A Novel Model for the Study of the Therapy of Flavivirus Infections Using the Modoc Virus

Virology 279, 27–37 (2001) doi:10.1006/viro.2000.0723, available online at http://www.idealibrary.com on

CORE Metadata, citation and similar papers at core.ac.uk

Provided by Elsevier - Publisher Connector A Novel Model for the Study of the Therapy of Flavivirus Infections Using the Modoc Virus

Pieter Leyssen,* Alfons Van Lommel,† Christian Drosten,‡ Herbert Schmitz,‡ Erik De Clercq,* and Johan Neyts*,1

*Rega Institute for Medical Research, Katholieke Universiteit Leuven, Leuven B-3000, Belgium; †Division of Histopathology, University Hospitals, Leuven B-3000, Belgium; and ‡Bernhard Nocht Institute of Tropical Medicine, Hamburg, Germany Received May 31, 2000; returned to author for revision July 21, 2000; accepted October 26, 2000

The murine Flavivirus Modoc replicates well in Vero cells and appears to be as equally sensitive as both yellow fever and dengue fever virus to a selection of antiviral agents. Infection of SCID mice, by either the intracerebral, intraperitoneal, or intranasal route, results in 100% mortality. Immunocompetent mice and hamsters proved to be susceptible to the virus only when inoculated via the intranasal or intracerebral route. Animals ultimately die of (histologically proven) encephalitis with features similar to Flavivirus encephalitis in man. Viral RNA was detected in the brain, spleen, and salivary glands of infected SCID mice and the brain, lung, kidney, and salivary glands of infected hamsters. In SCID mice, the interferon inducer poly IC protected against Modoc virus-induced morbidity and mortality and this protection was associated with a reduction in infectious virus content and viral RNA load. Infected hamsters shed the virus in the urine. This allows daily monitoring of (inhibition of) viral replication, by means of a noninvasive method and in the same animal. The Modoc virus model appears attractive for the study of chemoprophylactic or chemotherapeutic strategies against Flavivirus infections. © 2001 Academic Press Key Words: Flavivirus; Modoc; antiviral therapy; West Nile virus; tick-borne encephalitis virus; Japanese encephalitis virus.

INTRODUCTION an estimated 30,000 people a year become infected, despite the availability of a vaccine. The mortality rate Flaviviruses are enveloped, positive single-stranded RNA viruses that belong, together with the Hepaci- and associated with JEV infections is 30% and an additional Pestiviruses, to the family of the Flaviviridae (Trepo et al., 30% of the survivors suffer from long-lasting neurological 1997; Rice, 1996). Based on phylogenetic analysis of 58 sequelae (Schneider et al., 1974; Kalita and Misra, 1998). members, the genus Flavivirus was subdivided into three There are two subtypes of tick-borne encephalitis virus clusters: the mosquito-borne, the tick-borne, and the (TBEV): the Far Eastern, also referred to as Russian no-vector clusters (Kuno et al., 1998). Viruses belonging spring-summer encephalitis virus (RSSEV), and the Eu- to the first two cluster have a complex natural transmis- ropean, referred to as central European encephalitis sion cycle, which involves several natural hosts (mostly virus (CEEV). TBEV is transmitted to humans usually by mammals or birds) and vector(s), the latter being blood- the bite of a tick or occasionally following consumption sucking insects such as ticks or mosquitoes (Monath of unpasteurized milk. The mortality rates associated and Heinz, 1996). with the Far Eastern subtype and the European subtype Flaviviruses have a worldwide distribution and several are about 30 and 1–2%, respectively (Heinz and Mandl, of them have a major impact on human health. The WHO 1993; Dumpis et al., 1999; Ecker et al., 1999). estimates that each year about 50 million people be- Although the last large epidemic of Murray Valley come infected with the dengue fever virus (DENV) (http:// encephalitis (MVEV) occurred in 1974, new cases of www.who.int/inf-fs/en/fact117.html). At least 500,000 MVEV are regularly reported, especially in Western Aus- cases of dengue hemorrhagic fever (DHF) or dengue tralia (Mackenzie and Broom, 1995). The West Nile virus shock syndrome (DSS) are reported annually, of which (WNV) is mainly endemic around the Mediterranean Sea, Ϯ5% may have a fatal outcome, mainly in children under Africa, and the Middle East. In 1996, an outbreak of WNV the age of 15 (Halstead, 1992). Despite the existence of occurred in Romania with 393 cases, of which 352 were an effective vaccine, the yellow fever virus (YFV) causes with acute CNS involvement and 17 with fatal outcome more than 5000 cases a year with a mortality of 50% (Tsai et al., 1998; Han et al., 1999). More recently (late (Monath, 1987). For the Japanese encephalitis virus (JEV), August–September 1999), an outbreak of WNV caused over 1900 serologically proven infections in the New York

1 metropolitan area, of whom 77 had to be hospitalized To whom correspondence and reprint requests should be ad- with signs of encephalitis and of whom 7 died (Briese et dressed at Rega Institute for Medical Research, Minderbroedersstraat 10, Leuven B-3000, Belgium. Fax: (32) 16-33.73.40. E-mail: al., 1999; Lanciotti et al., 1999). During the late summer of [email protected]. 2000, 17 more cases (1 death) were reported in New

