(~) INSTITUTPASTEUR/ELsEVIER Paris 1991
Res. ViroL 1991, 142, 387-394
ELISA for detection of IgM and IgG antibodies to sandfly fever Sicilian virus R. Eitrem (1) (2), S. Vene O) and B. Niklasson (1) (3) a; Department o f Virology, National Bacteriological Laboratory. S-105 21 Stockholm, ~2~Department o f Virology, Karolinska Institute c/o SBL S-105 21 Stockholm, and ts~ National Defense Research Establishment, FOA-$, S-172 90 Sundbyberg (Sweden) SUMMARY
An enzyme-linked immunosorbent assay (ELISA) was developed to detect specific human immunoglobuIin G and M antibodies to sandfly fever Sicilian (SFS) virus. Acute and early convalescent serum pairs with 1> 7 days between the 2 specimens were available from 20 patients and all showed significant optical density (OD) increase and significant titre rise ( i> 4-fold) by IgG ELISA. However, negative or borderline-positive sera were found as late as 11 days after onset of symptoms when tested by IgG ELISA. Specific IgM antibodies were detected during the first week of symptoms, and maximum OD values were obtained during the first 4 weeks after onset of disease. The IgM OD values declined over the following 3-9 months. All sera collected later than 14 months post-onset were negative by IgM ELISA. The combination of early antibody response and the need to test only one serum specimen gives IgM ELISA an advantage over IgG ELISA in patient diagnosis. The IgG ELISA was also evaluated as a seroepidemiological tool and compared to a plaque reduction neutralization test (PRNT) using sera from a no" -al Cypriot population. Of 183 sera tested, 34 (19 %) were positive in plaque reducti¢,, neutralization tests (PRNT) and 113 (62 %) by IgG ELISA. A number of PRNT-negative sera were strongly positive by IgG ELISA and also by indirect immunofluorescence test, which may suggest the presence of a virus related to SFS in Cyprus which has not yet been isolated.
Key-words: Sandfly fever, Phlebovirus, Arbovirus; IgG, IgM, ELISA, Sicilian virus, Cyprus.
INTRODUCTION Sandfly fever (SF), or phlebotomus fever, has historically caused significant morbidity among non-native populations in the Mediterranean region (Pick, 1886; Doerr, 1909). Two viruses, SF Sicilian (SFS) and SF Naples (SFN), were
found to be responsible for extensive outbreaks of febrile illness among allied troops stationed in Italy during World War II (Sabin, 1951 ; Herfig and Sabin, 1964). Later, Toscana (TOS) virus was associated with aseptic meningitis in Italy (Verani et al., 1984; Leoncini, 1986). Several other SF viruses have recently been isolated and
Submitted March 27, 1991, accepted June 12, 1991. Correspondingauthor: Bo Niklasson,Departmentof Virology,NationalBacteriologicalLaboratory,S-10521 Stockholm(Sweden).
388
R. E I T R E M E T A L .
shown to infect man. However, their role as human pathogens is not clear (Peters and LeDuc, 1984). The n u m b e r o f phlebovirus serotypes (Bunyaviridae family) in the Old World, transmitted by phlebotomines and known to cause disease in humans, currently amounts to 3: SFS, SFN and TOS viruses; they all occur a r o u n d the Mediterranean basin (Peters and LeDuc, 1984). The awareness o f SF increased in Sweden when eases occurred among Swedish United Nations troops stationed in Cyprus (Niklasson and Eitrem, 1985). Seroepidemiological follow-up studies among Cypriots and the incidence among Swedish soldiers and tourists visiting Cyprus, showed that SF viruses are a c o m m o n cause o f infection on the island (Eitrem et el., 1990, 1991). SFS was found to be the p r e d o m i n a n t agent among the 3 viruses investigated. Here, we report the development and evaluation o f an ELISA for the detection o f IgG and IgM antibodies to SFS.
MATERIALS AND METHODS
Serum specimens
Acute-, early-convalescent- and late-convalescentphase sera (n = 137) were available from 36 Swedish n~.~L]~nLO o f ' a n f e IS.,LUIS.ILItIII~ vafnrn;nn ~ Fam IIUIll
I-s~..,~ '%-,-,~.[.,PZUD
I A , . - - ; ~~ l~O,J-l~gOOJ 1NO,~ I I"~OOX IkLI[UlIII
with clinical symptoms compatible with SF and diagnosed by indirect immunofluorescence test (IFT) as having been infected with SFS. To evaluate the IgG ELISA as a seroepidemiological tool, serum samples were also collected from Cypriot in- and out-patients (n = 132) who visited the medical, surgical, paedriatric or orthopaedic clinic at Paralimni Hospital, Cyprus, during the autumn of 1985 and the spring of 1986. Sera were randomly collected at the biochemical laboratory from patients regardless of suspected diagnosis, history of previous infectious diseases, or current symptoms. In o,der to obtain sera from younger age groups, sera were also collected from volunteer blood donors from Paralimni (n = 51).
ELISA IFT MEM MOI PBS OD
= = = = = =
enzyme-linked immunosorbent assay. indirect immunofluorescence test. minimal essential medium. multiplicity of infection. phosphate-buffered saline. optical density.
