Arboviruses are defined as " maintained in nature by a biological transmission cycle between permissive vertebrate hosts and haematophagous arthropods ". Some arboviruses do not fit this definition, but have been classified as such on ecological but not taxonomic bases. More than 100 arboviruses pathogenic for humans are recognised world-wide. Pathophysiology induced by these viruses gives raise to undifferentiated symptoms (flue-like, meningo-encephalitic, and haemorrhagic syndromes) that may lead to death. Haemorrhagic fever viruses belong to Arenaviridae, Bunyaviridae, Filoviridae and Flaviviridae. The majority of these viruses are found in tropical and subtropical countries, because of the large concentration of hosts and of vectors: this is the case of yellow fever virus, dengue, West Nile encephalitis, and Rift Valley fever (RVF). Several of these viruses are also found in temperate areas of central, eastern and southern Europe. In France, three viruses receive particular attention: Tick-borne encephalitis virus, the mosquito-borne West Nile virus, this last being sporadically introduced by migratory birds, and hantavirus transmitted to humans by its rodent reservoir. Many questions remain on molecular and epidemiological grounds for arenaviruses, flaviviruses, bunyaviruses and filoviruses. Researches in our Unit are focussed on six important viral diseases that can lead to severe encephalitis and haemorrhagic fevers of public health concern in several countries, and on particular in French countries overseas and in countries where Pasteur Institutes are established : yellow fever, dengue, West Nile, RVF, Ebola, haemorrhagic fever with renal syndrome (HFRS). Our research and expertise activities on arboviruses and haemorrhagic fevers are conducted in collaboration with our partners in the international network of Pasteur Institutes. These collaborative works are very fertile thanks to mutual exchanges of experience, materials, and technologies.
Dengue and West Nile viruses (V. Deubel)
Dengue is a viral disease transmitted to humans by Aedes mosquitoes in tropical and subtropical areas. The infectious agent is a flavivirus of the Flaviviridae family, subdivided in four serotypes (dengue-1, -2, -3 and 4). Dengue virus infection may be asymptomatic or cause a high fever (DF) with flue-like symptoms that rarely lead to haemorrhagic manifestations (dengue haemorrhagic fever, DHF) with vascular permeability and coagulopathy. A fatal issue can be observed with a severe hypovolemic shock (dengue shock syndrome, DSS) or hepatitis. According to WHO, the number of annual DF cases are close to 100 millions with an estimation of 30,000 deaths. Factors responsible of the severity of the disease, whether linked to the host or to the virus are not known. Our research programmes are focussed on the studies of molecular markers of dengue virus biodiversity that determine their virulence, and of the mechanisms that control virus interactions with its host cells.
West Nile virus is maintained in nature by biological transmission between Culex mosquitoes and birds. The major enzootic areas for West Nile virus are located in Africa and viruses are introduced in emerging areas by migratory birds. Humans and horses are susceptible to the virus but the low viraemia makes them dead-ends in the cycle. West Nile virus was considered mild, causing only rarely encephalitis. However, it has caused since five years high fever and deadly encephalitis. These severe manifestations are observed since 1996 in humans, horses and birds, during epidemics in Europe, northern Africa, Middle-east, Russia and the USA. Neuron lysis and necrosis in the central nervous system (CNS) with diffuse inflammatory cells are observed in autopsies of humans and animals. However, molecular markers of virus virulence and of host susceptibility are unknown. We are currently studying these factors.
Mechanisms of virulence of dengue virus virulence.(P. Desprès, M.P. Courageot, A. Catteau, M. Lucas, M.P. Frenkiel)
Dengue virus induces apoptosis in mouse infected neurons. We have studied viral and cellular factors that contribute to dengue virus pathogenicity. A comparative study of virus variants that differ in a murine model of neurovirulence has shown that viral determinants in the envelope protein E and helicase NS3 can modulate the kinetics of apoptotic cell death in response to dengue infection.
(Collaborations : C. Duarte Dos Santos, F. Rey)
Study of the maturation of dengue virus particle(M.P. Courageot, P. Desprès, M.P. Frenkiel)
The induction of apoptosis in dengue virus-infected cells is linked to the efficacy of viral morphogenesis in the endoplasmic reticulum (ER) and to the viral replication. We have studied the maturation of the envelope glycoproteins.
First, the study of virus morphogenesis has shown that heterodimerisation of prM (the precursor of the membrane M protein) and E proteins in the ER is a preliminary step in virion formation. PrM and E protein folding depends on the maturation of their N-oligosaccharides in the ER and is modified in the presence of a-glucosidase inhibitors. Virus production in supernatant fluid of infected cells is consequently blocked.