York. From July to September 1999, an outbreak of me- and Alberta (Canada) (Zarnke and Yuill, 1985). Neutral- ningoencephalitis associated with WNV occurred in ization tests using blood samples isolated from mam- Southern Russia with hundreds of cases and dozens of mals trapped in Alberta as well as from humans indicate deaths (Lvov et al., 2000). In the late summer and early the appearance of natural infection without disease fall of 2000, an outbreak in Israel was responsible for (Zarnke and Yuill, 1985). No arthropod vector was previ- more than 169 cases and 12 deaths (Siegel, 2000). ously demonstrated (Johnson, 1967). Based on cross- Although there are no recent reports of outbreaks or serological reactivity, the virus was classified as a Flavi- epidemics of Saint Louis encephalitis virus (SLEV), the virus (Casals, 1960; Varelas and Calisher, 1982; Calisher virus causing this disease is endemic in the western et al., 1989). A phylogenetic analysis based on an ϳ1-kb region of the United States and is responsible for severe fragment of 58 flaviviruses showed that MODV belongs disease (Kramer et al., 1997). The Powassan encephalitis to the no-vector cluster of flaviviruses. The no-vector and virus (POWV), transmitted by ticks, is endemic in Ontario the vector-transmitted viruses have a common ancestor and in the Southern Far East of Russia and may cause (Kuno et al., 1998). encephalitis after a period of fever and nonspecific We determined the complete genomic sequence of the symptoms (Mandl et al., 1993; Kolski et al., 1998). Thou- Modoc virus (to be published elsewhere). An overall sands of cases of infections with the Kyasanur Forest sequence similarity with the tick-borne and mosquito- disease virus (KFDV, endemic in India) have been re- borne flaviviruses of 54.3 and 52.3%, respectively, was ported since the discovery of the virus in 1957. The calculated. This is comparable to the sequence similarity annual incidence of virologically diagnosed cases is between mosquito-borne (including YFV, DENV, JEV, 400–500, with a case fatality rate of 3–5% (Pavri, 1989; WNV, MVEV, KUNV) and tick-borne (TBEV, POWV, LIV) Banerjee, 1996; Monath and Heinz, 1996). A small num- Flaviviruses, which is 53.7%. Furthermore, based on our ber of cases of Omsk hemorrhagic fever (OHFV) occur sequence data, the Modoc virus has the same genomic each year among rural residents in the Omsk region in organization and conserved motifs for enzymatic activity Russia (Monath and Heinz, 1996). Louping ill virus (LIV) is as the human flaviviruses. In the present study, we es- primarly a disease in sheep, although there have been at tablished an infection model for MODV in SCID mice and least 37 documented cases in man. The infection was hamsters. This model may be attractive for the study of probably acquired during laboratory experiments or di- antiviral strategies against Flavivirus infections, particu- agnosis (Davidson et al., 1991; McGuire et al., 1998). larly Flavivirus encephalitis. Despite the clinical impact of Flavivirus infections, there is as yet no effective therapy. The study of antiviral RESULTS chemoprophylaxis or chemotherapy is hindered by the lack of a convenient small animal model. Antiviral susceptibility Experimental infection of mice with JEV, YFV, or TBEV MODV produces a clear cytopathic effect in Vero cells. by intracerebral or peripheral inoculation was reported to To study whether the replication of MODV is equally cause morbidity and mortality. However, use of these susceptible to a selection of antiviral agents as the highly pathogenic viruses (BSL-3 or BSL-4), for man, replication of human flaviviruses, we compared the in- requires special laboratory facilities (Holzmann et al., hibitory effects of ribavirin, EICAR (the 5Ј-ethynyl deriva- 1990; Higgs and Gould, 1991; Kaluzova et al., 1994). tive of ribavirin) (De Clercq et al., 1991), tiazofurin, selena- Attenuated strains of YFV or DENV (BSL-2) may cause zofurin (an oncolytic C-nucleoside), and mycophenolic morbidity and even mortality in small laboratory animals, acid (MPA, an inhibitor of IMP-dehydrogenase) on the but only when the virus is inoculated directly into the replication of YFV, DENV, and MODV. As can be derived brain of newborn or suckling mice (Kawano et al., 1993; from Table 1, the antiviral efficacy of the different com- Bray et al., 1998). Several models were previously de- pounds against YFV, DENV, and MODV ranked as fol- scribed for DENV infection in mice, although most of lows: mycophenolic acid Ͼ EICAR Ͼ selenazofurin Ͼ these are labor intensive because of the use of human ribavirin Ͼ tiazofurin. This result was confirmed by a tumor cell engraftment (Lin et al., 1998; An et al., 1999; cytopathic effect (CPE) reduction test (data not shown). Johnson and Roehrig, 1999). The susceptibility of MODV to these compounds proved We here present a novel and convenient model for the more or less comparable to the susceptibility of the study of new strategies for the treatment of infections human flaviviruses YFV and DENV to these drugs. with flaviviruses and in particular Flavivirus encephalitis. To this end, we employed the murine Flavivirus Modoc. Detection of Modoc virus by means of RT-PCR with The Modoc virus (MODV) was previously isolated from Flavivirus-specific primers white-footed deer mice (Peromyscus maniculatus) collected in Modoc County, California, in 1958 (Johnson, Total RNA was extracted from the supernatant of 1967). Later on the virus was also isolated from deer MODV-infected Vero cells that showed extensive CPE. mice trapped in the wild in Oregon, Montana, Colorado, This RNA was subjected to RT-PCR using a set of uni- MODOC VIRUS INFECTIONS 29

TABLE 1 organs. Because organ extracts cause some toxicity on Susceptibility of the Modoc Virus, YFV and DENV cell cultures in which the titration is performed, the de- to Different Antiviral Agents tection limit of this methodology appeared to be ϳ100 PFU/100 mg tissue. We therefore decided to monitor viral ␮ a EC50 ( g/ml) replication by means of RT-PCR [detection limit 10 fg of viral RNA (data not shown)]. Total RNA was extracted Antiviral agent Modoc virus YFV DENV from the brain, salivary glands, lung, liver, kidney, and Ribavirin 34 Ϯ 17 22 Ϯ 14 38 Ϯ 11 spleen of intraperitoneally infected SCID mice that ex- EICAR 1.0 Ϯ 0.6 0.8 Ϯ 0.4 1.6 Ϯ 0.6 hibited obvious signs of paralysis. Following RT-PCR, Mycophenolic acid 0.05 Ϯ 0.03 0.06 Ϯ 0.04 0.10 Ϯ 0.06 viral RNA was detected in the brain, salivary glands, and Ϯ Ϯ Ϯ Tiazofurin 48 18 17 97325 spleen. No, or very little, viral replication was detected in Selenazofurin 11 Ϯ 4 2.6 Ϯ 1.1 12 Ϯ 5

a Concentration required to reduce virus-induced plaque formation on Vero cells by 50%.

versal Flavivirus primers (Pierre et al., 1994). Amplicons of YFV had a size of 674 bp; RT-PCR on RNA extracted from DENV-2-infected cultures produced two bands, one of 641 bp and another one of 554 bp; and RT-PCR on RNA isolated from MODV-infected cultures yielded a fragment of 483 bp (data not shown). This latter fragment was sequenced. An overall sequence similarity of 57.9 and 65.5% with DENV and YFV, respectively, was calculated. The entire sequence of the Modoc virus genome will be published elsewhere. The reasons for the different lengths of the fragments are the variation of the sequence and the size of the 3Ј UTR of the viral genome where the primer VD8 binds (GenBank accession number DENV: M84727; YFV: U54798; MODV: our unpublished results).

Modoc virus infections in mice The effect of different routes of inoculation with MODV on morbidity and mortality in SCID and NMRI mice, respectively, was studied. Direct intracerebral inoculation of the virus (104 PFU), in both SCID and NMRI mice, led to paralysis within 6 days and mortality within 9 days [Figs. 1A and 1B; mean day of paralysis (MDP): 6.2 Ϯ 0.4 (SCID) and 6.3 Ϯ 0.4 (NMRI); mean day of death (MDD): 9.0 Ϯ 0.4 (SCID) and 9.4 Ϯ 0.5 (NMRI); 25 animals]. Morbidity was characterized by ruffled fur, paralysis of the hind legs, and a wasting syndrome. Intraperitoneal (i.p.) and intranasal (i.n.) infection of SCID mice with MODV resulted in morbidity [Fig. 1A; MDP: 10.8 Ϯ 0.6 (i.p. route, Ͼ50 animals), 11.8 Ϯ 0.4 (i.n. route, Ͼ15 animals)] and mortality [MDD: 12.8 Ϯ 1.3 (i.p. route), 13.4 Ϯ 0.5 (i.n. route)]. Immunocompetent NMRI mice, infected via intraperitoneal inoculation, remained healthy throughout the experiment. When NMRI mice were inoculated via the intranasal route, 55% of the animals developed paralysis [MDP: 11.6 Ϯ 0.5 (11 animals)] and ultimately died Ϯ as a result of CNS involvement (MDD: 12.8 0.8) (Fig. FIG. 1. Paralysis (Ⅺ) and mortality (f) following intracerebral (i.c.), 1B). intraperitoneal (i.p.), or intranasal (i.n.) inoculation of SCID mice (A), We next wanted to monitor viral replication in different NMRI mice (B), and hamsters (C) with MODV. NT ϭ not tested. 30 LEYSSEN ET AL.

creased diuresis, resulting in an important dilution of viral RNA (before Day 9, average urine volumes were 2.1 Ϯ 0.8 ml and this increased to Ͼ50 ml urine/hamster/ day after Day 9 postinfection).