Sera from a rural Swedish population (n = 60) were included to determine the background activity in both IgM and IgG ELISA. In addition, 14 human sera with IgM antibodies to other viruses (3 measles, 1 rubella, 3 cytomegalovirus, 2 mumps, 5 EpsteinBarr virus infections) were tested in the SFS IgM ELISA. Acute-, early-convalescent- and late-convalescentphase sera were shipped overnight to the National Bacteriological Laboratory (NBL). Sera from Cyprus were stored at - 20°C and transported on dry ice to the NBL, Stockholm, Sweden, where all sera were stored at - 2 0 ° C until they were tested. Virus strains
The following virus strains were used: SFS (Sabin), SFN (Sabin) and TOS (ISS Phi 3) (Karabatsos, 1985). Virus strains were kindly provided by Dr. R. Tesh, Yale Arbovirus Research Unit, New Haven, CT, USA. Indirect immunofluorescence test
"Spot slides" were prepared for SFS, SFN and TOS viruses, respectively, by allowing approximately 1/3 infected and 2/3 uninfected Vero cells to attach to cleaned 10-well slides for 12 h. Cells were fixed to slides by exposure to cold anhydrous acetone (4°C) and the slides were stored at - 70°C until they were used. Sera were tested at 2-fold dilutions starting at 1/8. Fluorescein-isothiocyanate-conjugated sheep anti-human Ig (NBL product) was used to detect Ig to SFS, SFN and TOS viruses, respect'~vely. Plaque reduction neutralization test (PRNT)
Neutralizing antibodies against SFS, SFN and TOS were determined by PRNT described by Earley et ai. and modified according to McCown by adding DEAE-dextran, dimethyl sulphoxide and heparin to the agar overlays (Earley et al., 1967; McCown et ai., 1979). The second overlay containing neutral red was added on day 4 for SFS and TOS and on day 7 for SFN. Plaques were enumerated the following day. Sera reducing plaques by t> 80 070 at dilution 1/10 were considered positive.
PRNT SF SFN SFS TOS
= = = = =
plaque reduction neutralization test. sandfly fever. SF Naples (virus). SF Sicilian (virus). Toscana (virus).
lgG A N D IgM E L I S A FOR SA ND F L Y FEVER S I C I L I A N VIRUS
Immune reagents used in IgG and IgM ELISA EL1SA antigen was prepared from SFS-virusinfected Vero cells. Vero cells were infected with a MOI of 0.01 and cultivated for 6 days in Eagle's MEM supplemented with 2 °70 heat-inactivated foetal calf serum before harvest. After centrifugation (800 g), ceils and supernatant fluids were separated. Cell pellets were resuspended in 1/10 of the original volume (using the supernatant as diluent), sonicated, centrifuged again, and the supernatant was used as antigen in the ELISA. A negative control antigen was prepared from uninfected cells by the same procedure. For production of immune serum, mice were immunized intraperitoneally with 0.5 ml of 10 070suckling mouse brain homogenate of the virus mixed with Freund's complete adjuvant. The mice received 3 additional doses (identical to the first injection but with Freund's incomplete adjuvant) in 2-week intervals. Sera were collected 2 months after the initial immunization and were found to have a titre of 2,560 by PRNT.
Detection of IgG antibodies to SFS by ELISA A sandwich ELISA was employed as follows. Mouse anti-SFS antibody diluted 1/800 in coating buffer (0.05 M sodium carbonate pH 9.6) was adsorbed to each of 96 wells of a polystyrene microtitre plate (E1A-plate no. 3590, Costar, MA, USA) and incubated at 37°C for 1 h. Virus antigen diluted 1,1/lr~l,~,, in ELISA I-t,u~i . . . . . . (PBS without Mg ++ and Ca + + but with 0.05 070 Tween-20 and 0.5 070 bovine serum albumin) was added and plates were incubated at 37°C for 1 h, test serum was diluted in ELISA buffer and incubated at 37°C for 1 h, and alkalinephosphatase-conjugated swine anti-human IgG (Orion Diagnostica, Espoo, Finland) was diluted 1/100 in ELISA buffer and incubated at 37°C for 1 h. p-Nitrophenyl-phosphate (Sigma, St Louis, MO, USA) diluted in diethanolamine buffer (1 M diethanolamine pH 9.8, 0.5 mM MgC12) was used as the substrate. Washing between each step was done 6 times using washing buffer (saline with 0.05 070 Tween-20). The reaction was read after 30 min at room temperature in a spectrophotometer at 405 nm and expressed as optical density (OD). Optimal dilutions of all reagents used in the ELISA were determined by box titrations. All specimens were tested in duplicate with viral antigen and negative control antigen. The OD was calculated as the average OD with antigen minus the average OD wiwh negative control antigen. The borderline betweer, positive a~d negative was calculated as tile mean of the test result .~f 60 known negative sera (rural Swedish population) plus 3 standard devi-
389
ations. An OD of 0.100 or higher was considered positive. To adjust for plate-to-plate and day-to-day variations in the assay, a positive control serum was included on all plates. This control had an OD of 0.900 (in the linear interval of this IgG ELISA). The plate was accepted if the positive control serum had an OD between 0.800 and 1.000. In addition, all OD values on the plate were multiplied by a factor which made the positive control value equal 1.000. The sera were tested in 10-fold dilutions starting at 1/100 and the ELISA titre was calculated as the point at which the antibody titration curve crossed the A40s=0.100 (Sundqvist and Wahren, 1981). A significant antibody rise was defined as an OD increase of 100 070between paired sera using the single dilution or a 4-fold titre rise using the titration method.
Detection of lgM antibodies to SFS by ELISA A ~t-capture ELISA was employed as follows. Goat anti-human IgM (~t-chain-specific) (Cappel Laboratories, Cochranville, PA, USA) diluted 1/500 in coating buffer was adsorbed to each of the 96 wells in polystyrene microtitre plates (Cooke M-29 AR, Dynatec Laboratories, Alexandria, VA, USA) at 37°C for 2 h. Plates were consecutively incubated at 37°C for 1 h with human test serum diluted 1/100 in ELISA buffer, virus antigen diluted 1/50 in ELISA buffer with 0.2 % normal human serum, mouse antibody to SFS virus diluted 1/400 in ELISA buffer and alkaline-phosphatase-labelled goat anti-mouse IgG (Kirkegaard and Perry lab, Geithersburg, MD, USA, cat. no. 051506) diluted 1/400 in ELISA buffer, p-Nitrophenyl-phosphate diluted in diethanolamine buffer was used as substrate. Washing between each step was done 6 times using washing buffer. The reaction was read after 30 min at room temperature in a spectrophotometer at 405 nm and expressed as OD. Optimal dilutions of all reagents used, the OD results and adjustment for plate-to-plate and day-to-day variations were calculated as described above for the SFS IgG ELISA. The borderline between positive and negative was calculated as the mean of the test result of 60 known negative sera plus 3 standard deviations. An OD of 0.120 or higher was considered positive.
