Yellow Fever Virus

Alan DT Barrett, University of Texas, Galveston, Texas, USA

Yellow fever virus is a flavivirus that causes the disease yellow fever. The virus is found in tropical South America and subSaharan Africa. The disease varies from a febrile illness to a lethal haemorrhagic fever.
. History

Secondary article
Article Contents
. Classification

. Structure . Replication . Epidemiology

Classification
Yellow fever virus (YFV) is the prototype virus of the family Flaviviridae, which takes its name from the Latin for yellow (avus). The virus is a member of the genus Flavivirus, which contains 67 human and animal viruses (see below). On the basis of their ecology, aviviruses have been termed arboviruses, or arthropod-borne viruses, to denote the fact that many are transmitted between vertebrate hosts by mosquitoes or ticks. Serological relationships between aviviruses have been determined by antibodies (monoclonal or polyclonal) associated with the biological activities of haemagglutination-inhibition (HAI) and neutralization. Both biological activities are encoded by the envelope (E) protein of the virus. Complement-xation activity has been associated with the nonstructural protein one (NS1) protein. The aviviruses have been termed a Hall of Mirrors by Karl Johnson because of the extensive serological crossreactivities between dierent aviviruses. Neutralization tests have been used to identify eight serological subgroups, together with 17 viruses that were not assigned to any serogroups due to poor crossreactivity with other members of the Flavivirus genus (Calisher et al., 1989). Interestingly, although YFV is considered the prototype avivirus, it is serologically distinct from all other aviviruses identied to date. Recently, nucleotide sequencing of a region of the nonstructural protein 5 (NS5) gene of most aviviruses has resulted in identication of genetic relationships that closely follow those of serological relationships (Kuno et al., 1998); however, the genetic analysis has allowed more detailed relationships to be established than was possible by serological studies. The studies of Kuno et al. indicate that, on the basis of genetic homology, YFV is closely related to nine other aviviruses (Banzi, Bouboui, Edge Hill, Jugra, Saboya, Potiskum, Sepik, Uganda S and Wesselsbron viruses).

. Clinical Features . Control

century, with large epidemics reported in both the Old World and the New World. In addition, epidemics were also reported in Europe following movement of the virus in sailing ships. Grin Hughes is thought to have been the rst to use the term yellow fever in his book Natural History of Barbados (1750), with yellow derived from the jaundice which results from damage to the liver. There were many debates over the agent and transmission of yellow fever at the end of the nineteenth and start of the twentieth centuries. Carlos Findlay was the rst to propose that mosquitoes transmitted yellow fever. Subsequently, pioneering studies by Walter Reed and coworkers demonstrated that the agent that caused the disease was ltrable (i.e. it was a virus) and was transmitted by the bite of the mosquito Aedes aegypti. Thus, YFV was the rst animal virus shown to be transmitted by an insect. The virus itself was not isolated until 1927, when French and American researchers in Africa isolated the strains French viscerotropic and Asibi, respectively. During the 1930s, both of these wild-type strains were used to derive live attenuated vaccines known as the French neurotropic vaccine and 17D, respectively. The rst South American strain, JSS, was identied in Brazil in 1935, and the jungle transmission cycle involving sylvatic mosquitoes and nonhuman primates was rst identied in Brazil in 1932.

Structure
Morphology
Virus particles are spherical in shape, approximately 50 nm in diameter, and have a lipid envelope. The lipid envelope is derived from host cell membranes and constitutes 1520% of the total weight of the virus particle. Carbohydrates represent 910% of the weight of virus particles and are found as glycolipids and glycoproteins. The small basic capsid (C) protein surrounds the genome of the virus and the envelope contains two viral proteins known as envelope (E) and membrane (M). Two types of virions are recognized: mature extracellular virions contain M protein while immature intracellular virions contain precursor M
1

History
YFV has played an important role in the history of animal virology. The virus was rst recognized as a human disease of major public health importance in the eighteenth

ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net

Yellow Fever Virus

(prM), which is proteolytically cleaved during maturation to yield M protein.

antibodies. The M protein also contains epitopes recognized by neutralizing antibodies.