Second, we have studied properties of the sequence at the junction of capsid C and prM proteins that is required for the translocation of prM into the ER. (Collaboration : F. Pénin)
Third, the role of prM and E proteins in the induction of apoptosis pathway has been studied by expressing independently fragments of these proteins under the control of an inducible promoter. The intraluminal ectodomain of the structural M protein may have a role in triggering cell apoptosis.
Analysis of the biological properties of flavivirus nonstructural protein NS1(S. Alcon, F. Coulibaly, M.-T. Drouet, F. Mégret & M. Flamand)
NS1, the non-structural glycoprotein of flaviviruses, is essential for virus viability. It has been shown to participate in virus replication although no proper biological activity has been related to it. In addition to the cell-associated form of the protein, NS1 may be secreted in a host-dependent fashion from mammalian and not vector-derived insect cells. We have been interested in characterising the regulatory mechanism of NS1 secretion. In dengue type 1 virus-infected green monkey kidney Vero cells, we found that NS1 is secreted as a major soluble form, characterised by biochemical and biophysical means as a unique hexameric species. Complete processing of the complex-type sugar appears to be required for efficient release of soluble NS1 into the culture fluid of infected cells as suggested by the repressive effect of N-glycan processing inhibitors swainsonine and deoxymannojyrimicin. We further analysed NS1 secretion in vivo, and found the protein circulating in patient sera during the acute phase of the disease. Altogether, these results suggest that the secreted form of NS1 may specifically contribute to the pathophysiology of the disease.
(Collaborations: M. Arborio, S. Dartevelle, A. Falconar, B. Goud, R. Hellio, J. Krinsje-Locker, J. Lepault, F. Nato, F. Rey, A. Talarmin)
The Rift Valley fever virus (Michèle Bouloy)
Member of the Bunyaviridae family (genus Phlebovirus), the RVF is the cause of haemorrhagic fever in human, and of abortion and malformations of the fetus in ruminants. Epidemics have occurred in Kenya, Somalia, Tanzania and Mauritania in 1997-1999. In October 2000, the virus has caused severe epidemics in Saudi Arabia and Yemen. The virus is enveloped and contains a genome composed of three RNA segments (L, M, S) of negative strand with the S segment of ambisense polarity. This segment encodes for two proteins, the nucleoprotein N and a non-structural protein NSs. Our major efforts are dedicated to the study of the biological properties of this protein and of its role in virus pathogenesis.
Rift Valley fever virus transcription(Nicola Vahsen)
The S RNA segment encodes two proteins with an ambisense strategy. These proteins are translated from two RNAs of opposite polarity : RNA encoding N is of antigenomic sense, whereas RNA encoding NSs has a genomic sense. Using a minigenome containing the CAT (Chloramphenicol Acetyl Transferase) gene, we have demonstrated that two proteins, the N nucleoprotein and the L RNA polymerase, are required for the transcription. In this system, the viral proteins and the CAT protein are produced in the cells transfected with the plasmids containing the corresponding genes under the control of the T7 polymerase. Thus, protein synthesis requires the presence of the T7 polymerase that is expressed by the vaccinia virus vTF7.3. However, this virus produces an important cytopathic effect (CPE) that modifies the metabolism of the infected cell. To circumvent this problem, we have used another system already used for the influenza virus reverse genetic. It uses two cellular RNA polymerases, the RNA polymerase I to synthesise the minigenomes that are not capped at their 5'-extremities and the RNA polymerase II to synthesise mRNAs involved in viral protein translation. Two types of plasmids are being constructed ; the genomic plasmids S-CAT, that contain non coding sequences and the CAT gene in the antigenomic sense under the control of the promoter for polymerase I and plasmids pN and pL under the control of the promoter CMV. This system will allow us to study the RVF virus replication, to look for the role of NSs protein, and to establish a reverse genetic system.
The NSs protein(F. Yadani, A. Billecocq)
In most of RVF virus-infected cells, the NSs protein is localised in the cell nucleus where it forms filaments. To better understand the mechanisms of the filament formation, NSs protein gene has been expressed in the Semliki forest vector. The analysis of different mutants shows that the ten amino acids at the C-terminus of the protein are involved in the filament formation but not in the translocation into the nucleus. The two serine residues in the C-teminus fragment are phosphorylated presumably by the kasein kinase. Substitution of these amino acids by alanine residues in NSs does not block the filament formation, suggesting that the protein phosphorylation does not play a role in its polymerisation. The cytoplasmic form of NSs is associated with the betaglobulin and colocalises with microtubules.(Collaboration : P. Denoulet, P. Roux et P. Gounon)
Clone 13 : a natural mutant (A. Billecocq, P. Vialat)
The clone 13 of RVF virus has been selected from a strain isolated from a human case in Central African Republic. This virus contains a deletion of 70% of the coding region of NSs protein gene and is attenuated for rodents. NSs protein does not form filament in the nucleus of clone 13-infected cells and is rapidly degraded in the cytoplasm. Vero cells infected with this clone can be serially cultured without CPE and produce viral proteins. The reassortment of clone 13 genes with those of a wild type virus has shown that the S segment contains markers of attenuation. We have recently demonstrated that a major role of NSs protein is to block the production of interferon type I.