Histopathology The brains of hamsters that had been infected with the Modoc virus via intraperitoneal inoculation and that exhibited obvious signs of encephalitis were studied histologically. In parts of the cortex and bulbus olfactorius, FIG. 2. Detection of viral RNA in different organs of animals infected obvious destruction of the tissue was observed. In these with MODV via the intraperitoneal route. Organs were collected at the areas, infiltration of monocytes and lymphocytes was time the animals showed fulminant encephalitis. RNA was detected by noted. Infiltration with inflammatory cells was merely means of RT-PCR. Lane 1, brain; lane 2, salivary gland; lane 3, lung; lane 4, liver; lane 5, kidney; and lane 6, spleen. (A) SCID mice, 10 days observed in the gray matter of the cortex (Fig. 4B) and in postinfection. (B) Hamsters, 11 days postinfection. the bulbus olfactorius (Fig. 4C). Perivascular cuffing, one of the characteristics of encephalitis induced by flaviviruses, was observed in small or medium-size blood the lungs, liver, and kidneys (Fig. 2A). Next, the progres- vessels (Figs. 4D and 4E). The meninges were loosely sion of infection was monitored in the brain and spleen attached and swollen in the affected areas (Fig. 4B; of SCID mice that had been infected intraperitoneally compare with Fig. 4A in a nonaffected area). The struc- with MODV. Two mice were terminated daily (from Day 1 ture of the tissue was (almost) completely destroyed in to Day 10 postinfection) and total RNA was extracted and these areas (Fig. 4B). Around these acellular plaques subjected to RT-PCR. Viral RNA first became detectable (endematous necrotic foci), degenerated or pyknotic in the spleen at 5 days postinfection. In the brain, viral (deeply eosinophilic) neurons were observed, often in- RNA became detectable at 9 days postinfection. Signs of volved in neuronophagia or the formation of glial nod- paralysis developed about 2 days later (i.e., on Day 11 ules. Throughout the gray matter, groups of degenerating postinfection; data not shown). neurons were detected (Fig. 4F). In the cerebellum, only a small number of Purkinje neurons appeared to be Modoc virus infection in hamsters involved (Fig. 4G). Except for the olfactory bulbs, part of the cortex and the cerebellum, no other major signs of Although immunocompetent, 50–60% of 8- to 12-week- virus-induced pathology were noted in the brain. In the old hamsters that had been inoculated intraperitoneally spleen, large polymorphonuclear cells were detected in or intranasally with MODV became sick, developed signs of paralysis, and ultimately died because of encephalitis [intraperitoneal inoculation: MDP ϭ 9.2 Ϯ 0.8; MDD ϭ 11.9 Ϯ 1.3 (18 animals); intranasal inoculation: MPD ϭ 6.8 Ϯ 0.4; MDD ϭ 13.7 Ϯ 2.9 (7 animals)] (Fig. 1C). Total RNA extracted from the organs of three moribund animals was subjected to RT-PCR. Viral RNA was detected in the brain, salivary glands, lungs, and kidneys (Fig. 2B). Because viral RNA was also detected in the kidneys, in a next set of experiments, we collected urine samples during the course of the infection. RNA was extracted from these samples and subjected to RT-PCR. From Day 5 postinfection, viral RNA became detectable in the urine after which the signal intensified for several days, sug- gesting increasing viral titers (Fig. 3A). Detection of viral RNA in the urine is consistent with the detection of the FIG. 3. MODV RNA from Day 1 to Day 12 postinfection in the urine of virus in the kidneys. Using TaqMan technology for quan- 8- to 12-week-old hamsters (at each time point, the urine from four titative RT-PCR, a very low viral RNA load was detectable animals was collected). (A) RT-PCR analysis of viral RNA extracted from as of Days 3 and 4 postinfection, whereafter viral RNA 140 ␮l urine. (B) Quantitation of the viral RNA load in the urine by titers (quantified as PCR-units ϭ lowest template copy TaqMan RT-PCR (means Ϯ SD of two replicate experiments). PCRU ϭ number detectable in five of five replicate reactions de- PCR-units. One PCR-unit is defined to be the lowest template copy number detectable in five of five replicate reactions as determined by termined by limiting dilution series) sharply increased up limiting dilution series. Values above bars indicate the mean values. to Day 8 postinfection (Fig. 3B). From Day 9 on, viral titers Arrow indicates the day as of which an increased diuresis is observed started to drop. This is consistent with a markedly in- in infected animals. MODOC VIRUS INFECTIONS 31

Ultrastructural analysis MODV-infected Vero cell cultures were studied by means of electron microscopy. Upon infection, the intra- cellular organization appeared to be changed. Cellular pathology was characterized, as compared to the cytoplasm of a noninfected cell, by extensive proliferation and dilatation of the rough endoplasmic reticulum (RER) (Fig. 5A and insert), which is a characteristic for cells infected with a Flavivirus (Westaway, 1987; Chambers et al., 1990). Large amounts of viral particles egressed from

FIG. 4. Histological appearance (H&E) of the brain of MODV-infected hamsters and SCID mice at 10 days postinfection. (A) Cross section of the cerebral cortex in a nonaffected area of hamster brain (ϫ100). (B) Cross section of the cerebral cortex (same hamster as in A) in an affected area (acellular plaque), showing massive infiltration by inflammatory cells, loosely attached and swollen meninges, and destruction of the normal tissue structure (ϫ100). (C) Infiltration of the bulbus olfactorius of a MODV-infected hamster by inflammatory cells and destruction of the normal tissue structure (ϫ100). (D) Cross section through medium-size blood vessels in the gray matter of the cortex of a MODV-infected hamster at 15 days postinfection, showing perivascular cuffing (ϫ100). (E) Longitudinal section of a medium-size blood vessel in the gray matter of a MODV-infected hamster at 15 days postinfection, showing perivascular cuffing (ϫ100). (F) Degenerated or pyknotic (deeply eosinophilic) neurons accompanied by normal neurons in the cerebral cortex of a MODV-infected SCID mouse at 11 days postinfection (ϫ180). (G) One normal and two degenerating Purkinje cells in the cerebellum of a MODV-infected SCID mouse at 11 days postinfection (ϫ220). (H) Large polymorphonuclear cell in the red pulpa of the spleen of a MODV-infected hamster at 15 days postinfection (ϫ220). (I) Steatosis of the liver of a MODV-infected hamster at 11 days postinfection (ϫ180). the red pulpa (Fig. 4H). Whether this is a direct conse- quence of the replication of the virus or the result of the immune response remains unclear. In the liver, extensive steatosis was observed (Fig. 4I). No apparent his- FIG. 5. Ultrastructural analysis of (A) Vero cell 2 days postinfection, topathological changes were observed in the other or- showing dilated RER (scale bar ϭ 0.5 ␮m) compared to the RER of a gans studied. noninfected cell (arrows, insert); (B) Vero cell 3 days postinfection, In SCID mice that exhibited obvious signs of Modoc releasing large amounts of mature viral particles (scale bar ϭ 0.5 ␮m); encephalitis, however, little or no infiltration of the brain (C) affected Purkinje cell surrounded by intact tissue in the cerebellum ϭ with lymphocytes was observed (data not shown). This is of a MODV-infected SCID mouse at 11 days postinfection (scale bar 5 ␮m); (D) neuron (bottom right) located in an acellular plaque of the in accordance with the fact that SCID mice do not have cerebral cortex (tissue fragments) accompanied by a blood lymphocyte functional T and B cells. Also, despite the fact that these (top left) and a mononuclear cell (bottom left) of a MODV-infected mice had clinical signs of encephalitis, no other major hamster 15 days postinfection (scale bar ϭ 5 ␮m); (E) cytoplasm of a histopathological changes were noted. However, at the normal neuron of a SCID mouse (RER, top left; Golgi complex, bottom ϭ ␮ ultrastructural level, severe changes in the neurons were right; scale bar 1 m); and (F) cytoplasm of an infected neuron of a MODV-infected SCID mouse at 11 days postinfection, showing exten- observed (see below). This is in agreement with neuron sive dilatation, proliferation, and hypertrophy of the RER and vesicle dysfunction, which leads to encephalitis and ultimately structures containing virus-like particles (center right; scale bar ϭ death. 1 ␮m). 32 LEYSSEN ET AL.