RESULTS Detection of IgG antibodies to SFS by ELISA Patients with acute illness were serologically diagnosed by IFT in which a 4-fold titre rise between acute and convalescent sera occurred or
390
R. EITREM ET AL.
3,000
OD at 405 nm x 1000 -
2,500
.,
~..Y , o.:~.J~
...........
/+ ta
2,000
:"
-
o
y,
•
1,500
/
-
/
/ I
1,000
"
r
..,"
1" ."
'
500
~p~
/.
•
x
." "x,. i,,
.
. " ,.,,,,,
i+, SC
"
..\
",
+" - ' ' " I .4~ • " • S"
.
." -
/,;
i."
x.
.
II ,.. ..... "ll".
•
..
Jl s"
.I ~
" P;
/ .,,,,.. /
./
/ •/
/
•
•.
..J
m
.-S'.
."i,"
IK
• ""
--
...... "'""
• •
%
"'l..
....''""
"°,. o, ,°
.°°°
oi 1
3
5
7
~
9 11 13 15 17 19 21 23 25 27 29"'1 DAYS
3
TIME AFTER ONSET OF ILLNESS
5
7
%° ID
9 11 2
MONTHS
4 YEARS
Fig. 1. Acute-, early-convalescent-and late-convalescent-phase sera (n = 137) from 36 serologically (IFT) confirmed SFS-infected subjects tested by lgG ELISA. The results are expressed as OD values multiplied by 1,000 using a single dilution (1/100) technique. Results from individuals with /> 5 sera are connected with a line.
a positive titre (i> 8) in convalescent serum (when no acute serum was available) was noted. P R N T was used as a confirmatory test on 18 randomly selected patients. The IgG E L I S A results for all 137 acute-, early-convalescent- and late-convalescent-phase sera f r o m 36 serologically (IFT) confirmed SFS infections are seen in figures I and 2. Figure 1 shows IgG E L I S A OD results using a single dilution (1/100) technique; figure 2 shows the titres. Acute- and earlyconvalescent-phase serum pairs sampled during the first m o n t h after onset of symptoms with /> 7 days between the 2 specimens were available from 20 patients. All 20 patients showed a significant OD increase (>i 100 °70)and a significant titre rise (>1 4-fold) by IgG ELISA. O f the 6 patients from w h o m >/ 5 samples were obtained, both maximal titre and OD value was reached 1-12 m o n t h s after onset. Sera collected 2-5 years after onset showed a minor decline in
titre as well as in OD value. In one patient, the antibody response was very low, with the 3-year follow-up specimen being borderline negative. The remaining patients were all IgG-ELISApositive during the entire study period. Results of tests with 183 sera from Cyprus by P R N T and IgG E L I S A are seen in table I. A to-
I. Serum samples collected from in- and outpatients (n = 132) at Paralimni hospital and volunteer blood donors (n = 51) from Paralimni tested by IgG ELISA and PRNT. Table
IgG-ELISA + IgG-ELISAPRNT + P RNT Total
34 79 113
2 68 70
Total 36 147 183
IgG AND
IgM ELISA
FOR SANDFL Y FEVER
SICILIAN
391
VIRUS
1,000,000
100,000
i o
-%
10,000 nIJJ
I-
I-"
1,000
0
.-3 ILl
0
l/ " , ;I;"; " '/ 'o
./.'.¢ /.'J
•
O
,,"..
"":,
," °" .,4" ................. ;": ................ • ..... ,~.......... o... ,,"
e ...
••
"~
100 r
Z
10
o J d,
1
3
5
7
9 11 13 15 17 19 21 23 25 27 29-1
3
DAYS
5 7
9 11-2
MONTHS
4 YEARS
TIME AFTER ONSET OF ILLNESS
Fig. 2. Acute-, early-convalescent- and late-convalescent-phase sera (n = 137) from 36 serologically OFT) confirmed SFS-infected subjects tested by IgG ELISA.
The results are expressed as endpoint titres calculated as described elsewhere (Sundqvist and Wahren, 1981). Results from individuals with ~> 5 sera are connected with a line.
tal of 34/183 (19 °70) were found positive by P R N T and 113/183 (62 %) by IgG ELISA. All PRNT-positive samples except 2 (both with P R N T titres of 10) were IgG-ELISA-positive. The mean IgG ELISA OD were 0.469, 0.639 and 1.160 for the PRNT-positive samples with titres of 10, 40 and i> 160, respectively. Forty-four sera had an IgG ELISA OD o f m o r e than 0.500 and 20 had an OD > 1.000. A m o n g the 20 sera with an OD o f 1.000 or more, 10 were P R N T negative. These sera were all retested by P R N T and confirmed as negative and tested by IgG IFT and found to be positive with typical cytoplasmic staining at a titre o f 8 or more. IgG ELISA was positive for 24 Cypriots for w h o m P R N T was < 10 for SFS, SFN and TOS.
Detection of IgM antibodies to SFS by ELNSA IgM ELISA results for all I37 acute-, earlyconvalescent- and late-convalescent-phase sera from 36 confirmed SFS infections are seen in figure 3. lgM antibodies to SFS were detected 5 days after onset of symptoms. High ag~v~ EI_ISA OD were generally found 1 to 4 weeks after onset of symptoms. By 3-8 months, OD w¢:e declining. Sera collected 9-13 months after onset were negative or in the lower range of significance, and sera collected later than 14 mqr~ths after onset were all negative. All 14 human sera with lgM antibodies to other viruses (3 measles, 1 rubella, 3 cytome-
R. EITREM E T AL.
392
3,000
OD at 405 nm x 1000
x. \.
°
.
Q
a 2,000 O
° X
• ~
•
\
/i,
x.\
/:.
N
.
\%\
~ ~,~
o...~..
1,000
......
