Master of Science / Department of Diagnostic Medicine/Pathobiology / A. Sally Davis / Rift Valley fever virus (RVFV) is a mosquito-borne, zoonotic Phlebovirus that is a significant threat to ruminants and humans. RVFV is categorized as an overlap Select Agent by the Department of Health and Human Services and US Department of Agriculture. Therefore, the study of RVFV’s pathogenesis and the development of novel diagnostic tools for the prevention and control of outbreaks and virus spread is crucial. RVF is endemic to sub-Saharan Africa but has spread beyond the continent to the Arabian Peninsula indicating the competence of the virus to emerge in new areas. Thus, the high likelihood of RVF’s spread to other non- endemic countries also spurs the need for development and implementation of rapid diagnostic tests and surveillance programs. In the US, RVFV is a Select Agent, requiring BSL-3 enhanced containment practices for research work. First, we developed a method for the detection of RVFV RNA by reverse transcriptase real-time PCR (RT-qPCR) using non-infectious, formalin- fixed, paraffin-embedded tissues (FFPET). The results from FFPET RT-qPCR were compared to prior results for fresh-frozen tissues (FFT) RT-qPCR, as well as immunohistochemistry and histopathology completed on the same FFPET blocks. We developed a novel technique using a rapid and low cost magnetic bead extraction method for recovery of amplifiable RVFV RNA from FFPET. FFPET RT-qPCR can serve as an alternative tissue-based diagnostic test, which does not require a BSL-3 research facility. Second, we assessed the diagnostic accuracy and precision of a recombinant RVFV nucleoprotein based competitive ELISA (cELISA) assay to detect RVFV antibodies. The cELISA results were compared to the virus neutralization test, the gold standard serological assay for RVFV. This prototype cELISA is easy to implement, sensitive, specific, and safe test for the detection of antibodies to RVFV in diagnostic and surveillance applications. RVF is an important transboundary disease that should be monitored on a regular basis. The diagnostic tests developed and validated in this thesis could be used in endemic or non-endemic countries for the early detection of RVF and assist with the implementation of countermeasures against RVFV.

Doctor of Philosophy / Department of Diagnostic Medicine/Pathobiology / A. Sally Davis / William Wilson / Rift Valley fever virus (RVFV) is a zoonotic arbovirus that is a significant threat to livestock and humans. It is listed as #3 for most dangerous animal threats and is in the top 10 pathogens needing urgent research in preventative and control measures. Although RVFV has never been reported in the US or Europe, outbreaks outside the African continent have sparked renewed interest in developing diagnostics and vaccines to protect both agriculture and public health. Having specific and versatile diagnostics is critical for vaccine development and application. For example, diagnostic tools that aid in identifying key immunogens and understanding the virus-host interaction directly contribute to developing protective vaccines. Additionally, vaccines that are used prophylactically or in response to an outbreak require diagnostic tests to differentiate infected from vaccinated animals (DIVA). This is critical for assessing the return to ‘disease free’ status after an outbreak. Unfortunately, there are limited RVFV diagnostic tests that are versatile and DIVA compatible with the newest RVFV vaccines. We describe the development of several diagnostic tools that are DIVA compatible for detecting RVFV nucleic acid, antibodies, and antigens. First, we evaluate a fluorescence microsphere immunoassay (FMIA) for the detection of antibodies against a RVFV surface glycoprotein and the nucleocapsid protein. The targets developed in this assay provide the basis for a DIVA-compatible serological assay with a candidate RVFV Gn/Gc subunit vaccine, as well as, offer a multiplexing platform that can simultaneously screen for several ruminant diseases. Second, we describe a novel chromogenic in situ hybridization (ISH) assay to detect RVFV in formalin-fixed, paraffin-embedded (FFPE) tissues. This molecular assay offers a highly sensitive, multiplexing platform that detects RVFV RNA on the cellular level of diagnostic tissue samples. Moreover, we demonstrate the first application of ISH as a DIVA-compatible assay for candidate RVFV gene-deletion vaccines. Third, we provide working protocols for western blot (WB), immunohistochemistry (IHC), and immunofluorescence (IF) that use monoclonal or polyclonal antibodies against key RVFV antigens. These tools can be applied to pathogenesis research and used in the development of vaccine and therapeutic countermeasures against RVFV. The RVFV diagnostic methods developed and evaluated in this dissertation can serve as a model for developing diagnostic strategies for other transboundary animal diseases.