TABLE 2 means of RT-PCR) in the brain and spleen of MODV Effect of Pre- or Posttreatment with the Interferon-Inducer Poly IC infected SCID mice that had received either mock- or on MODV-Induced Morbidity and Mortality in Intraperitoneally In- poly IC treatment. Tissue samples were taken at a time fected SCID Mice at which untreated controls develop signs of paralysis (i.e., 11 days postinfection). Levels of viral RNA detected Number in the brain and spleen of infected SCID mice were with Number Condition paralysis MDP of death MDD

VC 5/5 12.8 Ϯ 1.9 5/5 15.4 Ϯ 1.9 Pretreatment 9/15*** 15.1 Ϯ 2.0* 13/15NS 23.0 Ϯ 7.2** Posttreatment 14/14 13.1 Ϯ 1.8NS 14/14 16.4 Ϯ 2.6NS

* P Ͻ 0.05, **P Ͻ 0.01, ***P Ͻ 0.001, NS ϭ not significant. Note. Adult SCID mice were either left untreated or treated intraperitoneally at Day Ϫ1 with 15 mg/kg poly IC (pretreatment condition) or at Day ϩ1 with 15 mg/kg poly IC (posttreatment condition). the cells 2–3 days after infection (Fig. 5B). The size of the viral particles was determined to be about 45 nm, which is very similar to the size of human flaviviruses (40–60 nm) (Monath and Heinz, 1996). Next, brain tissue samples from mice and hamsters that showed extensive signs of encephalitis were examined for ultrastructural changes. In the cerebellum of SCID mice and hamsters, Purkinje cells show dilatation and hypertrophy of the RER (Fig. 5C). In the cytoplasm of degenerative cortical neurons from SCID mice and hamsters, accumulation of vesicular structures, most likely derived from the ER or Golgi complex, was observed. Furthermore, dilatation, proliferation, and hypertrophy of the RER were noted (Fig. 5F), which is not the case in neurons of uninfected animals (Fig. 5E). In the acellular plaques observed in the brain of infected hamsters, mononuclear cells (mac- rophages or microglial cells) and infiltrating blood lymphocytes were distinguished (Fig. 5D).

