, data on both the genetic diversity of T. gondii and the clinical forms associated with this zoonotic disease are scarce [11–14]. However, in West and Central Africa, it was shown that ocular toxoplasmosis could be an important health issue [15] and that cerebral toxoplasmosis was found to be the second cause of death after tuberculosis among HIV-positive patients [10]. In addition, a recent study conducted in France has described cases of severe toxoplasmosis imported from tropical Africa in immunocompetent patients and associated with strains belonging to the Africa 1 lineage [16]. Therefore, further data on the diversity of Toxoplasma from this continent are required to better understand the eco-evolution of the pathogen and the risks for human health.

The Republic of Benin, formerly Dahomey, is located in West Africa, surrounded by Togo, Nigeria, Burkina Faso and Niger. Currently, Benin has an important maritime commerce through its International Sea port of Cotonou [17] with importation of vehicles, rice, fuels and palm oil (https://www.tresor.economie.gouv.fr/Pays/BJ/commerce-exterieur-du-benin) and exportation of raw cotton (34.5% of exportations), coconuts, Brazil nuts and cashews and gold to, inter alia, India, Bangladesh, China, United Arab Emirates and Vietnam (oec.world/en/profile/country/ben/).

During the 17^th century, the region was referred to as the Slave Coast [18]. Benin was a major source of African slaves during the so-called Atlantic triangular trade: raw cotton was shipped from America to Europe where it was manufactured and then sent to Africa. There, it was traded for slaves who were deported to Americas (an estimated two to three millions of people were deported from the Bight of Benin; [19,20]) (http://revealinghistories.org.uk/africa-the-arrival-of-europeans-and-the-transatlantic-slave-trade/articles/commerce-and-collecting.html).

Recent evidence has shown that T. gondii was already present in Africa 2,000 years ago [21]. However, such extensive and long-term intercontinental exchanges may have fueled the dissemination of many organisms to and from coastal Africa, and this could include pathogens such as non-autochtonous lineages of T. gondii through the accidental (e.g. rats) or deliberate (e.g. cats) importation of their hosts from Europe [4,22]. Interestingly, invasive species introduced into Africa may have long remained along the coasts for a while before being disseminated further inland due to late development of inland transports (e.g. rats and mice in Senegal [23,24]; rats along the Benin-Niger corridor [25]). It is thus feasible that this past maritime trade have shaped differences in genetic diversity and population structure of T. gondii between continents, but also between coastal and inland areas within Africa. In such context, the main objective of the present study is to map T. gondii genetic diversity across diverse regions of Benin and to compare the observed diversity to that found elsewhere in Africa but also Brazil, final destination of millions of slaves for three centuries.

Methods

Results

Isolation, virulence and genotyping of T. gondii isolates from Benin

Out of the 119 seropositive poultry, 69 isolations were attempted and 45 ones were successful. Among the latter ones, 39 could be genotyped using 15-MS genotyping (Fig 1): 36 strains were Africa 1 (the eight typing markers matching with reference strain “FOU”; see S1 Table) with 27 distinct MLGs. They were virulent for laboratory mice with the rapid development of an ascitic fluid usually followed by euthanasia or death after puncture, with 50% of mice being dead at day 10 p.i. (Fig 2). Two others strains from the region of Ouidah were found to be clones of Type III (non-virulent for mice) and produced cysts in the brain of three laboratory mice four weeks p.i. (respectively a total of 80, 128 and 24 cysts per brain).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Geographical distribution of Senegalese (modified from Galal et al., 2019), Beninese and Gabonese Toxoplasma gondii strains.