La fièvre de la Vallée du Rift (FVR) est une zoonose émergente provoquée par un virus. La FVR affecte principalement le bétail. Chez l'Homme, la maladie peut évoluer sous deux formes mortelles: une fièvre hémorragique et une encéphalite. Malgré l'importance du fonds génétique dans l'issue de la FVR, l'identité des gènes responsables reste inconnue. Nous avons étudié les facteurs génétiques qui déterminent la sensibilité de la lignée de souris MBT/Pas et la résistance de la lignée BALB/c à la FVR. Nous avons identifié trois QTLs sur les chromosomes 2, 5 et 11, nommés, respectivement, Rvfs1, Rvfs3 et Rvfs2. Une infection des lignées congéniques correspondantes, C.MBT Rvfs1, -2 et -3, a confirmé le rôle de la région Rvfs2 dans la sensibilité à la FVR. Une analyse pathophysiologique a montré que les souris C.MBT Rvfs2 et BALB/c développent précocement une hépatite. Les souris C.MBT Rvfs2 meurent de cette hépatite aiguë. En revanche, les souris BALB/c régénèrent leur foie, ce qui leur permet de mieux tolérer l'atteinte du foie. La majorité des souris BALB/c sont décédées plus tard d'une encéphalite. Ces observations montrent que les modèles étudiés reproduisent chacune des deux formes de la maladie observées chez l'Homme. Nous avons produit des lignées sous-congéniques pour la région Rvfs2 et avons testé leur sensibilité à la FVR. Les résultats ont été croisés avec une analyse des variants de structure et de régulation présents dans l'intervalle. Cette stratégie a permis d'identifier 3 gènes candidats : Rnf213, Cd7 et Fasn. La fonction du gène Fasn suggère qu'il puisse jouer un rôle lors de la régénération du foie et ainsi être le facteur génétique que nous recherchons. / Rift Valley fever (RVF) is an emerging zoonosis caused by an arbovirus. The disease affects mainly livestock, but it can have a severe impact on human health. In humans, RVF may progress into fatal outcomes due to acute hepatitis or encephalitis. Despite the influence of the host genetic background on the outcome of the disease, the identity of causative genes remains unknown. We studied the genetic factors determining the susceptibility of MBT/Pas and the resistance of BALB/c mouse strains to RVF. We identified 3 QTLs linked to survival on chromosomes 2, 5 and 11 and named them, respectively, Rvfs1, Rvfs3 and Rvfs2. The infection of the corresponding congenic strains, C.MBT Rvfs1, 2 and 3, confirmed the role of Rvfs2 on the susceptibility to RVF. A pathophysiological investigation showed that both C.MBT Rvfs2 and BALB/c mice exhibit early onset severe hepatitis. However, while C.MBT Rvfs2 died rapidly from liver failure, BALB/c mice tolerated the liver disease and regenerated the hepatic tissue. These mice eventually died at a later stage from encephalitis. These observations indicate that each of the studied mouse models recapitulates one form of the human disease. We generated subcongenic strains harboring the Rvfs2 region and tested their susceptibility to RVF. The results were combined with high-throughput analyses of the structural and regulatory variants found in the region. Our combined approach allowed the identification of three candidate genes: Rnf213, Cd7 and Fasn. The function of the Fasn gene suggests that it could play a role in the mechanism of liver regeneration and, thus, Fasn is our best candidate gene to account for the susceptibility phenotype.