Effect of treatment with an interferon inducer on Modoc virus replication and associated mortality To validate the Modoc virus model for future use in antiviral studies, we sought to assess whether a correlation exists between the protective effect on morbidity and mortality of a therapeutic or prophylactic agent, on one hand, and an inhibitory effect on viral replication, on the other hand. Treatment with ribavirin did not protect mice from MODV-induced morbidity and mortality, nor did it cause a reduction in viral RNA in the infected organs (data not shown). We therefore studied the effect of the interferon inducer poly IC. To this end, SCID mice were treated with a single dose of poly IC at 15 mg/kg FIG. 6. MODV RNA detected in the spleen and brain of MODV- and 24 h later were infected intraperitoneally with MODV. infected SCID mice that received pretreatment or posttreatment with Under these conditions, treatment with poly IC resulted poly IC (experiment parallel to the experiment presented in Table 2). (A) in significantly fewer animals that developed signs of Gel electrophoresis of the amplicons obtained by OneStep RT-PCR. (B) paralysis and also caused a 6- to 7-day delay in virus- Semiquantitation of the viral load using GAPDH as an internal standard. (C) Titration of 10% tissue homogenates on Vero cells for infectious induced mortality. When poly IC was first given 1 day virus load (dotted line indicates threshold of detection). (D) Quantitation after the infection, all protective activity was lost (Table of viral load using TaqMan technology (each bar represents the 2). In a parallel experiment, viral RNA was monitored (by means Ϯ SD of four replicate experiments). MODOC VIRUS INFECTIONS 33 markedly reduced in those animals that had been pre- tick-borne encephalitis virus). Since these enzymatic ac- treated with poly IC as was assessed by means of tivities imply specific catalytic residues and configura- semiquantitative RT-PCR (Figs. 6A and 6B), by quantita- tions of amino acids in the active site, antiviral agents tive RT-PCR (TaqMan technology) (Fig. 6D) and by titra- that may be active against one Flavivirus may also be tion for infectious virus content (Fig. 6C). Thus, a protec- expected to show broad-spectrum activity against other tive effect of poly IC on virus-induced morbidity and flaviviruses (as we demonstrate in cell culture experi- mortality translates in a reduction of titers of infectious ments; see Table 1). Only those drugs that show broad- virus and viral RNA. spectrum anti-Flavivirus activity would be of interest for further development. Such drugs could then be easily DISCUSSION evaluated in the Modoc models, which may be instru- mental in the preclinical development of anti-Flavivirus The availability of a convenient small animal model drugs. would be very useful for the study of strategies for the The Modoc virus particle has a size of ϳ45 nm, which treatment or prevention of Flavivirus infections. Monkeys is comparable to the size of other flaviviruses (40–60 nm) infected with flaviviruses such as YFV, DENV, or JEV may (Monath and Heinz, 1996). Ultrastructurally, infected cells develop infection, but because of the cost involved and show changes (proliferation of the endoplasmic reticu- the restricted availability of monkeys, the number of lum) similar to those observed in cells infected with other studies that can be carried out by using these animals is flaviviruses (Westaway, 1987; Chambers et al., 1990). limited. Some flaviviruses (e.g., DENV, YFV) cause mor- In immunocompetent mice (here NMRI), MODV bidity and mortality in mice, but only when inoculated causes morbidity and mortality, which is 100% when virus intracerebrally and when virus strains are used that have is inoculated directly into the brain and 55% when inoc- been adapted by serial passage in the brain of suckling ulated via the intranasal route. However, MODV appears mice. Other flaviviruses (e.g., TBEV, JEV) may cause in- to be highly pathogenic to SCID mice when inoculated fection in mice following peripheral inoculation, although peripherally (i.e., via either the intraperitoneal or the these viruses require high containment laboratory facil- intranasal route, 100% mortality). Viral replication is de- ities, which is of course an obstacle for intensive re- tected in the salivary glands and spleen of infected SCID search with animals. The Modoc virus was first isolated mice and, in a later stage, the virus becomes detectable from deer mice in Colorado in 1958 (Johnson, 1967) and, in the brain, followed a couple of days later by death as except for some initial reports, no further research has a result of CNS involvement. Although no major his- been carried out with this virus. Based on cross-serology, topathological changes were observed in the brains of the Modoc virus was classified as a Flavivirus (Casals, SCID mice that had obvious signs of Modoc virus-in- 1960; Varelas and Calisher, 1982; Calisher et al., 1989). In duced encephalitis, ultrastructural analysis of the neu- a comparative study, Kuno and colleagues (1998) ampli- rons clearly revealed a cellular pathology characteristic fied an ϳ1-kb fragment of the genome of 58 different for Flavivirus infection. flaviviruses, including MODV, using a primer set that Although immunocompetent mice are not susceptible anneals to the highly conserved 3Ј region of the Flavivi- to the virus when infected via the intraperitoneal route, rus genome. Phylogenetic analysis based on these se- 50–60% of the inoculated hamsters became sick and quences revealed that MODV belongs to the no-vector died (as a result of CNS involvement) after intranasal or cluster, which further branches into three clades, all of intraperitoneal inoculation with the virus. MODV-induced which are murine or bat-associated viruses (Kuno et al., encephalitis in hamsters was characterized by infiltration 1998). of the cortex and bulbus olfactorius by monocytes and The Modoc virus can be readily propagated in Vero lymphocytes and massive destruction of the tissue struc- cells and elicits an obvious cytopathic effect. MODV virus ture. Large polymorphonuclear cells could be detected in replication appeared to be more or less as equally sen- the spleen and extensive steatosis was observed in the sitive as YFV and DENV to the inhibitory activity of dif- liver. Interestingly, the virus was shed in the urine of ferent drugs that inhibit Flavivirus replication, including these infected hamsters (and not SCID mice). This is in ribavirin, EICAR (the 5-ethynyl analog of ribavirin), and agreement with detection of viral RNA in the kidneys of mycophenolic acid (MPA). The overall genome organiza- infected hamsters. That the virus is shed in the urine of tion of MODV is similar to that of other flaviviruses (our infected hamsters offers the possibility to assess the unpublished results) and genes that encode potential effect of therapy on viral replication in one and the same antiviral targets such as the NS3 (encoding the protease, animal and over a relatively long time span, simply by NTPase, and helicase) and NS5 (encoding the methyl- (daily) monitoring of viral RNA in the urine (by means of transferase and RNA-dependent RNA polymerase) con- the quantitative RT-PCR assay we have established). tain the same conserved motifs for enzymatic activity as To validate the SCID model for use in antiviral studies, those of flaviviruses that are important human pathogens we sought to prove that protection against MODV dis- (e.g., dengue fever virus, Japanese encephalitis virus, ease progression, brought about by either a therapeutic 34 LEYSSEN ET AL. or a prophylactic agent, correlates with a reduction in Compounds viral RNA. Although ribavirin did inhibit the replication of Mycophenolic acid (MPA) was purchased from Sigma YFV, DENV, and MODV in vitro, the drug was not potent (St. Louis, MO) and ribavirin [1-(␤-D-ribofuranosyl)-1,2,4- enough to inhibit the replication of the virus in infected triazole-3-carboxamide] was from ICN (Costa Mesa, CA). SCID mice and hamsters (i.e., neither a delay in virus- EICAR (5-ethynyl-1-␤-D-ribofuranosylimidazole-4-carbox- induced morbidity and mortality nor a reduction in viral amide) was synthesized by Dr. A. Matsuda (Hokkaido RNA was observed). This is in agreement with the lack of University, Sapporo, Japan). Tiazofurin was provided by activity of ribavirin against experimental infections with Dr. R. Cooney and Dr. D. G. Johns (National Cancer DENV or YFV in monkeys (Huggins, 1989; Malinoski et Institute, NIH, Bethesda, MD). Poly inosinic-polycytidilic al., 1990). Because of the absence of any activity of acid (Poly IC) was purchased from Sigma. ribavirin in the Modoc model, we decided to study whether the interferon inducer poly IC delays or prevents Plaque-reduction assay virus-induced mortality and, if so, whether this correlates with a reduction in viral RNA load. Poly IC was previously Serial dilutions of the test compounds were added to shown to elicit a protective effect against JEV in mice and confluent Vero cell cultures that were grown in 96-well monkeys (Singh and Postic, 1970; Worthington et al., microtiter trays. Cells were infected with approximately 1973; Harrington et al., 1977). Pretreatment of infected 20 PFU (plaque-forming units) of either YFV, DENV, or SCID mice with poly IC resulted in a significant delay in MODV. Cultures were further incubated at 37°C for 7 virus-induced mortality, which correlated with a marked days, after which they were fixed with 70% ethanol and reduction in viral RNA load in infected organs. These stained with a 2% Giemsa solution. Plaques were then findings demonstrate that the Modoc virus model may be counted under an inverted microscope. The antiviral ac- convenient for antiviral drug study purposes. tivity of the compounds is expressed as the 50% effective In conclusion, the Modoc models in SCID mice and concentration (EC50) or the concentration required to hamsters may offer the possibility of evaluating novel reduce plaque formation by 50%. strategies for the treatment of infections with flaviviruses, including Flavivirus-induced encephalitis. Titration for infectious virus content Animals were killed by ether anesthesia, spleen and MATERIALS AND METHODS brain were dissected aseptically, and 10% (w/v) tissue homogenates were prepared in MEM supplemented Cells and viruses with 2% FCS. Confluent Vero cell cultures grown in 96- well trays were infected with 10-fold serial dilutions of African green monkey kidney (Vero) cells were grown the tissue homogenates and incubated at 37°C for 1 h, in minimum essential medium (MEM; Gibco, Paisley, after which the inoculum was removed. Cultures were Scotland) supplemented with 10% inactivated fetal calf then washed twice with warm medium and further incu- serum (FCS; Integro, Zaandam, Holland), 1% L-glutamine, bated at 37°C for 7–9 days, after which the cultures were and 0.3% bicarbonate. MODV (ATCC VR-537) was ob- evaluated for virus-induced cytopathic effect and titer tained from the American Type Culture Collection (ATCC, was expressed in cell culture infective dose 50 (CCID ). Rockville, MD). The cytopathic dengue virus serotype 2 50 (strain New Guinea) was kindly provided by Dr. V. Deubel RT-PCR (Institut Pasteur, Paris, France) and the yellow fever virus used was the vaccine strain Stamaril (Pasteur Merieux). Viral RNA was extracted from 140 ␮l of either cell All viruses were propagated in Vero cells at 37°C. Con- culture supernatant or urine of infected animals using fluent cultures of Vero cells were infected with the dif- the QIAamp Viral RNA kit (QIAgen, Hilten, Germany). For ferent viruses and incubated at 37°C until extensive CPE the isolation of RNA from cell cultures, cell debris was was observed (generally 7–9 days postinfection). At that first spun down with two consecutive centrifugation time, culture medium was collected, cell debris was steps at 3500 rpm in a Minifuge-T. For the isolation of pelleted by centrifugation, and the supernatant was ali- RNA from tissue samples, the RNeasy Mini kit (Qiagen) quoted and stored at Ϫ80°C. was used. The following reaction mixture was prepared for the reverse transcription step: 31.5 ␮l purified RNA; 10 ␮lof5ϫ RT-buffer (Amersham, Roosendaal, Belgium); Animals dATP, dGTP, dCTP, and dTTP, each at a final concentra- Eight- to 12-week-old Gold hamsters and SCID (severe tion of 0.5 mM; and 1.2 ␮M VD8 primer (5Ј-GGGTCTC- combined immunodeficiency) or NMRI mice (weighing CTCTAACCTCTAG-3Ј). Following a 1-min incubation of 16–20 g) were used throughout the experiments. All an- the RNA template at 94°C, annealing of the primer was imals were bred at the Rega Institute (the SCID mice allowed to proceed at room temperature. Subsequently, under germ-free conditions). 2.5 ␮l enzyme mixture was added, consisting of 1.35 ␮l MODOC VIRUS INFECTIONS 35