'. ',
500
0
I,~-~;~,, 3
5
,
r
i . ,
I I r.,
.,,.'|"'!~'~'-~-
7 9 11 13 15 17 19 21 23 25 27 29"1 3 DAYS TIME AFTER ONSET OF ILLNESS
S 7 9 11rt2-4 MONTHS YEARS
Fig. 3. Acute-, early-convalescent- and late-convalescent-phase sera (n = 137) from 36 serologically OFT) confirmed SFS-infected subjects tested by !gM ELISA. The results are expressed as OD values multiplied by 1,000 using a single dilution (1/100) technique. Results from individuals with I> 5 sera are connected with a line.
galovirus, 2 mumps, 5 Epstein-Barr virus infections) were negative in the SFS IgM ELISA.
DISCUSSION
Previous studies have shown that SF is endemic in Cyprus (Eitrem et al., 1990). Antibody prevalence studies have found high rates among native Cypriots of SFS, SFN and TOS viruses (Eitrem et al., unpublished observations). That serosurvey found the highest antibody prevalence rates by neutralization assay to SFN. However, studies of the incidence in Cyprus among both Swedish tourists arid Swedish UN soldiers have found SFS virus to be the dominatin~ aetiologic agent. Both SFS and SFN viruses have recently
been isolated from patient sera (Eitrem et al., 1990). Despite the high antibody prevalence rate among the local population in endemic areas aiid despite the significant morbidity caused by SF viruses among non-immune populations (military troops and tourists) temporarily visiting endemic areas, little documentation of the disease among the native populations is available. It has been speculated that infection at an early age may be mild or subclinical, but few studies have been undertaken to answer this question. The clinical picture with fever and general malaise as described in most textbooks suggests that serological confirmation is necessary to diagnose SF. Sensitive and specific serological tests are necessary for early diagnosis in order to ex-
IgG A N D I g M E L I S A F O R S A N D F L Y F E V E R S I C I L I A N [qRUS
clude an a e d o l o g y which requires specific treatment and to help define the complete clinical range o f the disease. Serological tests are also needed for the discovery o f new endemic foci. In the present study, we evaluated IgG and IgM E L I S A for detecting antibodies to SFS virus by testing sera f r o m Swedish patients with symptoms compatible with SF and confirmed by IFT. IgM E L I S A proved useful as a diagnostic tool testing a single serum at one dilution only. Seroconversions as well as significant titre rises were found in all paired sera tested by IgG ELiSA. Negative or borderline-positive sera were found as late as 11 days after onset o f symptoms. The c o m b i n a t i o n o f a detectable early antibody response and the possibility o f testing a single serum specimen gives IgM E L I S A an advantage over IgG E L I S A in patient diagnosis. However, it should be r e m e m b e r e d that the present study was p e r f o r m e d on sera f r o m patients w h o had most likely experienced their first SF infection. It is necessary to further test sensitivity and specificity in an endemic area where several SF viruses are present. IgG E L I S A was c o m p a r e d with standard P R N T as a seroepidemioiogical tool by testing sera from a n o r m a l Cypriot population. IgG E L I S A f o u n d a very high (62 070) antibody prevalence rate compared to P R N T (19 070). The data suggest that IgG E L I S A will detect all sera positive by P R N T . However, a high proportion o f the sera were antibody-positive by IgG ELISA but -negative by P R N T . Because s o m e o f these sera were strongly positive by IgG E L I S A and positive by IFT, they m a y contain antibodies to a serologically related but as yet unidentified virus.
Acknowledgement
We thank Marona Engvall, Jan Lundstr6m and Sven Bj6rsten for excellent technical support and Manolis Stylianou, New Larnaca Hospital, Cyprus, for collecting serum specimens.
393
ELISA et d~tection des anticor-ps !gM et lgG contre le virus sicilien de la fibvre/t phl6botomes
Deux s&ums pr61ev6s /l i> 7 jours d'intervalle chez 20 sujets en phase aigu~ et en phase pr6coce de convalescence ont tous montr~ une 61~vation significative de la densit6 optique (DO) et du titre (>I 4 fois) des IgG en ELISA; mais des r6sultats n6gatifs ou linfites ont 6td not6s en ELISA au 11".jour apr6s le d6but des symptfmes. Des anticorps IgM ont 6t6 d6cel6s pendant la premiere semaine de la symptomatologie, et l'on observe des valeurs maximales de la DO durant les quatre premieres semaines; ces valeurs diminuent au cours des 3-9 mois. Tousles s6rums recueillis apr~s le 14e mois se sont montr~s n6gatifs en ELISA pour les IgM. La r~ponse anticorps pr~coce et l'int~r& d'un diagnostic sur un ~chantillon unique de s&um donnent /l I'ELISA lgM un avantage certain sur I'ELISA IgG. Le dosage des IgG par ELISA, compar6 au test de r6duct';on/neutralisation en plaque (PRNT), a 6t~ utile pour l'6tude s~ro-6pid6miologique d'une population normale de l'ile de Chypre. Sur les 183 s~rums examines, 14 (19 °/0) 6taient positifs en PRNT et 113 (62 %) l'~taient en ELISA. Certains s6rums PRNT n~gatifs 6taient fortement positifs en ELISA IgG ainsi qu'en irnmunofluorescence indirecte. Ces observations sugg&ent la presence/t Chypre d'un virus apparent6 au virus sicilien de la fi~vre/l phl~botome, mais qui n'a pas encore ~t~ isol_~. Mots-clds: Fi~vre ~ phl6botomes, Arbovirus, Phi~bovirus; IgG, IgM, ELISA, Virus sicilien, Chypre.