Lumpy skin disease (LSD) and Rift Valley fever (RVF) are transboundary viral diseases occurring in Africa and the Middle East (e.g. Israel, Saudi Arabia and Yemen) with increasing potential for global spread. Although the role of wildlife in the epidemiology of these diseases is still not clearly understood, the African buffalo (Syncerus caffer) is thought to play a role in the epidemiology of these diseases. This study sought to expand our understanding of the role of buffalo in the maintenance of RVF and LSD by determining seroprevalence to these viral diseases in buffalo during the inter-epidemic period. Lumpy skin disease is endemic in Africa, and has spread to the Middle East (e.g. Israel); consequently there is a high risk of lumpy skin disease virus (LSDV) expanding its geographical distribution to other areas and due to its economic importance it is included in the list of Notifiable Diseases of the World Organization of Animal Health (OIE). The African buffalo is also suspected to play a role in the epidemiology of RVF. Like LSD, RVF was, until recently, only endemic in Africa. However, it spread to the Arabian Peninsula (Saudi Arabia and Yemen) in 2000 exacerbating concerns that it will extend to other regions of the world. Studies have already established that competent mosquito vectors for RVFV exist in North America and other parts of the world. A total of 248 buffalo sera was tested for antibodies to capripoxviruses and neutralising antibodies against LSDV and RVFV using an indirect enzyme-linked immunosorbent assay (I-ELISA) as well as the serum neutralisation test (SNT). The samples were obtained from the Kruger National Park (KNP) and Hluhluwe-iMfolozi Park (HiP) in South Africa. The prevalence of antibodies to LSDV and RVFV in the sera tested was 70/248 (28.2%) and 15/248 (6.1%), respectively using an I-ELISA. The LSDV I-ELISA, using a sheeppox virus as antigen, has not been validated for use in African buffalo. The high percentage of LSDV positive antibody results obtained in this study is however a concern. Results obtained is in contrast with other published results as well as results obtained with the SNT for antibodies against LSDV. The SNT is currently the gold standard for LSDV antibody testing. Using this test for LSDV in this study, 5/66 (7.6 %) samples tested positive. The results of the RVF I-ELISA, which had previously been validated for use in the African buffalo, correlated with the SNT results. From 12 SNT RVFV-positive sera, 3 (25%) had very high SNT titres of 1:640. Neutralising antibody titres of more than 1:80 were found in 80% of the positive sera tested. Eleven buffaloes (4.4% of the total samples) also showed evidence of antibodies to both LSDV and RVFV. The results obtained in this study complement other reports indicating the role of African buffalo in the epidemiology of these diseases during inter-epidemic periods. / Dissertation (MSc)--University of Pretoria, 2012. / Veterinary Tropical Diseases / unrestricted

Impulsive differential equations are useful for modelling certain biological events. We present three biological applications showing the use of impulsive differential equations in real-world problems. We also look at the effects of stability on a reduced two-dimensional impulsive HIV system. The first application is a system describing HIV induction-maintenance therapy, which shows how the solution to an impulsive system is used in order to find biological results (adherence, etc). A second application is an HIV system describing the interaction between T-cells, virus and drugs. Stability of the system is determined for a fixed drug level in three specific regions: low, intermediate and high drug levels. Numerical simulations show the effects of varying drug levels on the stability of a system by including an impulse. We reduce these two models to a two-dimensional impulsive model. We show analytically the existence and uniqueness of T-periodic solutions, and show how stability changes when varying the immune response rate, the impulses and a certain nonlinear infection term. The third application shows how seasonal changes can be incorporated into an impulsive differential system of Rift Valley Fever, and looks at how stability may differ when impulses are included. The analysis of impulsive differential systems is crucial in developing more realistic mathematical models for infectious diseases.

During March 2008 a suspected outbreak of Rift Valley fever was reported on a farm in the Bela-Bela area, Limpopo Province of South Africa. The affected dairy farm, where no vaccination programme against RVF were practiced, applies an intensive farming system with 300 Holstein Friesland cattle (calves included) as well as 200 Pedi sheep on the farm. Seven calves died on this farm but no apparent clinical disease was reported in cattle as well as in sheep. During the outbreak blood samples from cattle and sheep were taken and the animals were re-sampled 8 weeks later. A set of sera was also collected from cattle on a neighbouring farm. The aim of the study was to determine the extent of the outbreak by evaluating if the virus had also infected other animals on the affected farm as well as on a neighbouring farm. During the first blood collection 233 samples were taken from cattle and 73 from sheep on the affected farm; 55 blood samples were taken from cattle on a neighbouring farm. A second blood collection was only done on the affected farm and 234 cattle and 85 sheep were bled. All the sera collected were tested by an IgM-capture ELISA and by an indirect IgG ELISA. Selected IgM positive (n=14), IgG positive (n=23) and samples negative for both IgM and IgG (n=19) were then tested by the serum neutralization test (SNT). Sera from IgM positive (14) and negative (20) animals were also tested by a TaqMan PCR. Results from the affected farm showed that 7% (16/233) of cattle samples were IgMpositive and 13.7% (32/233) IgG positive at the first collection of samples, and 2% were IgM-positive at the second sample collection. The number of cattle positive for RVF virus-specific IgG antibodies increased by 20.3% when compared to the first bled. Only 1.4% of sheep were both positive for anti-RVF virus IgM and IgG antibodies at the first collection; IgM-positive cases decreased to 1.2%, while IgG-positive cases increased to 2.4% at the second bled. Although no IgM-positive cattle could be found on the neighbouring farm, 5.5% of cattle were IgG-positive. The SNT confirmed most of the ELISA results. Three samples that tested positive for anti-RVF virus IgM and one anti-RVF virus IgG positive sample using ELISA tested negative using the SNT. Two samples that tested negative for both IgM and IgG antibodies using ELISA, tested low positive (1:10 and 1:20) using the SNT. All samples tested using a TaqMan PCR were negative. On the affected farm, apart from the seven calves that died, cattle were also infected. There was evidence of virus circulation on the neighbouring farm but the negative PCR results indicate that at the time the animals were sampled they were not viraemic. How the virus was introduced onto the farm is not clear. The possibility of low level virus circulation in animals and the reactivation of virus from endemic foci by the breeding of vector competent mosquitoes on the low-lying area on the farm in Bela-Bela may have led to ideal circumstances for an outbreak to occur. The fact that mostly cattle seroconverted suggests a higher host preference of the local population of mosquitoes for cattle rather than sheep. Copyright / Dissertation (MSc)--University of Pretoria, 2011. / Veterinary Tropical Diseases / unrestricted

No description available.