Map of the distribution of Toxoplasma gondii of isolates fully genotyped by 15-MS genotyping. Size of pie charts correlate with the total number of isolates and colors indicate the different clonal lineages found.

https://doi.org/10.1371/journal.pntd.0008980.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Survival curve of Swiss mice infected by Africa 1 strains from Benin.

https://doi.org/10.1371/journal.pntd.0008980.g002

During the isolation protocol, seven strains lead to an apparently asymptomatic infection, with a positive serology but no brain cyst by direct microscopic examination and unsuccessful genotyping probably due to insufficient DNA amount (positive sample in qPCR but Ct > 34) or unsuccessful sub-inoculation of frozen aliquots back to France. Among those isolation attempts, only one of the inoculated digest could be genotyped and corresponded to Africa 4 variant TUB2 (293pb instead of 291pb) (S1 Table).

Genetic structure and genotypes geographical distribution

The different genotypes found in Benin and in two other African countries, Senegal [4,14] and Gabon [11] are illustrated in Fig 1. In Benin, Africa 1 is the dominant genotype from South to North.

The phylogenetic distance-based tree separates the strains from Africa, Europe and South America into three different haplogroups that correspond to Africa 1, Type III and Africa 4 genotypes (Fig 3). No clear structure was retrieved within the Africa 1 lineage. In contrast, the Type III group appears divided into three clusters: one made of only Senegalese genotypes, a second one with only Gabonese genotypes and a third one with the single Type III genotype from Benin together with some strains from Senegal and Gabon.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Global Neighbor-joining tree of genotypes inferred from calculated Cavalli-Sforza distances for 15 microsatellite markers.

https://doi.org/10.1371/journal.pntd.0008980.g003

Africa 1 population genetics within Benin and comparison with strains from Senegal, Gabon and Brazil

The sample size of Africa 1 strains allowed us to focus on that particular lineage and to perform population genetic analyses on 27 Africa 1 MLGs from Benin, six from Gabon, 12 for Senegal and 19 from Brazil (Table 1). Values point to similar diversity within Senegal and Brazil, a slightly higher diversity in Benin and a lower one in Gabon, where a possibly more clonal population is observed, although this was based on a smaller sampling size.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Toxoplasma gondii genetic diversity and genotypic diversity.

https://doi.org/10.1371/journal.pntd.0008980.t001

No significant genetic differentiation was found between pairs of localities within Benin (Table 2). The same analysis performed worldwide (i.e. Benin, Senegal Gabon and Brazil) shows that Benin and Senegal are the only two countries with no significant differentiation (p-value = 0.37) (Table 3) although no obvious common allele was found between each strains of these populations through the MS-genotyping. The minimum spanning network indicates that the Africa 1 strains isolated in Africa and Brazil are closely related (Fig 4); however, no country-specific genetic affinities are observed.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Minimum spanning network (MSN) showing the relationships between multilocus genotypes (MLGs) of Africa 1 lineage from Benin, Senegal, Gabon and Brazil.

Thick and dark lines show MLGs that are more closely related to each other.

https://doi.org/10.1371/journal.pntd.0008980.g004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. P-value for each population pair of Beninese regions across all loci (Fisher’s method).

https://doi.org/10.1371/journal.pntd.0008980.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. P-value for each population pair of countries across all loci (Fisher’s method).

https://doi.org/10.1371/journal.pntd.0008980.t003

Discussion

Over the last two decades, the picture of the genetic diversity of Toxoplasma gondii in Africa has been slowly drawn thanks to accumulating data and an increasing sampling size [13]. Nevertheless, West Africa remains one of the least documented region of the world in terms of number of isolates and genotypes described. This study significantly contributes to filling such a gap of knowledge with the description of 39 T. gondii genotypes from Benin, compared to available microsatellite data of T. gondii diversity from West (Senegal) and Central Africa (Gabon). It also allows us to investigate the influence of intercontinental exchanges and the potential heritage of the triangular trade on shaping T. gondii genetic diversity.

We found three different T. gondii lineages that circulate in Benin: Type III, Africa 4 and Africa 1. The Type III clone (2 strains), corresponding to a cosmopolitan lineage already described in various countries on the Africa continent [13], were found in Ouidah, a former Portuguese fort where slave export was once extensive [37]. Type III distribution and circulation remain mysterious, with description of this genotype all around the globe and in both anthropized and isolated areas [4,11,38].