RNase-free water, 95 units HPRI ribonuclease inhibitor BSA (Sigma-Aldrich, Munich, Germany), and 0.4 ␮lof (Amersham), and 3.75 units RAV-2 reverse transcriptase ThermoScript Reverse Transcriptase/Platinum Taq En- (Amersham). The mixture was incubated at 45°C for zyme mix. Prepared RNA (2 ␮l) was added to each 1.5 h. The PCR reaction mixture was prepared as follows: reaction. Thermal cycling in a LightCycler (Roche Molec- 35.3 ␮l DNase-free water, 5 ␮l cDNA, 5 ␮l10ϫ PCR- ular Biochemicals, Mannheim, Germany) involved re- buffer, 2 ␮l dNTP-mix (100 ␮M of each dNTP), 1.2 ␮M verse transcription at 55°C for 20 min, denaturation at primer VD8, 1.2 ␮M primer EMF1 [5Ј-TGGATGACSACK- 95°C for 5 min, and 40 cycles of 95°C for 5 s and 60°C GARGAYATG-3Ј (S ϭ C/G, K ϭ T/G, R ϭ A/G, Y ϭ C/T)] for 20 s (Wittwer et al., 1997). were mixed with 3.5 units Super TAQ (HT Biotechnology, In the course of PCR, probe ModP1 was digested by Cambridge, UK). The following PCR program was run 5Ј-nuclease activity of Taq DNA polymerase when spe- (Thermocycler Techne; Progene, Cambridge, UK): 5 min cifically annealed to a generated Modoc PCR-product at 95°C, 30 cycles of 45 s at 95°C, 45 s at 50°C, 45 s at (Holland et al., 1991). Probe-digestion liberated the FAM 72°C, and for the final extension 5 min at 72°C. EMF1 from the TAMRA dye, causing an increase in FAM-spe- and VD8 are universal Flavivirus primers that anneal in cific fluorescence during PCR (Livak et al., 1995). FAM very conserved regions of the NS5 protein and the 3Ј fluorescence was measured on detection channel F1 UTR, respectively (Pierre et al., 1994). (530 nm) and divided by fluorescence measured on channel F2 (640 nm) for normalization. As a quantifica- Semiquantitative RT-PCR tion standard for real-time PCR, a strong positive Modoc virus stock was diluted in negative human plasma prior A calculated amount of 15 ng of total RNA isolated to RNA preparation. The amount of one PCR-unit of from tissue fragments was used in a semiquantitative Modoc RNA was defined to be the lowest possible dilu- RT-PCR assay using the OneStep RT-PCR Kit (Qiagen). tion step that could be amplified in five of five replicate GAPDH served as an internal control and a primer ratio reactions. The quantification procedure using the Light- MODV/GAPDH of 2/1 was used, as recommended by the Cycler was previously described (De Silva et al., 1998). manufacturer. Amplification primers were GAPDH for- Urine samples were analyzed in two replicate reactions ward primer: 5Ј-GGT GAA GGT CGG TGT GAA CG-3Ј; and tissue samples, in four. GAPDH reversed primer: 5Ј-ATG TTC TGG ACA ACC CCA CG-3Ј; MODV forward primer: 5Ј-GGG TGC TAA Animal experiments CTT GGA TGT GGA-3Ј; MODV reversed primer: 5Ј-GGG CTG TTG GCA TTG ATT CTT-3Ј. After amplification (an- SCID mice or NMRI mice were inoculated with 104 nealing at 64°C, 30 cycles) and separation of the ampli- PFU of MODV via either the intraperitoneal (i.p., 200 ␮l), cons (GAPDH: 609 b, MODV 743 b) by agarose gel intracerebral (i.c., 50 ␮l), or intranasal route (i.n., 20 ␮l). electrophoresis, the gel was scanned and the two bands The animals were examined daily for signs of morbidity were quantified using the ImageMaster program (Hoeffer and mortality. Statistical significance of differences in the Pharmacia Biotech, Uppsala, Sweden). mean day of death was assessed by means of the Student’s t test and differences in the number of survivals Quantification of Modoc virus RNA by 5Ј-nuclease by means of the ␹2 test. To study the progression of real-time RT-PCR MODV infection in SCID mice, the animals were infected intraperitoneally with MODV and two mice were termi- RNA preparation was performed using the Qiagen nated daily. Brain, salivary glands, lung, liver, kidney, and Viral RNA Kit (Qiagen) according to the manufacturer’s spleen were removed and total RNA was extracted from instructions. For elution of RNA, the columns were incu- these organs and subsequently used for RT-PCR. Eight- bated with 50 ␮l of RNase-free water at 80°C. Primers to 12-week-old hamsters were infected intraperitoneally ModS1 (5Ј-CCA GGA CAA GTC ATG TGG TAG C-Ј3) and with 104 PFU of MODV. At least three animals were used ModAS1 (5Ј-TCC CAA AGA TGT TCC TCA CCT T-3Ј) were for each experimental condition. The animals were used to amplify a 101-bp segment in the NS5 region of housed individually in metabolic cages. Urine samples Modoc virus. For detection of PCR products, probe were collected daily and RNA was extracted and subse- ModP1 (5Ј-CAG AAT GGG CCA AGT TGC TTC CAG T-3Ј) quently used for RT-PCR. was used. It was labeled with FAM (6-carboxy-fluores- cein) at the 5Ј-end and TAMRA (6-carboxy-N,N,NЈ/,NЈ- Histology, histopathology, and electron microscopy tetramethylrhodamine) at the 3Ј-end. The probe was phosphorylated at its 3Ј-end to prevent elongation during Vero cells that had been infected with MODV and that PCR. RT-PCR was done using the Platinum RT-PCR Ther- exhibited an extensive cytopathic effect were fixed in a moScript One-Step System (Life Technologies, buffered glutaraldehyde solution for 30 min, washed with ␮ Karlsruhe, Germany). A 20- l reaction volume contained isotonic phosphate buffer, postfixed with a 1% OsO4 10 ␮lof2ϫ reaction buffer, 280 nM each of primers solution for 1 h, collected, dehydrated, and embedded in ModS1 and ModAS1, 140 nM of probe ModP1, 0.8 ␮g Dow epoxy resin for electron microscopy. Sections (1 36 LEYSSEN ET AL.