(Collaboration : O. Haller)
Phylogenic analysis (M. Bouloy, A.A. Sall)
The coding sequence of NSs protein has been used to compare RVF virus strains isolated in different areas of Africa for 50 years. The strains are classified in three clades : western Africa, central and eastern Africa, Egypt. Sequencing of a strain isolated by Dr R. Swanepoel (Joanesburg, South Africa) during the epidemics in Kenya, 1997-1998, has shown its close genetic relationship with a strain isolated in 1991 in Madagascar, in the eastern Africa cluster.
National Reference Centre and WHO Collaborative Centre for Arboviruses and Haemorrhagic fevers .Development of new techniques (A. Billecocq, D. Coudrier, M.T. Drouet, B. Murgue, S. Murri, H Zeller, M Bouloy)
The recombinant proteins NP and VP40 of the Gabonese 1994 strain of Ebola virus have been expressed in E. Coli and purified. The glycoprotein GP of the same strain has been produced in Vero cells using the Semliki forest vector. These proteins have been used in an ELISA test for diagnosis of Ebola-infected patients or for the study of the serological prevalence in areas presumably endemic for the virus.
Two methods of rapid diagnosis using RT-PCR light-cycling have been established for RVF and Puumala (PUU) viruses. (Collaborations : A. Sall, D. Garin)
Outbreaks: detection and investigations(B. Murgue, S. Murri, D. Coudrier, I. Marendat, H. Zeller)
The CNR detected West Nile virus in horses on September 6th, 2000, and initiated a co-ordinated response including multiple partners (Ministry of Health, of Agriculture, and of Environment and other agencies) for covering the different aspects: veterinarian, medical, entomological, and ornithological.
In October 2000, haemorrhagic fevers were observed in central Guinea. The CNR was involved in etiological diagnosis and confirmed a yellow fever outbreak. Then it made the laboratory follow-up of the outbreak that was still ongoing at the end of the year 2000. (Collaborations : S. Zientara, J Hars, JP Durand, L Koivogui, M. N'Faly, J ter Meulen, C. Lagneau.)
Dengue (diagnosis, imported dengue)(B. Murgue, S. Murri, I. Marendat, H. Zeller)
Commercial tests for dengue diagnosis were compared with the same panel of sera in two locations: Paris (CNR) and Marseille (Institut de Médecine Tropical du Service de Santé des Armées). Results are under evaluation.
A prospective study on the importance of imported dengue in France was done and results are under analysis. (Collaborations: JP Durand, C Sadorge, L Pupin, X. Deparis)
Prediction and prevention of hemorrhagic fever with renal syndrome(H. Zeller, D. Coudrier, M. Bouloy, A. Billecocq, B. Murgue)
Clinical cases of hemorrhagic fever with renal syndrome (HFRS) of hantavirus infection are reported in the north-eastern part of France. Clethrionomys glareolus (bank whole) is the specific rodent reservoir for PUU virus that is present in France. Epidemics of HFRS are reported on a three year basis with the last outbreak in 1999 (115 cases). Similar situation is observed in Belgium. During epidemics, most cases are recorded in the Ardennes department. In 2000, the CNR which performed the serological diagnosis detected 68 cases (mean age 40 years, sex ratio M/F 4:1). A study is conducted for a better understanding of the HFRS epidemiology for prediction of human risks. The rodent aspect includes studies of population dynamics, spatial distribution and dynamics of the incidence of infection, and epidemiological models. Captures of rodent are performed in specific sites for measurements and blood sampling, and then release. A very low prevalence of PUU antibodies in Clethrionomys glareolus was recorded in the different sites this year. Specimens are now processed in Nancy for PUU serology, following the technology transfer there and confirmed in Paris. New tools using PUU recombinant antigen expressed in Semliki forest virus vector are under evaluation for human and rodent diagnosis.
Otherwise, a first study on hantavirus detection in rodents and in humans was conducted in Cambodia and in Madagascar with the Pasteur Institute partners. In both countries, high prevalence of hantavirus antibodies were recorded in rodents. A comparative phylogenic study of RNA sequences on the S segment has shown that hantaviruses in this region are genetically related to Asian type (Seoul-like) and strongly linked to the rodent species. (Collaborations: C. Penalba, M. Artois, P. Vuillaume, D. Pontier, F. Sauvage, Boué, C. Janot, J.M. Reynes)