One single Africa 4 sample was successfully genotyped from Boukoumbé, a village at the Togolese border, 80 km east of Natitingou in a remote area. Africa 4 lineage (equivalent to RFLP genotype ToxoDB#20) is a widespread lineage in Africa (Egypt, Ethiopia, Senegal, Mali and Gambia), and was repeatedly found in Asia (China, Sri Lanka, United Arab Emirates) [13]. However, the strain isolated in this study was an Africa 4 variant TUB2 (equivalent to ToxoDB#137 [39]), which is a quite uncommon variant of this lineage since it was only described in three African countries (Senegal, Ghana and Mali) [4,13]. This sample was directly genotyped from a digest since the strain isolation “failed”: no clinical virulence in laboratory mice and no cerebral cyst were observed despite the positive serological test as well as the positive qPCR reaction on the brain. However, the low DNA amount detected in mouse brains (Ct = 38.37) could explain the genotyping failure. This atypical pattern has already been observed in a previous study on Senegalese strains belonging to the Africa 4 lineage. In our study, four other isolation attempts within the same village (Boukoumbé), without successful genotyping, also showed a similar isolation pattern. Given the geographical proximity and the sharing of this phenotypic trait, it is highly probable that these strains also belong to the Africa 4 lineage. These observations may highlight a bias in the widely used T. gondii isolation protocol. The use of a mouse model could disallow the isolation and/or characterization of certain strain genotypes by selecting a part of the diversity inducing a lower DNA amount in the host brain (seemingly, the case for the Africa 4 lineage). The use of knocked-out immunosuppressed mice [40] may solve such a problem but brings ethical concerns and other technics like cell culture are to optimize. However, these two methods would be complicated to implement without proper infrastructures to maintain sterile conditions, which are usually lacking in tropical regions. Considering these limits, our protocol remains the reference technique for now.

Aside from these two lineages, Africa 1 strains were the most widespread (36/39 characterized strains) and were found all over Benin. This lineage is predominant in tropical Africa since it was found in 11 countries (from Guinea to Democratic Republic of the Congo and more sparsely in East Africa) [13] and is regularly detected in France among imported cases in African patients with symptomatic toxoplasmosis by the National Reference Centre (CNR) for Toxoplasmosis [16]. Interestingly, two Africa 1 strains from Cotonou were respectively identical for the 15 MS to two other strains isolated from a pregnant woman and an immunocompromised patient in Denmark [41]. Although the use of microsatellites may lead to some homoplasy, such a risk is probably limited here by the number of markers used, especially since our 15 loci are distributed along 11 out of the 14 Toxoplasma chromosomes [28]. Therefore, these patients were probably imported cases (infection in Africa), but due to the absence of associated data, we cannot confirm that these patients were from, or travelled to Benin.

In Benin, no genic differentiation between the Africa 1 populations from the four sampled regions was observed, which may support a recent dissemination of this lineage across the country, or reflect an extensive circulation of strains between the four Beninese regions. Its nearly exclusive occurence in all sampled regions is in great contrast with what was previously observed in Senegal [4] and Gabon [11] where two to five lineages are co-occurring in most regions (Fig 2). However, it has to be noted that port regions, the most likely regions in which introduced lineages are expected to be found, were not thoroughly sampled in Gabon as it is the case in Senegal and Benin. So how such different diversity patterns between these three countries could be explained?

First, the absence of the European Type II in Benin and notably on the coast, in contrast to the hypothesis of an introduction of this lineage during the triangular trade in Senegal [4], is surprising since the two countries had a rather similar precolonial and colonial history with Europe. Benin may have had an even more prominent role in the triangular slave trade, with nearly two million slaves leaving the Bight of Benin from 1501–1867 (and more than one million just from Ouidah) while “only” 145.000 slaves were deported from Saint-Louis in Senegal [19]. If the difference is not in the route of introduction of type II strains, the explanations could be related to different factors enabling the persistence of this lineage in these new territories.