␮m) stained with toluidin blue were examined with a light De Silva, D., Reiser, A., Hermann, M., Tabiti, K., and Wittwer, C. (1998). microscope. Sample sections (50 nm) of areas of interest Rapid genotyping and quantification on the LightCycler with hybrid- were collected on grids, stained with uranyl acetate and ization probes. Biochemica 2, 12–15. Dumpis, U., Crook, D., and Oksi, J. (1999). Tick-borne encephalitis. Clin. lead citrate, and examined with a Philips CM10 electron Infect. Dis. 28, 882–890. microscope. Moribund animals with severe signs of pa- Ecker, M., Allison, S. L., Meixner, T., and Heinz, F. X. (1999). Sequence ralysis were terminated by ether anesthesia and were analysis and genetic classification of tick-borne encephalitis viruses transcardially perfused with 20 ml of a buffered 4% form- from Europe and Asia. J. Gen. Virol. 80, 179–185. aldehyde solution for histology or with 20 ml of buffered Ghosh, S. N., Goverdhan, M. K., Sathe, P. S., Chelliah, S. C., Naik, S. V., Godbole, P. V., and Banerjee, K. (1990). Protective effect of 6-MFA, a 2.5% glutaraldehyde solution for electron microscopical fungal interferon inducer against Japanese encephalitis virus in bon- purposes. Organs were processed as described above. net macaques. Indian J. Med. Res. 91, 408–413. For histology, fixed tissue samples were embedded in Goverdhan, M. K., Kulkarni, A. B., Gupta, A. K., Tupe, C. D., and Ro- paraffin and sections were stained with hematoxylin and drigues, J. J. (1992). Two-way cross-protection between West Nile and eosin (H&E) according to standard procedures. Japanese encephalitis viruses in bonnet macaques. Acta Virol. 36, 277–283. Halstead, S. B. (1992). The XXth century dengue pandemic: Need for ACKNOWLEDGMENTS surveillance and research. World Health Stat. Q. 45, 292–298. Han, L. L., Popovici, F., Alexander, J. J., Laurentia, V., Tengelsen, L. A., Pieter Leyssen is a Research Fellow from the Instituut voor Weten- Cernescu, C., Gary, J. H., Ion, N. N., Campbell, G. L., and Tsai, T. F. schappelijk Onderzoek & Technologie (IWT) and Johan Neyts is a (1999). Risk factors for West Nile virus infection and meningoenceph- postdoctoral research fellow of the Fonds voor Wetenschappelijk alitis, Romania, 1996. J. Infect. Dis. 179, 230–233. Onderzoek-Vlaanderen (FWO). We thank C. Callebaut and I. Aerts for Harrington, D. G., Hilmas, D. E., Elwell, M. R., Whitmire, R. E., and excellent editorial assistance. This work was supported by a grant from Stephen, E. L. (1977). Intranasal infection of monkeys with Japanese the Flemisch Institute for the stimulation of scientific-technological encephalitis virus: Clinical response and treatment with a nuclease- research in the industry (IWT/SB/981025/Leyssen), Geconcerteerde resistant derivative of poly (I):poly (C). Am. J. Trop. Med. Hyg. 26, Onderzoeksacties (GOA: Project No. 95/5), and the Fonds voor Weten- 1191–1198. schappelijk Onderzoek (G. 0122-00). Heinz, F. X., and Mandl, C. W. (1993). The molecular biology of tick- borne encephalitis virus. APMIS 101, 735–745. REFERENCES Higgs, S., and Gould, E. A. (1991). Differences in fusogenicity and mouse neurovirulence of Japanese encephalitis viruses. Arch. Virol. An, J., Kimura, K. J., Hirabayashi, Y., and Yasui, K. (1999). Development 119, 119–133. of a novel mouse model for dengue virus infection. Virology 263, Holland, P. M., Abramson, R. D., Watson, R., and Gelfand, D. H. (1991). 70–77. Detection of specific polymerase chain reaction product by utilizing Angsubhakorn, S., Moe, J. B., Marchette, N. J., Latendresse, J. R., the 5Ј 3 3Ј exonuclease activity of Thermus aquaticus DNA poly- Palumbo, N. E., Yoksan, S., and Bhamarapravati, N. (1987). Neuro- merase. Proc. Natl. Acad. Sci. USA 88, 7276–7280. virulence detection of dengue virus using rhesus and cynomolgus Holzmann, H., Heinz, F. X., Mandl, C. W., Guirakhoo, F., and Kunz, C. monkeys. J. Virol. Methods 18, 13–24. (1990). A single amino acid substitution in envelope protein E of Arroyo, J. I., Apperson, S. A., Cropp, C. B., Marafino, B. J., Jr., Monath, tick-borne encephalitis virus leads to attenuation in the mouse T. P., Tesh, R. B., Shope, R. E., and Garcia, B. M. (1988). Effect of model. J. Virol. 64, 5156–5159. human gamma interferon on yellow fever virus infection. Am. J. Trop. Huggins, J. W. (1989). Prospects for treatment of viral hemorrhagic Med. Hyg. 38, 647–650. fevers with ribavirin, a broad-spectrum antiviral drug. Rev. Infect. Dis. Banerjee, K. (1996). Emerging viral infections with special reference to 11(Suppl. 4), S750–S761. India. Indian J. Med. Res. 103, 177–200. Johnson, A. J., and Roehrig, J. T. (1999). New mouse model for dengue Bray, M., Men, R., Tokimatsu, I., and Lai, C. J. (1998). Genetic determi- virus vaccine testing. J. Virol. 73, 783–786. nants responsible for acquisition of dengue type 2 virus mouse Johnson, H. N. (1967). Ecological implications of antigenically related neurovirulence. J. Virol. 72, 1647–1651. mammalian viruses for which arthropod vectors are unknown and Briese, T., Jia, X.-Y., Huang, C., Grady, L. J., and Lipkin, W. I. (1999). avian associated soft tick viruses. Jpn. J. Med. Sci. Biol. 20, 160–166. Identification of a Kunjin/West Nile-like flavivirus in brains of patients Kalita, J., and Misra, U. K. (1998). EEG in Japanese encephalitis: A with New York encephalitis. Lancet 354, 1261–1262. clinico-radiological correlation. Electroencephalogr. Clin. Neuro- Calisher, C. H., Karabatsos, N., Dalrymple, J. M., Shope, R. E., Porter- physiol. 106, 238–243. field, J. S., Westaway, E. G., and Brandt, W. E. (1989). Antigenic Kaluzova, M., Eleckova, E., Zuffova, E., Pastorek, J., Kaluz, S., Kozuch, O., relationships between flaviviruses as determined by cross-neutral- and Labuda, M. (1994). Reverted virulence of attenuated tick-borne ization tests with polyclonal antisera. J. Gen. Virol. 70, 37–43. encephalitis virus mutant is not accompanied with the changes in Casals, J. (1960). Antigenic relationships between Powassan and Rus- deduced viral envelope protein amino acid sequence. Acta Virol. 38, sian spring-summer encephalitis viruses. Can. Med. Assoc. J. 82, 133–140. 355–358. Kawano, H., Rostapshov, V., Rosen, L., and Lai, C. J. (1993). Genetic Chambers, T. J., Hahn, C. S., Galler, R., and Rice, C. M. (1990). Flavivirus determinants of dengue type 4 virus neurovirulence for mice. J. Virol. genome organization, expression, and replication. Annu. Rev. Micro- 67, 6567–6575. biol. 44, 649–688. Kolski, H., Ford, J. E., Richardson, S., Petric, M., Nelson, S., Jamieson, F., Davidson, M. M., Williams, H., and Macleod, J. A. (1991). Louping ill in Blaser, S., Gold, R., Otsubo, H., Heurter, H., and MacGregor, D. (1998). man: A forgotten disease. J. Infect. 23, 241–249. Etiology of acute childhood encephalitis at The Hospital for Sick De Clercq, E., Cools, M., Balzarini, J., Snoeck, R., Andrei, G., Hosoya, M., Children, Toronto, 1994–1995. Clin. Infect. Dis. 26, 398–409. Shigeta, S., Ueda, T., Minakawa, N., and Matsuda, A. (1991). Antiviral Kramer, L. D., Presser, S. B., Hardy, J. L., and Jackson, A. O. (1997). activities of 5-ethynyl-1-␤-D-ribofuranosylimidazole-4-carboxamide Genotypic and phenotypic variation of selected Saint Louis enceph- and related compounds. Antimicrob. Agents Chemother. 35, 679– alitis viral strains isolated in California. Am. J. Trop. Med. Hyg. 57, 684. 222–229. MODOC VIRUS INFECTIONS 37