References
Doerf, R., Frank, K. & Taussig, S. (1909), Das Pappatacifieber, ein epidemisches Drei-Tage-Fieber im Adriatischen Kiistengebiete Oesterreich-Ungarns. Wien. Kiln. Wschr., 22, 609-610. Earley, E., Peralta, P.H. & Johnson, K.M. (1967), A plaque neutralization method for arboviruses. Proc. Soc. exp. Med., 125, 741-747. Eitrem, R., Niklasso~, B. & Weiland~ O. (!991)7 Sandfiy fever among Sw, dish tourists. Scand. J. infect. Dis. (in press). Eitrem, R., Vene, S. & Niklasson, B. (1990), Incidence of sandflv fever among, Swedish United Nations soldiers on Cyprus during 1985.Amer. J. trop. Med. Hyg., 43, 207-211. Hertig, M. & Sabin, A.B. 0964), Sandfly fever (Pappata-
394
R. E I T R E M E T A L .
ci, Phlebotomus, three-day-fever), in "Preventive medicine in World War II: volume 7" (Coates, J.B.) (pp. 109-174). Communicable diseases. US Government Printing Office, Washington, DC. Karabatsos, N. (1985), International catalogue of arboviruses including certain other viruses of vertebrates. Third ed. American Society of Tropical Medicine & Hygiene. San Antonio Texas. (Naples pp. 893-894; Sicilian pp. 895-896; Toscana pp. 1027-1028). Leoncini, F., Bartolozzi, D., Banchi, S., Balducci, M., Fratiglioni, L., CacioUi, S. & Rence, A. (1986), Toscana virus : a new phlebovirus, aetiological agent of acute inflammatory disease of the human central nervous system. G. Mal. infett., 38, 649-52. McCown, J.M., Brandt, W.E., Bancroft, W.H. & Russell, P.K. (1979), Dimethyl sulfoxide enhancement of Phiebotomus fever virus plaque formation. Amer. J. trop. Med. Hyg., 28, 733-739.
Niklasson, B. & Eitrem, R. (1985), Sandfly fever among Swedish UN troops in Cyprus. Lancet, I, 1212. Peters, C.J. & LeDuc, J.W. (1984), Bunyaviruses, phleboviruses and related viruses, in "Textbook of human virology" (Belshe, R.B.) (pp. 572-576). PSG Publ. Co., Littleton, MA. Pick, A. 0986), Zur Pathologie und Therapie einer eigentumlischen endemischen Krankheitsform. Wien. Med. Wschr., 36, 1141-45 and 1168-71. Sabin, A.B. (1951), Experimental studies on Phlebotomus (pappataci, sandfly) fever during World War II. Arch. ges. Virusforsch., 4, 367-410. Sundqvist, V. & Wahren, B. (1981), An interchangeable ELISA for cytomegalovirus antigen and antibody. J. Virol. Methods, 2, 301-312. Verani, P., Nicoletti, L. & Ciufoloni, M.G. (1984), Antigenic and biological characterization of Toscana virus, a new Phlebotomus fever group virus isolated in Italy. Acta. Virol., 28, 39-47.
Res. ViroL 1991, 142, 387-394
ELISA for detection of IgM and IgG antibodies to sandfly fever Sicilian virus R. Eitrem (1) (2), S. Vene O) and B. Niklasson (1) (3) a; Department o f Virology, National Bacteriological Laboratory. S-105 21 Stockholm, ~2~Department o f Virology, Karolinska Institute c/o SBL S-105 21 Stockholm, and ts~ National Defense Research Establishment, FOA-$, S-172 90 Sundbyberg (Sweden) SUMMARY
An enzyme-linked immunosorbent assay (ELISA) was developed to detect specific human immunoglobuIin G and M antibodies to sandfly fever Sicilian (SFS) virus. Acute and early convalescent serum pairs with 1> 7 days between the 2 specimens were available from 20 patients and all showed significant optical density (OD) increase and significant titre rise ( i> 4-fold) by IgG ELISA. However, negative or borderline-positive sera were found as late as 11 days after onset of symptoms when tested by IgG ELISA. Specific IgM antibodies were detected during the first week of symptoms, and maximum OD values were obtained during the first 4 weeks after onset of disease. The IgM OD values declined over the following 3-9 months. All sera collected later than 14 months post-onset were negative by IgM ELISA. The combination of early antibody response and the need to test only one serum specimen gives IgM ELISA an advantage over IgG ELISA in patient diagnosis. The IgG ELISA was also evaluated as a seroepidemiological tool and compared to a plaque reduction neutralization test (PRNT) using sera from a no" -al Cypriot population. Of 183 sera tested, 34 (19 %) were positive in plaque reducti¢,, neutralization tests (PRNT) and 113 (62 %) by IgG ELISA. A number of PRNT-negative sera were strongly positive by IgG ELISA and also by indirect immunofluorescence test, which may suggest the presence of a virus related to SFS in Cyprus which has not yet been isolated.
Key-words: Sandfly fever, Phlebovirus, Arbovirus; IgG, IgM, ELISA, Sicilian virus, Cyprus.
INTRODUCTION Sandfly fever (SF), or phlebotomus fever, has historically caused significant morbidity among non-native populations in the Mediterranean region (Pick, 1886; Doerr, 1909). Two viruses, SF Sicilian (SFS) and SF Naples (SFN), were
found to be responsible for extensive outbreaks of febrile illness among allied troops stationed in Italy during World War II (Sabin, 1951 ; Herfig and Sabin, 1964). Later, Toscana (TOS) virus was associated with aseptic meningitis in Italy (Verani et al., 1984; Leoncini, 1986). Several other SF viruses have recently been isolated and
Submitted March 27, 1991, accepted June 12, 1991. Correspondingauthor: Bo Niklasson,Departmentof Virology,NationalBacteriologicalLaboratory,S-10521 Stockholm(Sweden).
388
R. E I T R E M E T A L .
shown to infect man. However, their role as human pathogens is not clear (Peters and LeDuc, 1984). The n u m b e r o f phlebovirus serotypes (Bunyaviridae family) in the Old World, transmitted by phlebotomines and known to cause disease in humans, currently amounts to 3: SFS, SFN and TOS viruses; they all occur a r o u n d the Mediterranean basin (Peters and LeDuc, 1984). The awareness o f SF increased in Sweden when eases occurred among Swedish United Nations troops stationed in Cyprus (Niklasson and Eitrem, 1985). Seroepidemiological follow-up studies among Cypriots and the incidence among Swedish soldiers and tourists visiting Cyprus, showed that SF viruses are a c o m m o n cause o f infection on the island (Eitrem et el., 1990, 1991). SFS was found to be the p r e d o m i n a n t agent among the 3 viruses investigated. Here, we report the development and evaluation o f an ELISA for the detection o f IgG and IgM antibodies to SFS.