One difference between Senegal and Benin is the climate [42]: Senegal is characterized by a typically Sahelian (in the north) and a Sudanian climate (in the south) while Benin has a more tropical climate, from tree savannas (in the north) to subequatorial rain regime (in the south). The possible relationships between this marked climatic demarcation and the Toxoplasma genotypes diversity has already been discussed [13]. In this article, Galal and collaborators proposed that tropical Africa has been isolated from new hosts for quite a long time due to natural barriers such as the Sahara desert and a dense tropical forest, thus resulting in a singular T. gondii diversity within hosts living in the tropical area with autochtonous lineages such as Africa 1 [13]. Highly productive biotopes such as the tropical environment probably allow potential hosts of T. gondii to allocate a large amount of energy to the evolution of their resistance to pathogens that are abundant there [43]. This could lead to a general increase in the virulence of pathogens [44], including strains of Toxoplasma circulating in that environment. These processes would have allowed the selection of more virulent lineages (as seen with experimental mouse models) such as Africa 1. This could also explain the lack of success of the type II lineage to propagate in such tropical environments [13,22,45] both in Africa such as Benin and Gabon, and in South America due to a lack of adaptation of this lineage to this new environment (host immune system and competition with pre-existing and better-adapted indigenous lineages).

The Africa 1 lineage is highly virulent (if not lethal) in laboratory mice [4,11], but also in wild-derived mice [46]. In Senegal, since their introduction to the coast through past maritime trade, the spatial expansion of invasive rodents, mainly the house mouse Mus musculus domesticus and the black rat Rattus rattus, was driven by road- and river-mediated commercial exchanges [24,47]. It was found that strains of Africa 1 lineage are today seldom found in the port regions of Senegal where the house mouse is the most abundant commensal rodent [24,47] and where non-virulent lineages (type II, followed by type III and Africa 4) appear to be far more common [4]. The opposite was noticed in the inland region of Kedougou where the house mouse has not disseminated yet (Fig 2). To explain this contrasting pattern in the geographical distribution of T. gondii lineages in Senegal, Galal and collaborators [4] proposed that the house mouse could have allowed the introduction of the intercontinental lineages type II and type III in the port regions of Senegal and the maintenance of this lineages in these regions, beside the non-virulent lineage Africa 4. The successful invasion of these regions by the house mouse would have gradually eliminated Africa 1 lineage from these areas due to its marked virulence for this species, which is probably the most important prey of cats in this areas. In Benin, a large small mammal survey could not provide any evidence of house mice with the only exception of Cotonou Seaport [25]. This rodent species has therefore not (yet) influenced T. gondii populations in this country. In support to such a hypothesis, a similar match between the relative abundance of house mice and the frequency of virulent (Africa 1 and/or 3) vs. non-virulent (Type III) Toxoplasma strains was observed in Gabon (Franceville vs. Makokou; Fig 2) [48].

At the opposite, the most abundant rodents captured in Benin are the native African Mastomys natalensis, beside the invasive black rats R. rattus and Norway rats R. norvegicus [25]. Mastomys natalensis was experimentally shown to be resistant to the virulent strains of Type I [49]; and is therefore probably also resistant to Africa 1 since it shares common virulence alleles with Type I [50]. Rattus norvegicus is known to be highly resistant to mouse-virulent T. gondii strains [51]. No data about the resistance of R. rattus to the different T. gondii strains is available, but its importance as a significant reservoir for T. gondii is doubtful, given that it is an arboreal species, probably having a limited exposition to the oocysts on the soil. The only study from Africa reporting data about the prevalence of T. gondii infection in R. rattus was from Senegal, where a very low molecular prevalence was found [14].