Physiochemical properties
Mature virions sediment at approximately 200 S and have a buoyant density in sucrose of 1.19 g mL 2 1. Virions are stable at mild alkaline pH (8.09.0); they are inactivated by acidic pH, temperatures of 398C or higher, detergents, organic solvents and ultraviolet light.

Replication
The initial events of the virus replication cycle involve the binding of virus to cell receptor(s), which is mediated by the E protein. No putative cell receptor molecules have been reported to date. Uptake of the virus particle into cells is via receptor-mediated endocytosis followed by pH-dependent membrane fusion activity to release the virus nucleocapsid into the cytoplasm. The input virus does not contain a viral RNA-dependent RNA polymerase. Thus, the positive-sense genomic RNA is translated to generate the nonstructural proteins required for replication of the virus, including the RNAdependent RNA polymerase. RNA replication is associated with membranes and begins with transcription of the input genomic RNA to synthesize complementary negative-sense strands, which are then used as templates to transcribe positive-sense genomic RNAs. These are synthesized by a semiconservative mechanism involving replicative intermediates (i.e. containing double-stranded regions as well as nascent single-stranded molecules) and replicative forms (i.e. duplex RNA molecules). Synthesis of negative-sense RNA continues throughout the replication cycle. Proliferation and hypertrophy of intracellular membranes is a common feature of avivirus-infected cells both in cell cultures and in animals. The replication complex sediments with membranous fractions of infected cell extracts. Translation starts at the rst initiation codon (AUG) of the single long open reading frame. Once the polyprotein has been translated, it is processed by cellular proteases (signal peptidases) and the viral NS2B-NS3 serine
3NCR RNA

Genome
The genome consists of a single-stranded, positive-sense (i.e. infectious) ribonucleic acid (RNA) of approximately 11 000 nucleotides. The genome of the 17D-204 vaccine virus is 10 862 nucleotides in length (Rice et al., 1985). The 5-terminus of the genome possesses a type I cap (m-7GpppAmp) followed by the conserved dinucleotide AG. There is no terminal poly(A) tract at the 3-terminus. The gene order is C-prM-E-NS1-NS2A-NS2B-NS3NS4A-2K-NS4B-NS5 (Rice et al., 1985) (Figure 1). All viral proteins are produced initially as a single polyprotein that is posttranslationally processed by viral and cellular proteases.

Proteins
As stated above, the virions contain three structural proteins C, M and E. The C protein has a basic charge to facilitate its interaction with the viral genome. The E protein is glycosylated at one to three sites, depending on the strain; prM is also potentially glycosylated at one to three sites. The E protein is the viral haemagglutinin (i.e. the protein that binds to red blood cells) and contains most of the epitopes recognized by neutralizing antibodies. Thus, the E protein is the major target of neutralizing
5NCR cap Structural proteins

Nonstructural proteins

prM

NS1

NS2A

NS2B

NS3

NS4A 2K NS4B

NS5

Polyprotein

NS3 2K

Posttranslational processing C pr prM M Signal peptidase site Unique site NS2BNS3 protease site E NS1 NS2A NS2B NS3 NS3 NS5 NS4A NS4B NS5

NS3 Protease, helicase, NTPase Methyltransferase, RNA polymerase

Figure 1 Organization of the flavivirus genome and expression of proteins (genes are not drawn to scale).

ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net

Yellow Fever Virus

protease (Figure 1) to generate structural and nonstructural proteins. In addition to the three structural proteins C, prM and E, eight nonstructural (NS) proteins are found in virus-infected cells: NS1, NS2A, NS2B, NS3, NS4A, 2K, NS4B and NS5. Few of the nonstructural proteins have been studied in detail. NS3 is a multifunctional protein. The N-terminal one-third of the protein forms the viral serine proteinase complex together with NS2B. The Cterminal portion of NS3 contains an RNA helicase domain involved in RNA replication, as well as an RNA triphosphatase activity involved in the formation of the 5terminal cap structure of the viral RNA. Two enzymic activities have been assigned to NS5: the RNA-dependent RNA polymerase and the methyltransferase activity necessary for methylation of the 5 cap structure. NS1 is an unusual nonstructural protein as it is glycosylated at two sites. The functions of NS2A, NS4A and NS4B are poorly understood; however, current evidence suggests that NS2A, NS2B, NS4A and NS5 are all part of the replication complex and that NS1 is involved in RNA synthesis and virus assembly (Chambers et al., 1990). Virus particles can rst be observed in the rough endoplasmic reticulum, which is believed to be the site of virus assembly (i.e. interaction of genomic positive-sense RNA molecules and structural proteins C, prM and E). Progeny virions assemble by budding through intracellular membranes into cytoplasmic vesicles. These immature virions (prM rather than M protein) are then transported through the membrane systems of the host secretory pathway to the cell surface, where exocytosis occurs. Shortly before virion release, in the late Golgi or transGolgi network, the prM protein is cleaved by furin or a furin-like cellular protease to generate mature virions that contain M protein. Immature virions have low infectivity compared to mature virions. Flavivirus-infected cells also release a noninfectious subviral particle (containing prM and E proteins) that has a lower sedimentation coecient than whole virus (70 S versus 200 S) and exhibits haemagglutination activity (slowly sedimenting haemagglutinin, SHA). Host cell macromolecular synthesis is not shut o during virus replication and is not decreased until cytopathic eect is evident late in the infection process. The virus replicates in a variety of vertebrate and mosquito cell cultures. The major animal hosts are primates. Note that, although YFV is described as a mosquito-borne virus, there have been reports of isolation of the virus from ticks in Africa, indicating that the virus replicates in both mosquito and tick cells.

Epidemiology
YFV is only found in tropical South America and subSaharan Africa (Figure 2); the virus is not found in

Asia. Current YF activity is reported in the World Health Organization Weekly Epidemiological Record. The reason for the lack of YFV in Asia is not understood. The mosquito vector, Ae. aegypti, is found in Asia and laboratory studies indicate that Asian strains of Ae. aegypti can transmit YFV. A number of hypotheses have been proposed to explain this phenomenon but none have been proven. Yellow fever is considered to be a reemerging disease due to the increase in its incidence in the last 20 years. It is estimated that there are approximately 200 000 cases each year, of which the vast majority ( 4 90%) of cases occur in West Africa. The case-fatality rate is approximately 20%. There are continual reports of yellow fever in West Africa, whereas there are only occasional reports from East Africa; the last was in Kenya during 19921993. The majority of yellow fever activity in South America is found in the Amazon Basin, with occasional reports from surrounding areas, including Trinidad. Approximately, 200 cases are reported from South America each year with a case-fatality rate of 65%. This high gure is in part due to identication of the disease in some areas by histopathological examination of livers from fatal cases. This has led to speculation that the true incidence of yellow fever may be 10-fold greater than reported numbers of cases. The majority of cases in the Americas are reported from Peru. To date, the complete genomes of only four strains of YFV have been determined: three from West Africa and one from Trinidad. The molecular epidemiology of YFV has been poorly studied but all of the studies indicate that at least four genotypes of virus exist: two in Africa (West Africa, and Central and East Africa) and two in South America (one in Peru and Bolivia, and the other in all other parts of Central and South America) (Lepiniec et al., 1994; Chang et al., 1995; Wang et al., 1996). The genome of the West African genome is 4050 nucleotides longer than that of other genotypes, owing to duplication of nucleotide sequences in the 3 noncoding region. Phylogenetic analyses indicate that YFV originated in Africa, and that the virus was introduced into the Americas from West Africa, probably by the slave trade. Although genotypes of YFV can be identied based on nucleotide variation, there is very little variation at the amino acid level and all evidence indicates that the 17D vaccine is eective at controlling disease caused by all genotypes of YFV. Although primates are the vertebrate host of YFV, the virus infects a number of animal species. It causes a viscerotropic disease in humans, monkeys and hedgehogs. Specically, the virus infects the liver, kidneys and heart, and is not associated with encephalitis; however, it causes neurotropic disease in mice, hamsters and guinea-pigs, usually only after intracerebral inoculation. A number of vertebrates (e.g. sloths) have been found to have antibodies against YFV, but no viruses have been isolated. The role of these vertebrates in the ecology of the virus is unclear.

ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net

Yellow Fever Virus

Mo
Western Sahara Mauritania

roc

co
Algeria

Tunisia

Libya

Egypt

Mali Senegal Gambia Guinea Guinea Bissau Burkina Faso

Niger Chad Nigeria

m Ca er

Sudan

Eritrea Djibouti

Sierra Ivory Leone Coast Liberia

Ghana Togo Benin

Ethiopia n oo Central African ia Republic al m Democratic So Equatorial Republic Kenya Guinea of Congo Gabon Rwanda Burundi Cabinda Tanzania

Co

Angola Zambia Namibia

Zimbabwe Botswana

Swaziland (a) South Africa Lesotho

Venezuela Colombia Ecuador

Peru Brazil Bolivia

Pa

Chile

rag

ua

Argentina

(b)

Figure 2 Geographic areas in (a) Africa and (b) South America (shown in yellow) where yellow fever disease has been reported. Not all areas have continual yellow fever disease.

ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net

Uruguay

Suriname

French Guiana

Guyana

M oz am Malawi bi qu e

Ug an

ngo

da

Yellow Fever Virus

YFV is found in tropical regions of Africa and South America, and is transmitted between primates by diurnally active, tree hole-breeding mosquitoes. Humans infected as part of this sylvatic cycle have what is termed jungle yellow fever (Figure 3a). The main vector in Africa is Ae. africanus, while in South America it is Haemagoggus species. Other mosquito species involved in transmission include Ae. furcifer, Ae. vittatu, Ae. luteocephalus and Ae. simpsoni in Africa, and Sabethes chloropterus in South America. Primate species involved as vertebrate hosts of the virus also dier by geographic area. Ae. aegypti is a domestic vector associated with urban yellow fever (Figure 3b), the transmission of YFV between humans. Although periodically reported in Africa, there have been no cases of urban yellow fever in the Americas since 1942, even though Ae. aegypti has reinfested the majority of tropical South America in the last 20 years. Although yellow fever is considered to be a mosquito-borne disease, Amblyomma variegatum ticks have been shown to be naturally infected with the virus in Central Africa. Arthropods are involved in biological transmission of the virus, i.e. virus multiplies in the arthropod. Specically, the arthropod acts as a vector to transmit the virus between vertebrate hosts. Vertical (transovarial) transmission enables the virus to survive in eggs during the dry season. If an arthropod bites a virus-infected vertebrate that has a viraemia, the virus is transmitted to the arthropod when it takes its blood meal. Virus infects the midgut epithelium and disseminates from the midgut into the haemolymph and subsequently other tissue, in particular salivary glands and the reproductive tract. When the arthropod takes its next meal, virus present in the salivary glands is injected into a new vertebrate host. It takes between 7 and 14 days for virus to be taken up during one feeding and spread to the salivary glands for transmission to a susceptible vertebrate host (the extrinsic incubation period). It has been established that mosquitoes can be infected with as few as 10 plaque-forming units of virus, equating to viraemias of 104 plaque-forming units per millilitre of blood. Infection of the mosquito reproductive tract is thought to be important, as this enables eggs to become infected and allow overwintering of the virus when the climate is dry and/or cold. Diagnosis can only be achieved in a few laboratories with capabilities for serological testing, usually detection of anti-YFV immunoglobulin M (IgM) by enzyme-linked immunosorbant assay (ELISA) or an antigen-captureELISA. Other diagnostic procedures include isolation of virus from liver or serum in mice, mosquitoes or cell culture, although this is only possible during the rst 34 days of clinical symptoms of disease. Alternatively, liver histopathology demonstrating necrosis of hepatocytes and formation of Councilman bodies (see below) can be used for diagnosis. It is important to note that in areas of Africa and South America where the disease is found, the initial clinical symptoms are similar for many tropical diseases,