L'objectif de cette thèse est le développement d'un vaccin recombinant capripoxvirus protégeant contre la variole des ruminants, la Fièvre de la Vallée du Rift (FVR) et la Peste des Petits Ruminants (PPR) comme modèle vaccinal destiné aux pays atteints par ces infections. Une première partie de ce travail a consisté en une enquête sérologique en Tunisie pour évaluer les prévalences PPR et FVR. L'enquête menée a montré une séroprévalence PPR de 7,6% et l'absence de FVR. Le risque lié à une infection par le virus de la fièvre de la vallée du Rift n'est pas nul en raison de l'identification des vecteurs compétents Culex theileri et Culex pipiens dans les zones échantillonnées. L'élaboration du vaccin capripoxvirus FVR-PPR porte sur l'expression des gènes NSmGN-FVR et H-PPR où chacune des valences est insérée dans le site de la thymidine kinase et le site d'un analogue du récepteur à l'interleukine 8 respectivement. Le vecteur choisi pour la souche vaccinale Kenya Sheeppox-1. Bien que nos travaux aient conduit à l'obtention du capripoxvirus double recombinant, ce dernier n'a pu être purifié. L'alternative a donc été d'évaluer l'effet protecteur et l'immunogénicité induits par le simple recombinant capripoxvirus- NSmGN-FVR, qui est un produit de l'étape intermédiaire dans l'élaboration du double recombinant FVR-PPR. L'effet protecteur de notre construction a été validé par deux expérimentations chez des souris Mus m. musculus MBT/Pas, avec épreuve infectieuse. Le nombre de doses administrées, les voies d'administration ont été déterminants dans cette protection justifiée par l'obtention d'anticorps neutralisants anti-FVR. L'étude de l'immunogénicité a été réalisée sur un modèle caprin sans épreuve infectieuse, une séroconversion FVR a été observée. La lymphoprolifération et le typage des sous populations lymphocytaires ont été analysés. / The aim of this thesis was to develop a capripoxvirus based recombinant vaccine against ruminant pox, Rift Valley fever (RVF) and peste des petits ruminants (PPR) considered as a vaccine model for countries affected by these infections. The first part of the work consisted in a serological survey conducted in Tunisia to detect the PPR and RVF presence. A PPR seroprevalence of 7.6% has been found and no antibodies against RVF were detected. However, the risk of infection with rift valley fever virus persists since competent vectors such as Culex pipiens and Culex theileri has been identified in the sampled areas. The development of the RVF-PPR vaccine candidate is based on the NSmGN-FVR and H-PPR gene expression - where each of the genes is inserted into the thymidine kinase and the Interleukin 8 receptor analogue genes, respectively. The vector chosen is the vaccine strain Sheeppox Kenya-1. Although the double recombinant RVF-PPR has been produced, it could not be purified. The alternative was to evaluate the protection and the immunogenicity of the single recombinant capripoxvirus NSmGN-FVR, which is a product of an intermediate step of the process of the double recombinant preparation. The protection of our vaccine candidate has been performed by two mice experiments in Mus m. musculus MBT/Pas, with challenge. The number of doses, the route of administration played a key role in the protection confirmed by the presence of neutralizing anti-RVF antibodies. The study of the immunogenicity of the vaccine candidate was conducted in goats without challenge, RVF seroconversion has been shown. Lymphoproliferation studies and lymphocytes subpopulations typing have been analysed.