The composition of the host community, particularly the absence of M. m. domesticus from Beninese territory outside the port of Cotonou, together with the probable resistance of these hosts may explain the absence of type II lineage and the Africa 1 lineage predominance in Benin. Our results would therefore support the hypothesis of an important role of T. gondii hosts in giving a selective advantage to some T. gondii strains [52,53], shaping its population structure [4,14]. This would mainly concern hosts that actually play an important role in maintaining the transmission cycle of T. gondii, namely rodents. These small mammals are indeed the main prey of the domestic cat and are considered the main reservoir of T. gondii in the domestic environment. If true, one would expect that the Toxoplasma genetic diversity observed in 2018 in Benin will change in coming decades since the dissemination of invasive rodent species (especially the house mouse) is most probably an ongoing process in Benin [25]. Studies on the comparative virulence of different T. gondii strains on native African and invasive rodents should be helpful to test this hypothesis.

Although Africa 1 lineage has mainly been described throughout the tropical zone of the African continent (West and Central Africa) with numerous strains, it has also been reported more anecdotally in Brazil (equivalent to PCR-RFLP haplogroup 6) [29–35,39]. In previous studies using fewer markers, genetic proximity had already been found between some of these strains from the two continents and the authors hypothesized a role of the transatlantic slave trade in the 18th and 19th centuries [12,13]. In our study, despite a significant genic differentiation between certain geographical populations of Africa 1 lineage, other analyses (Neighbor-Joining tree, minimum spanning network) showed no genetic differentiation between strains of this lineage. It is expectable that, since the collapse of the trade activities linking Africa to South America by the end of the transatlantic trade, geographical populations of this lineage may have gradually diverged due to their isolation. In a former study, Su and collaborators also came to that conclusion using PCR-RFLP approach by classifying Africa 1 equivalent strains from Brazil mainly in the haplogroup 6, such as FOU (the Africa 1 reference), but also in the haplogroup 14 with an Africa 3 equivalent [39]. However, their analyses found a moderate genetic separation between these two groups, probably influenced only by the geographical separation. Unfortunately, our microsatellites analyses did not allow us to clearly certify the origin of Brazilian Africa 1 strains. Despite that, the wide distribution of this lineage in Africa (from Senegal to Uganda) and its high prevalence in most countries of West and Central Africa is supportive of an introduction starting from Africa to South America, as only limited populations of this lineage were found only in Brazil (T. gondii diversity have been thoroughly explored in South America) [13,45]. The inference of the Africa 1 evolutive history will probably require the use of more informative methods, such as high throughput genomics to build phylogenies based on complete genomes of these strains, which has not been done yet with T. gondii due to the lack of Africa 1 genomes published from these two regions [54].

As with many zoonoses, exploration of the wild cycle for a global understanding of a pathogen epidemiology should be considered, like in the Pendjari Park (Benin), but remains difficult on safety grounds. For T. gondii, except for a few single strains [4], no data exist for African wildlife and characterizing this T. gondii diversity will be one of the challenges in the coming years. Nevertheless, with this work, we point to the environmental differences potentially explaining the diversity observed in Benin, Senegal and Gabon.

Previous studies made it clear that port areas and human mediated activities seem to play a crucial role in shaping the epidemiology for a large number of zoonotic diseases such as toxoplasmosis [4,12,22]. These issues raise health challenges for the future in our globalized trading world. In this context, it remains important to continue studying T. gondii genetic diversity in the world and especially in Africa where data is still fragmented, particularly concerning the clinical aspects of toxoplasmosis in relation to its genetic diversity.

Supporting information

Acknowledgments

We would like to thank Florent Engelmann, the director of the Institut de Recherche pour le Développement (IRD) de Cotonou and its staff for their collaboration on the field mission in Benin. We also thank Dorothée Kindé-Gazard for allowing use to use her laboratory for the isolation protocol at the Faculté des Sciences de la Santé in Cotonou, and Melkior Kouchade, superior officer of Water, Forests and Hunting and Nagoya focal point in Benin for meeting and helping us build the Protocol on Access and Benefit sharing. Finally, yet importantly, we are incredibly grateful to the poultry owners in Benin for accepting to participate to this study.