(a)

(b)
Figure 3 Transmission cycles of Yellow fever virus. The virus is transmitted between vertebrate hosts by various species of mosquitoes. The jungle (or sylvatic) cycle (a) involves monkeys and tree hole-breeding mosquitoes; the urban cycle (b) involves humans and the Aedes aegypti mosquito.

ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net

Yellow Fever Virus

particularly in mild cases. Also, progression to haemorrhagic fever does not always indicate yellow fever, as other viral haemorrhagic fevers are found in Africa and South America.

Clinical Features
YFV is classied as Biosafety Level 3 (indigenous or exotic agents with a potential for respiratory transmission, and which may cause serious and potentially lethal infection) (US Department of Health and Human Services, 1999). Following the bite of an infected mosquito, virus enters the vertebrate host via mosquito saliva. Initial virus replication takes place at or near the point of infection and spreads to lymph nodes. The viraemia enables the virus to spread to the spleen, bone marrow, heart, kidneys and liver. The liver is the target organ of the virus. The virus initially infects Kuper cells, followed by hepatocytes. Injury to hepatocytes is considered to be due directly to virus-induced eects rather than immunopathology, and is characterized by eosinophilic degeneration (also known as Councilman bodies). Necrosis of the liver results in loss of liver function, causing the skin of patients to turn yellow, hence the name given to the virus. Clinical disease varies from mild febrile inuenza-like infection to a haemorrhagic fever and fatality in 1020% of cases. Typical disease involves infection with the virus via a mosquito bite, followed by an incubation period of 36 days. Clinical symptoms occur suddenly and include fever (39408C), headache, muscular pains, nausea and vomiting, which persist for the next 35 days. The patient is viraemic during this phase and is able to spread the virus to blood-feeding mosquitoes. Following a brief period of remission for up to 2 days, the intoxication phase includes fever, nausea, vomiting and jaundice for a further 35 days. The latter includes increase in serum transaminase levels (aspartate aminotransferase (AST) and alanine aminotransferase (ALT)). YF disease is characterized by greater increases in AST levels compared to ALT. In severe cases there is haemorrhagic fever with bleeding from dierent orices, kidney dysfunction (detected as increase in albuminuria) and dehydration. Death takes place 710 days after the rst clinical symptoms. Survivors take 12 weeks to convalesce. Note that in mild cases patients may not have the intoxication phase. Virus infection results in the rapid appearance of an immune response. Neutralizing antibodies appear approximately 7 days postinfection and eliminate the virus. The role of the cellular immune response is not at present understood.

Control
There is no antiviral treatment to control yellow fever infections, rather supportive therapy is the norm. The
6