Kuno, G., Chang, G. J., Tsuchiya, K. R., Karabatsos, N., and Cropp, C. B. Monath, T. P., and Heinz, F. X. (1996). Flaviviruses. In “Fields’ Virology” (1998). Phylogeny of the genus Flavivirus. J. Virol. 72, 73–83. (B. N. Fields, D. M. Knipe, P. M. Howley, et al., Eds.), Vol. 1, 3rd ed., pp. Lanciotti, R. S., Roehrig, J. T., Deubel, V., Smith, J., Parker, M., Steele, K., 961–1034. Lippincott-Raven, Philadelphia. Crise, B., Volpe, K. E., Crabtree, M. B., Scherret, J. H., Hall, R. A., Pavri, K. (1989). Clinical, clinicopathologic, and hematologic features of Mackenzie, J. S., Cropp, C. B., Panigrahy, B., Ostlund, E., Schmitt, B., Kyasanur Forest disease. Rev. Infect. Dis. 11(Suppl. 4), S854–S859. Malkinson, M., Banet, C., Weissman, J., Komar, N., Savage, H. M., Pierre, V., Drouet, M. T., and Deubel, V. (1994). Identification of mos- Stone, W., McNamara, T., and Gubler, D. J. (1999). Origin of the West quito-borne flavivirus sequences using universal primers and re- Nile virus responsible for an outbreak of encephalitis in the north- verse transcription/polymerase chain reaction. Res. Virol. 145, 93– eastern United States. Science 286, 2333–2337. 104. Lin, Y. L., Liao, C. L., Chen, L. K., Yeh, C. T., Liu, C. I., Ma, S. H., Huang, Rice, C. M. (1996). Flaviviridae: The viruses and their replication. In Y. Y., Huang, Y. L., Kao, C. L., and King, C. C. (1998). Study of dengue “Fields’ Virology” (B. N. Fields, D. M. Knipe, P. M. Howley, et al., Eds.), virus infection in SCID mice engrafted with human K562 cells. J. Virol. Vol. 1, 3rd ed., pp. 931–959. Lippincott-Raven, Philadelphia. 72, 9729–9737. Liu, C. T., and Griffin, M. J. (1982). Changes in body fluid compartments, Schneider, R. J., Firestone, M. H., Edelman, R., Chieowanich, P., and tissue water and electrolyte distribution, and lipid concentrations in Pornpibul, R. (1974). Clinical sequelae after Japanese encephalitis: A rhesus macaques with yellow fever. Am. J. Vet. Res. 43, 2013–2018. one year follow-up study in Thailand. Southeast Asian J. Trop. Med. Livak, K. J., Flood, S. J., Marmaro, J., Giusti, W., and Deetz, K. (1995). Public Health 5, 560–568. Oligonucleotides with fluorescent dyes at opposite ends provide a Siegel, I. J. (2000). Twelve die of West Nile virus in Israel. BMJ 321, 724. quenched probe system useful for detecting PCR product and nu- Singh, B., and Postic, B. (1970). Enhanced resistance of mice to virulent cleic acid hybridization. PCR Methods Appl. 4, 357–362. Japanese B encephalitis virus following inactivated vaccine and poly Lvov, D. K., Butenko, A. M., Gromashevsky, V. L., Larichev, V. P., Gaid- I:C. J. Infect. Dis. 122, 339–342. amovich, S. Y., Vyshemirsky, O. I., Zhukov, A. N., Lazorenko, V. V., Trepo, C., Vierling, J., Zeytin, F. N., and Gerlich, W. H. (1997). The first Salko, V. N., Kovtunov, A. I., Galimzyanov, K. M., Platonov, A. E., Flaviviridae symposium. Intervirology 40, 279–288. Morozova, T. N., Khutoretskaya, N. V., Shishkina, E. O., and Tsai, T. F., Popovici, F., Cernescu, C., Campbell, G. L., and Nedelcu, N. I. Skvortsova, T. M. (2000). Isolation of two strains of West Nile virus (1998). West Nile encephalitis epidemic in southeastern Romania. during an outbreak in southern Russia, 1999. Emerg. Infect. Dis. 6, Lancet 352, 767–771. 373–376. Varelas, W. I., and Calisher, C. H. (1982). Antigenic relationships of Mackenzie, J. S., and Broom, A. K. (1995). Australian X disease, Murray flaviviruses with undetermined arthropod-borne status. Am. J. Trop. Valley encephalitis and the French connection. Vet. Microbiol. 46, Med. Hyg. 31, 1273–1284. 79–90. Westaway, E. G. (1987). Flavivirus replication strategy. Adv. Virus Res. Malinoski, F. J., Hasty, S. E., Ussery, M. A., and Dalrymple, J. M. (1990). 33, 45–90. Prophylactic ribavirin treatment of dengue type 1 infection in rhesus Wittwer, C. T., Ririe, K. M., Andrew, R. V., David, D. A., Gundry, R. A., monkeys. Antiviral Res. 13, 139–149. and Balis, U. J. (1997). The LightCycler: A microvolume multi- Mandl, C. W., Holzmann, H., Kunz, C., and Heinz, F. X. (1993). Complete genomic sequence of Powassan virus: Evaluation of genetic ele- sample fluorimeter with rapid temperature control. BioTechniques ments in tick-borne versus mosquito-borne flaviviruses. Virology 194, 22, 176–181. 173–184. Worthington, M., Levy, H., and Rice, J. (1973). Late therapy of an arbo- McGuire, K., Holmes, E. C., Gao, G. F., Reid, H. W., and Gould, E. A. virus encephalitis in mice with interferon and interferon stimulators. (1998). Tracing the origins of louping ill virus by molecular phyloge- Proc. Soc. Exp. Biol. Med. 143, 638–643. netic analysis. J. Gen. Virol. 79, 981–988. Zarnke, R. L., and Yuill, T. M. (1985). Modoc-like virus isolated from wild Monath, T. P. (1987). Yellow fever: A medically neglected disease. deer mice (Peromyscus maniculatus) in Alberta. J. Wildl. Dis. 21, Report on a seminar. Rev. Infect. Dis. 9, 165–175. 94–99.