Master of Public Health / Department of Diagnostic Medicine/Pathobiology / Michael W. Sanderson / A zoonotic disease is any disease or infection that is naturally transmissible from vertebrate animals to humans. Over 200 zoonoses have been described (Zoonoses and the Human-Animal-Ecosystems Interface, 2013). Many zoonotic viruses are arboviruses, viruses transmitted by an infected, blood-sucking, arthropod vector (Hunt, 2010). There are several endemic arboviruses in the United States; some foreign arboviruses, such as Rift Valley fever (RVF) virus, are potential bioterrorism agents (Dar, 2013). Arboviruses, both endemic and foreign, threaten public health (Gubler, 2002) and therefore disease surveillance, vector control and public education are all vital steps in minimizing arboviral disease impact in the United States. Mosquito-borne disease threats, such as West Nile virus and Rift Valley fever, are constant concerns in the United States and globally. Current strategies to prevent and control mosquito-borne diseases utilize vector distribution, seasonal and daylight timing, and variation in population numbers. Climate factors, such as availability of still water for development of immature mosquitoes, shade, and rainfall, are known to influence population dynamics of mosquitoes. Using 1995-2011 mosquito population surveillance data from Fort Riley, Kansas, we compared population numbers of Culex tarsalis (Diptera: Culicidae), a vector of several arboviruses including West Nile virus and potentially Rift Valley fever, to a satellite-derived index of climate, the Normalized Difference Vegetation Index (NDVI) anomaly. No correlation between the population numbers and NDVI anomaly was observed, which contrasts with results from similar analyses in other locations. These findings suggest a need for continued investigation into mosquito population dynamics in additional ecological regions of the United States to better describe the heterogeneity of environment-population relationships within and among mosquito species.

La Fièvre de la Vallée du Rift (VFVR) est une arbovirose zoonotique affectant principalement les ruminants et les humains. Son éco-épidémiologie complexe implique de nombreuses espèces de vecteurs, d'hôtes et de voies de transmission. Ainsi, différents mécanismes de transmission et d'émergence sont impliqués dans la circulation du virus de la FVR (VFVR) et ceux-ci dans des écosystèmes contrastés d'Afrique, de la Péninsule Arabique et des îles du sud-ouest de l'Océan Indien, dont l'île de Madagascar.Par sa superficie, sa grande diversité éco-climatique et sa faune et flore endémique, Madagascar est considérée comme une île continent. On y retrouve, en effet, des écosystèmes variés plus ou moins favorables aux moustiques : semi-arides dans le sud-ouest, humides et froids sur les hautes terres centrales, per-humide dans l'est et humides et chaud dans le nord-ouest. Madagascar a été affectée par deux épidémies de FVR en 1990-91 puis 2008-09. Une étude menée lors de la dernière épidémie a montré que le virus avait largement diffusé dans l'île de façon hétérogène.Compte tenu de la complexité des mécanismes de transmission de la FVR et de la diversité des écosystèmes de Madagascar, nous avons supposé que cette hétérogénéité spatiale était due à des mécanismes de transmission et d'émergence qui variaient en fonction des écosystèmes de l'île. Ainsi, le premier objectif de ce travail de thèse étaient de déterminer les mécanismes et les dynamiques de transmission de la FVR inhérents aux différents écosystèmes de Madagascar. Le second objectif a été d'identifier les mécanismes d'émergence de la FVR à Madagascar et de déterminer s'il sera possible, et nécessaire, de prédire cette émergence à l'échelle des écosystèmes.Dans le cadre de ce travail de thèse deux enquêtes sérologiques nationales, l'une bovine (2008) et l'autre humaine (2011-13) ont, premièrement, été analysées par un modèle linéaire mixte généralisé afin d'identifier les facteurs environnementaux et comportementaux favorables à la circulation du virus chez les bovins et les humains. Deuxièmement, deux enquêtes sérologiques bovines, l'une réalisée en 2008 et l'autre en 2014, ont été analysées pour reconstruire la dynamique de transmission de la FVR dans les différents écosystèmes de l'île. Cette reconstruction a été réalisée à partir de données de séroprévalence et d'âge inclues dans un modèle Bayésien hiérarchique pour estimer la force d'infection annuelle de 1992 à 2014. Enfin, afin de faire le lien biologique avec les résultats des travaux menés à une échelle nationale et de décrire les mécanismes de transmission à une échelle fine, des enquêtes longitudinales entomologiques et sérologiques ont été réalisées entre 