antiviral agent ribavirin has shown antiviral activity in cell culture when used at high concentrations, but this has not been correlated in animal studies. The major method of control is to use insecticides to control the mosquito vector. A live attenuated vaccine known as 17D is used to prevent yellow fever. This vaccine was developed in the 1930s by Max Theiler and colleagues and was responsible, in part, for him being awarded a Nobel Prize. Theiler and coworkers took the wild-type strain Asibi and passaged it in embryonated chicken eggs to derive strain 17D after 176 passages. Today, strain 17D is used as two substrains, known as 17D-204 and 17DD; both are used as vaccines. The 17D vaccine is probably the nest vaccine developed to date. One dose of vaccine induces protective immunity. The World Health Organization requires booster vaccination every 10 years but immunity is probably lifelong, as demonstrated by studies in volunteers. There are occasional allergic reactions to the egg antigens in the vaccine and the vaccine is contraindicated in pregnant women, immunosuppressed individuals and children under the age of 69 months, depending on the country, due to the risk of postvaccinal encephalitis. There is a very low rate of reversion to virulence (only 21 cases from over 300 million doses administered), and there is no record of a vaccinated individual succumbing to yellow fever. Reversion to virulence results in neurological disease and not the viscerotropic disease shown by wild-type virus. French workers developed a second live yellow fever vaccine in the 1930s by taking the wild-type French viscerotropic virus (FVV) and passaging it through mouse brain to develop the French neurotropic vaccine (FNV). The FNV was used from the 1930s until the early 1980s in French-speaking countries in Africa. Although the vaccine was very ecacious, it was associated with a relatively high rate of postvaccinal accidents in children, resulting in the vaccine being discontinued. Both FNV and 17D vaccine viruses have lost mosquito competence compared with wild-type YFV. The molecular basis of attenuation and virulence of YFV has been investigated. The genome of wild-type strain Asibi has been compared to the 17D-204 and 17DD vaccine substrains (Hahn et al., 1987). There are 20 amino acid dierences between the 17D vaccine substrains and wild-type strain Asibi, plus four nucleotides in the 3 noncoding region (Table 1). Eight of the 20 amino acid dierences are found in the E protein, suggesting that this protein is involved in the attenuation of wild-type strain Asibi (Figure 1). Comparison of the genome of wild-type FVV with vaccine strain FNV reveals 77 nucleotide dierences encoding 35 amino acid substitutions (Wang et al., 1995). Neither attenuation of Asibi nor FVV viruses involved nucleotide changes in the 5 noncoding region of the genome, suggesting that this region is not involved in the attenuation of virulence of wild-type YFV. Although Asibi and FVV were attenuated by very dierent routes (chicken and mouse tissue, respectively), both attenuation

ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net

Yellow Fever Virus

Table 1 Nucleotide and amino acid dierences between wild-type Asibi virus and attenuated 17D vaccines Amino Acid 36 52 170 173 200 299 305 380 407 307 118 167 172 183 109 485 146 95 836 900 Wildtype Asibi Leu Gly Ala Thr Lys Met Ser Thr Ala Ile Met Thr Thr Ser Ile Asp Val Ile Glu Pro U U G A 17D-204 and 17DD vaccines Phe Arg Val Ile Thr Ile Phe Arg Val Val Val Ala Ala Phe Leu Asn Ala Met Lys Leu C C A C

Nucleotide

Protein M E

and is thought to be due to antibody-dependent cellmediated cytotoxicity or complement-xing activity of anti-NS1 antibodies. The importance of yellow fever immunization is emphasized by the two fatal cases in US residents since 1996, both of whom travelled to South America without prior immunization (Centers for Disease Control and Prevention, 2000).