MATERIALS AND METHODS
Serum specimens
Acute-, early-convalescent- and late-convalescentphase sera (n = 137) were available from 36 Swedish n~.~L]~nLO o f ' a n f e IS.,LUIS.ILItIII~ vafnrn;nn ~ Fam IIUIll
I-s~..,~ '%-,-,~.[.,PZUD
I A , . - - ; ~~ l~O,J-l~gOOJ 1NO,~ I I"~OOX IkLI[UlIII
with clinical symptoms compatible with SF and diagnosed by indirect immunofluorescence test (IFT) as having been infected with SFS. To evaluate the IgG ELISA as a seroepidemiological tool, serum samples were also collected from Cypriot in- and out-patients (n = 132) who visited the medical, surgical, paedriatric or orthopaedic clinic at Paralimni Hospital, Cyprus, during the autumn of 1985 and the spring of 1986. Sera were randomly collected at the biochemical laboratory from patients regardless of suspected diagnosis, history of previous infectious diseases, or current symptoms. In o,der to obtain sera from younger age groups, sera were also collected from volunteer blood donors from Paralimni (n = 51).
ELISA IFT MEM MOI PBS OD
= = = = = =
enzyme-linked immunosorbent assay. indirect immunofluorescence test. minimal essential medium. multiplicity of infection. phosphate-buffered saline. optical density.
Sera from a rural Swedish population (n = 60) were included to determine the background activity in both IgM and IgG ELISA. In addition, 14 human sera with IgM antibodies to other viruses (3 measles, 1 rubella, 3 cytomegalovirus, 2 mumps, 5 EpsteinBarr virus infections) were tested in the SFS IgM ELISA. Acute-, early-convalescent- and late-convalescentphase sera were shipped overnight to the National Bacteriological Laboratory (NBL). Sera from Cyprus were stored at - 20°C and transported on dry ice to the NBL, Stockholm, Sweden, where all sera were stored at - 2 0 ° C until they were tested. Virus strains
The following virus strains were used: SFS (Sabin), SFN (Sabin) and TOS (ISS Phi 3) (Karabatsos, 1985). Virus strains were kindly provided by Dr. R. Tesh, Yale Arbovirus Research Unit, New Haven, CT, USA. Indirect immunofluorescence test
"Spot slides" were prepared for SFS, SFN and TOS viruses, respectively, by allowing approximately 1/3 infected and 2/3 uninfected Vero cells to attach to cleaned 10-well slides for 12 h. Cells were fixed to slides by exposure to cold anhydrous acetone (4°C) and the slides were stored at - 70°C until they were used. Sera were tested at 2-fold dilutions starting at 1/8. Fluorescein-isothiocyanate-conjugated sheep anti-human Ig (NBL product) was used to detect Ig to SFS, SFN and TOS viruses, respect'~vely. Plaque reduction neutralization test (PRNT)
Neutralizing antibodies against SFS, SFN and TOS were determined by PRNT described by Earley et ai. and modified according to McCown by adding DEAE-dextran, dimethyl sulphoxide and heparin to the agar overlays (Earley et al., 1967; McCown et ai., 1979). The second overlay containing neutral red was added on day 4 for SFS and TOS and on day 7 for SFN. Plaques were enumerated the following day. Sera reducing plaques by t> 80 070 at dilution 1/10 were considered positive.
PRNT SF SFN SFS TOS
= = = = =
plaque reduction neutralization test. sandfly fever. SF Naples (virus). SF Sicilian (virus). Toscana (virus).
lgG A N D IgM E L I S A FOR SA ND F L Y FEVER S I C I L I A N VIRUS
Immune reagents used in IgG and IgM ELISA EL1SA antigen was prepared from SFS-virusinfected Vero cells. Vero cells were infected with a MOI of 0.01 and cultivated for 6 days in Eagle's MEM supplemented with 2 °70 heat-inactivated foetal calf serum before harvest. After centrifugation (800 g), ceils and supernatant fluids were separated. Cell pellets were resuspended in 1/10 of the original volume (using the supernatant as diluent), sonicated, centrifuged again, and the supernatant was used as antigen in the ELISA. A negative control antigen was prepared from uninfected cells by the same procedure. For production of immune serum, mice were immunized intraperitoneally with 0.5 ml of 10 070suckling mouse brain homogenate of the virus mixed with Freund's complete adjuvant. The mice received 3 additional doses (identical to the first injection but with Freund's incomplete adjuvant) in 2-week intervals. Sera were collected 2 months after the initial immunization and were found to have a titre of 2,560 by PRNT.
Detection of IgG antibodies to SFS by ELISA A sandwich ELISA was employed as follows. Mouse anti-SFS antibody diluted 1/800 in coating buffer (0.05 M sodium carbonate pH 9.6) was adsorbed to each of 96 wells of a polystyrene microtitre plate (E1A-plate no. 3590, Costar, MA, USA) and incubated at 37°C for 1 h. Virus antigen diluted 1,1/lr~l,~,, in ELISA I-t,u~i . . . . . . (PBS without Mg ++ and Ca + + but with 0.05 070 Tween-20 and 0.5 070 bovine serum albumin) was added and plates were incubated at 37°C for 1 h, test serum was diluted in ELISA buffer and incubated at 37°C for 1 h, and alkalinephosphatase-conjugated swine anti-human IgG (Orion Diagnostica, Espoo, Finland) was diluted 1/100 in ELISA buffer and incubated at 37°C for 1 h. p-Nitrophenyl-phosphate (Sigma, St Louis, MO, USA) diluted in diethanolamine buffer (1 M diethanolamine pH 9.8, 0.5 mM MgC12) was used as the substrate. Washing between each step was done 6 times using washing buffer (saline with 0.05 070 Tween-20). The reaction was read after 30 min at room temperature in a spectrophotometer at 405 nm and expressed as optical density (OD). Optimal dilutions of all reagents used in the ELISA were determined by box titrations. All specimens were tested in duplicate with viral antigen and negative control antigen. The OD was calculated as the average OD with antigen minus the average OD wiwh negative control antigen. The borderline betweer, positive a~d negative was calculated as tile mean of the test result .~f 60 known negative sera (rural Swedish population) plus 3 standard devi-
389
ations. An OD of 0.100 or higher was considered positive. To adjust for plate-to-plate and day-to-day variations in the assay, a positive control serum was included on all plates. This control had an OD of 0.900 (in the linear interval of this IgG ELISA). The plate was accepted if the positive control serum had an OD between 0.800 and 1.000. In addition, all OD values on the plate were multiplied by a factor which made the positive control value equal 1.000. The sera were tested in 10-fold dilutions starting at 1/100 and the ELISA titre was calculated as the point at which the antibody titration curve crossed the A40s=0.100 (Sundqvist and Wahren, 1981). A significant antibody rise was defined as an OD increase of 100 070between paired sera using the single dilution or a 4-fold titre rise using the titration method.