2015 et 2016 dans un écosystème à risque. Et ceci, afin de décrire la transmission saisonnière du VFVR chez les ruminants associée à la dynamique de transmission des vecteurs potentiels.Nos résultats ont montré que la région du nord-ouest de l'île est une région à risque de transmission. D'une part, elle est constituée d'environnements associant une forte densité de bovins à des zones humides, inondables et irriguées, favorables aux espèces d'Anopheles et Culex. D'autre part, le VFVR semble avoir circulé de façon relativement intense lors de la période inter-épizootique de 1992 à 2007, puis sa transmission a soudainement augmenté en 2007-2008, ce qui est concomitant avec l'apparition des foyers de FVR en 2008. Pour finir, 6 ans après l'épidémie de FVR à Madagascar, le virus semble toujours circuler à bas bruit dans la région. Cette circulation étant probablement due à une transmission vectorielle favorisée par l'abondance de vecteurs potentiels dans la région.Les résultats de ces différents travaux nous ont permis de présenter des hypothèses de transmission dans les différents écosystèmes de l'île et ainsi de proposer des stratégies de surveillance, de prévention et de lutte contre la FVR adaptées au contexte de Madagascar. / Rift Valley fever (RVF) is a zoonotic vector-borne disease affecting ruminants and humans. Its complex eco-epidemiology involves several species of vectors, hosts and transmission routes. These particularities allowed the circulation of RVF virus (RVFV) in a variety of ecosystems involving different transmission and emergence mechanisms. Indeed, the RVFV has affected contrasted eco-regions in Africa, Arabian Peninsula and South-West Indian Ocean islands, including Madagascar.Madagascar is considered as a continent island due to its ecological diversity and its endemicity level of the flora and the fauna. In particular, the variation of the Malagasy ecosystems (semi-arid in the south, humid and cold in the highlands, humid and warm in the north-west and per-humid in the east) has an impact in their presence and /or the relative abundance of some mosquito species. Madagascar was heavily affected by RVF in 1990-91 and 2008-2009, with evidence of a large and heterogeneous spread of the disease.Thus considering the diversity of RVF eco-epidemiological cycles and the variety of Malagasy ecosystems, we hypothesized that, in Madagascar, the mechanisms of transmission would be different according to these ecosystems. Therefore, the first objective of this thesis was to understand the mechanisms and the dynamics of transmission of RVFV in the different ecosystems. The second objective was to determine the mechanisms of emergence of RVFV and if it would be necessary and possible to predict the emergence of RVFV outbreaks according to the ecosystems.Firstly, we analyzed both cattle and human serological data performed at the national level using generalized linear mixed models to identify the environmental and behavioral factors associated with RVF transmission in both cattle and human. Secondly, we reconstructed the dynamic of transmission of RVF in the different Malagasy ecosystems. Seroprevalence data of cattle of known age were fitted using Bayesian hierarchical models to estimate the annual force of infection from 1992 to 2014. Thirdly, to understand the biological process link to the mechanisms of transmission at the national scale, we investigated the fine scale mechanisms of transmission of RVFV in pilot area of an at-risk region. We, thus, performed both longitudinal entomological and serological surveys between 2015 and 2016, in order to describe the seasonal transmission of RVFV among ruminants and its association with the dynamics of RVFV potential vectors.Our results showed that the northwestern part of Madagascar is an at-risk region for RVFV transmission. On one hand, it is characterized by high cattle densities associated with humid, floodplain and irrigated areas suitable for RVFV potential vector like Anopheles and Culex species. On the other hand, RVFV had probably circulated intensively in the region during the 1992-2007 inter-epizootic period and its transmission increased suddenly in 2007-08, almost concomitantly with the first outbreaks recorded in 2008. Finally, RVFV was still circulated in the northwestern region at low level, 6 years after the last epidemic. This circulation is likely due to vectorial transmission favoring by the abundance of several potential vectors of RVFV in this pilot region.Finally, our better understanding of the mechanisms of transmission of RVFV throughout Madagascar allowed us to propose hypothesis of transmission in different ecosystems of Madagascar and consequently refine strategies for RVF surveillance and prevention.