References
Calisher CH, Karabatsos N, Dalrymple JM et al. (1989) Antigenic relationships among aviviruses as determined by cross-neutralization tests with polyclonal antisera. Journal of General Virology 70: 3743. Centers for Disease Control and Prevention (2000) Fatal yellow fever in a traveller returning from Venezuela, 1999. Morbidity and Mortality Weekly Report 49: 303305. Chambers TJ, Hahn CS, Galler R and Rice CM (1990) Flavivirus genome organization, expression, and replication. Annual Reviews of Microbiology 44: 649688. Chang G-JJ, Cropp BC, Kiney RM, Trent DW and Gubler DJ (1995) Nucleotide sequence variation of the envelope protein gene identies two distinct genotypes of yellow fever virus. Journal of Virology 69: 57735780. Hahn CS, Dalrymple JM, Strauss JH and Rice CM (1987) Comparison of the virulent Asibi strain of yellow fever virus with the 17D vaccine strain derived from it. Proceedings of the National Academy of Sciences of the USA 84: 20192023. Jennings AD, Gibson CA, Miller BR et al. (1994) Analysis of a yellow fever virus isolated from a fatal case of vaccine-associated human encephalitis. Journal of Infectious Diseases 169: 513519. Kuno G, Chang G-JJ, Tsuchiya KR, Karabatsos N and Cropp CB (1998) Phylogeny of the genus Flavivirus. Journal of Virology 72: 73 83. Lepiniec L, Dalgarno L, Houng VTQ et al. (1994) Geographic distribution and evolution of yellow fever viruses based on direct sequencing of genomic cDNA fragments. Journal of General Virology 75: 417423. Rice CM, Lenches EM, Eddy SR et al. (1985) Nucleotide sequence of yellow fever virus: implications for avivirus gene expression and evolution. Science 229: 726733. US Department of Health and Human Services (1999) Biosafety in Microbiological and Biomedical Laboratories, 4th edn. Washington DC: US Government Printing Oce. Wang E, Ryman KD, Jennings AD et al. (1995) Comparison of the genomes of the wild-type French viscerotropic strain of yellow fever virus with its vaccine derivative French neurotropic vaccine. Journal of General Virology 76: 27492755. Wang E, Weaver SC, Shope RE et al. (1996) Genetic variation in yellow fever virus: Duplication in the 3 noncoding region of strains from Africa. Virology 225: 274281.

NS1 NS2A

NS2B NS3 NS4A NS4B

10367 10418 10800 10847

3NCR 3NCR 3NCR 3NCR

processes were found to share two common amino acid changes at membrane protein M amino acid 36 and NS4B protein amino acid 95. It is not known if these common changes were fortuitous or are involved in the attenuation of viscerotropism of wild-type YFV and/or loss of mosquito competence of the vaccine viruses. The molecular basis of reversion to virulence of the 17D vaccine has been examined. The 17D vaccine has caused only one fatal case of postvaccinal encephalitis, which was in a 3-year-old girl in 1965. The virus recovered from this patient, P-16065, was compared to the 17D-204 vaccine virus administered to the girl and found to dier at three amino acid positions, 155 and 303 in the E protein and amino acid 72 in the NS4B protein (Jennings et al., 1994). The contribution of one or more of these amino acid changes to reversion to virulence phenotype is not known. The mechanism of protective immunity is poorly understood, although production of neutralizing antibodies appears to correlate with immunity. Most neutralizing antibodies recognize epitopes of the E protein. Cellmediated immunity correlates with nonstructural proteins, in particular NS3. Experimental studies have demonstrated that NS1 can induce protective immunity. The mechanism does not involve neutralization of the virion

Further Reading
Augustin G (1909) History of Yellow Fever. New Orleans: Searcy and Pfa. Monath TP (ed.) (1988) Yellow fever. In: The Arboviruses: Ecology and Epidemiology, pp. 139231. Boca Raton: CRC Press. Monath TP (1991) Yellow fever: Victor, Victoria? Conqueror, conquest? Epidemics and research in the last forty years and prospects for the future. American Journal of Tropical Medicine and Hygiene 45: 143.

ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net

Yellow Fever Virus

Monath TP (1998) Yellow fever. In: Plotkin SA and Orenstein WA (eds) Vaccines, 3rd edn, pp. 815879. Philadelphia: WB Saunders. Monath TP and Heinz FX (1996) Flaviviruses. In: Fields BN, Knipe DM and Howley PM (eds) Fields Virology, 3rd edn, pp. 9611034. Philadelphia: Lippincott-Raven.

Rice CM (1996) Flaviviridae: the viruses and their replication. In: Fields BN, Knipe DM and Howley PM (eds) Fields Virology, 3rd edn, pp. 931960. Philadelphia: Lippincott-Raven. Tomari O (1999) Impact of yellow fever on the developing world. Advances in Virus Research 53: 534.

ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net