Detection of lgM antibodies to SFS by ELISA A ~t-capture ELISA was employed as follows. Goat anti-human IgM (~t-chain-specific) (Cappel Laboratories, Cochranville, PA, USA) diluted 1/500 in coating buffer was adsorbed to each of the 96 wells in polystyrene microtitre plates (Cooke M-29 AR, Dynatec Laboratories, Alexandria, VA, USA) at 37°C for 2 h. Plates were consecutively incubated at 37°C for 1 h with human test serum diluted 1/100 in ELISA buffer, virus antigen diluted 1/50 in ELISA buffer with 0.2 % normal human serum, mouse antibody to SFS virus diluted 1/400 in ELISA buffer and alkaline-phosphatase-labelled goat anti-mouse IgG (Kirkegaard and Perry lab, Geithersburg, MD, USA, cat. no. 051506) diluted 1/400 in ELISA buffer, p-Nitrophenyl-phosphate diluted in diethanolamine buffer was used as substrate. Washing between each step was done 6 times using washing buffer. The reaction was read after 30 min at room temperature in a spectrophotometer at 405 nm and expressed as OD. Optimal dilutions of all reagents used, the OD results and adjustment for plate-to-plate and day-to-day variations were calculated as described above for the SFS IgG ELISA. The borderline between positive and negative was calculated as the mean of the test result of 60 known negative sera plus 3 standard deviations. An OD of 0.120 or higher was considered positive.
RESULTS Detection of IgG antibodies to SFS by ELISA Patients with acute illness were serologically diagnosed by IFT in which a 4-fold titre rise between acute and convalescent sera occurred or
390
R. EITREM ET AL.
3,000
OD at 405 nm x 1000 -
2,500
.,
~..Y , o.:~.J~
...........
/+ ta
2,000
:"
-
o
y,
•
1,500
/
-
/
/ I
1,000
"
r
..,"
1" ."
'
500
~p~
/.
•
x
." "x,. i,,
.
. " ,.,,,,,
i+, SC
"
..\
",
+" - ' ' " I .4~ • " • S"
.
." -
/,;
i."
x.
.
II ,.. ..... "ll".
•
..
Jl s"
.I ~
" P;
/ .,,,,.. /
./
/ •/
/
•
•.
..J
m
.-S'.
."i,"
IK
• ""
--
...... "'""
• •
%
"'l..
....''""
"°,. o, ,°
.°°°
oi 1
3
5
7
~
9 11 13 15 17 19 21 23 25 27 29"'1 DAYS
3
TIME AFTER ONSET OF ILLNESS
5
7
%° ID
9 11 2
MONTHS
4 YEARS
Fig. 1. Acute-, early-convalescent-and late-convalescent-phase sera (n = 137) from 36 serologically (IFT) confirmed SFS-infected subjects tested by lgG ELISA. The results are expressed as OD values multiplied by 1,000 using a single dilution (1/100) technique. Results from individuals with /> 5 sera are connected with a line.
a positive titre (i> 8) in convalescent serum (when no acute serum was available) was noted. P R N T was used as a confirmatory test on 18 randomly selected patients. The IgG E L I S A results for all 137 acute-, early-convalescent- and late-convalescent-phase sera f r o m 36 serologically (IFT) confirmed SFS infections are seen in figures I and 2. Figure 1 shows IgG E L I S A OD results using a single dilution (1/100) technique; figure 2 shows the titres. Acute- and earlyconvalescent-phase serum pairs sampled during the first m o n t h after onset of symptoms with /> 7 days between the 2 specimens were available from 20 patients. All 20 patients showed a significant OD increase (>i 100 °70)and a significant titre rise (>1 4-fold) by IgG ELISA. O f the 6 patients from w h o m >/ 5 samples were obtained, both maximal titre and OD value was reached 1-12 m o n t h s after onset. Sera collected 2-5 years after onset showed a minor decline in
titre as well as in OD value. In one patient, the antibody response was very low, with the 3-year follow-up specimen being borderline negative. The remaining patients were all IgG-ELISApositive during the entire study period. Results of tests with 183 sera from Cyprus by P R N T and IgG E L I S A are seen in table I. A to-
I. Serum samples collected from in- and outpatients (n = 132) at Paralimni hospital and volunteer blood donors (n = 51) from Paralimni tested by IgG ELISA and PRNT. Table
IgG-ELISA + IgG-ELISAPRNT + P RNT Total
34 79 113
2 68 70
Total 36 147 183
IgG AND
IgM ELISA
FOR SANDFL Y FEVER
SICILIAN
391
VIRUS
1,000,000
100,000
i o
-%
10,000 nIJJ
I-
I-"
1,000
0
.-3 ILl
0
l/ " , ;I;"; " '/ 'o
./.'.¢ /.'J
•
O
,,"..
"":,
," °" .,4" ................. ;": ................ • ..... ,~.......... o... ,,"
e ...
••
"~
100 r
Z
10
o J d,
1
3
5
7
9 11 13 15 17 19 21 23 25 27 29-1
3
DAYS
5 7
9 11-2
MONTHS
4 YEARS
TIME AFTER ONSET OF ILLNESS
Fig. 2. Acute-, early-convalescent- and late-convalescent-phase sera (n = 137) from 36 serologically OFT) confirmed SFS-infected subjects tested by IgG ELISA.
The results are expressed as endpoint titres calculated as described elsewhere (Sundqvist and Wahren, 1981). Results from individuals with ~> 5 sera are connected with a line.
tal of 34/183 (19 °70) were found positive by P R N T and 113/183 (62 %) by IgG ELISA. All PRNT-positive samples except 2 (both with P R N T titres of 10) were IgG-ELISA-positive. The mean IgG ELISA OD were 0.469, 0.639 and 1.160 for the PRNT-positive samples with titres of 10, 40 and i> 160, respectively. Forty-four sera had an IgG ELISA OD o f m o r e than 0.500 and 20 had an OD > 1.000. A m o n g the 20 sera with an OD o f 1.000 or more, 10 were P R N T negative. These sera were all retested by P R N T and confirmed as negative and tested by IgG IFT and found to be positive with typical cytoplasmic staining at a titre o f 8 or more. IgG ELISA was positive for 24 Cypriots for w h o m P R N T was < 10 for SFS, SFN and TOS.
Detection of IgM antibodies to SFS by ELNSA IgM ELISA results for all I37 acute-, earlyconvalescent- and late-convalescent-phase sera from 36 confirmed SFS infections are seen in figure 3. lgM antibodies to SFS were detected 5 days after onset of symptoms. High ag~v~ EI_ISA OD were generally found 1 to 4 weeks after onset of symptoms. By 3-8 months, OD w¢:e declining. Sera collected 9-13 months after onset were negative or in the lower range of significance, and sera collected later than 14 mqr~ths after onset were all negative. All 14 human sera with lgM antibodies to other viruses (3 measles, 1 rubella, 3 cytome-
R. EITREM E T AL.
392
3,000
OD at 405 nm x 1000
x. \.
°
.
Q
a 2,000 O
° X
• ~
•
\
/i,
x.\
/:.
N
.
\%\
~ ~,~
o...~..
1,000
......
'. ',
500
0
I,~-~;~,, 3
5
,
r
i . ,
I I r.,
.,,.'|"'!~'~'-~-
7 9 11 13 15 17 19 21 23 25 27 29"1 3 DAYS TIME AFTER ONSET OF ILLNESS
S 7 9 11rt2-4 MONTHS YEARS
Fig. 3. Acute-, early-convalescent- and late-convalescent-phase sera (n = 137) from 36 serologically OFT) confirmed SFS-infected subjects tested by !gM ELISA. The results are expressed as OD values multiplied by 1,000 using a single dilution (1/100) technique. Results from individuals with I> 5 sera are connected with a line.
galovirus, 2 mumps, 5 Epstein-Barr virus infections) were negative in the SFS IgM ELISA.
DISCUSSION
Previous studies have shown that SF is endemic in Cyprus (Eitrem et al., 1990). Antibody prevalence studies have found high rates among native Cypriots of SFS, SFN and TOS viruses (Eitrem et al., unpublished observations). That serosurvey found the highest antibody prevalence rates by neutralization assay to SFN. However, studies of the incidence in Cyprus among both Swedish tourists arid Swedish UN soldiers have found SFS virus to be the dominatin~ aetiologic agent. Both SFS and SFN viruses have recently
been isolated from patient sera (Eitrem et al., 1990). Despite the high antibody prevalence rate among the local population in endemic areas aiid despite the significant morbidity caused by SF viruses among non-immune populations (military troops and tourists) temporarily visiting endemic areas, little documentation of the disease among the native populations is available. It has been speculated that infection at an early age may be mild or subclinical, but few studies have been undertaken to answer this question. The clinical picture with fever and general malaise as described in most textbooks suggests that serological confirmation is necessary to diagnose SF. Sensitive and specific serological tests are necessary for early diagnosis in order to ex-
IgG A N D I g M E L I S A F O R S A N D F L Y F E V E R S I C I L I A N [qRUS
clude an a e d o l o g y which requires specific treatment and to help define the complete clinical range o f the disease. Serological tests are also needed for the discovery o f new endemic foci. In the present study, we evaluated IgG and IgM E L I S A for detecting antibodies to SFS virus by testing sera f r o m Swedish patients with symptoms compatible with SF and confirmed by IFT. IgM E L I S A proved useful as a diagnostic tool testing a single serum at one dilution only. Seroconversions as well as significant titre rises were found in all paired sera tested by IgG ELiSA. Negative or borderline-positive sera were found as late as 11 days after onset o f symptoms. The c o m b i n a t i o n o f a detectable early antibody response and the possibility o f testing a single serum specimen gives IgM E L I S A an advantage over IgG E L I S A in patient diagnosis. However, it should be r e m e m b e r e d that the present study was p e r f o r m e d on sera f r o m patients w h o had most likely experienced their first SF infection. It is necessary to further test sensitivity and specificity in an endemic area where several SF viruses are present. IgG E L I S A was c o m p a r e d with standard P R N T as a seroepidemioiogical tool by testing sera from a n o r m a l Cypriot population. IgG E L I S A f o u n d a very high (62 070) antibody prevalence rate compared to P R N T (19 070). The data suggest that IgG E L I S A will detect all sera positive by P R N T . However, a high proportion o f the sera were antibody-positive by IgG ELISA but -negative by P R N T . Because s o m e o f these sera were strongly positive by IgG E L I S A and positive by IFT, they m a y contain antibodies to a serologically related but as yet unidentified virus.
Acknowledgement
We thank Marona Engvall, Jan Lundstr6m and Sven Bj6rsten for excellent technical support and Manolis Stylianou, New Larnaca Hospital, Cyprus, for collecting serum specimens.
393
ELISA et d~tection des anticor-ps !gM et lgG contre le virus sicilien de la fibvre/t phl6botomes
Deux s&ums pr61ev6s /l i> 7 jours d'intervalle chez 20 sujets en phase aigu~ et en phase pr6coce de convalescence ont tous montr~ une 61~vation significative de la densit6 optique (DO) et du titre (>I 4 fois) des IgG en ELISA; mais des r6sultats n6gatifs ou linfites ont 6td not6s en ELISA au 11".jour apr6s le d6but des symptfmes. Des anticorps IgM ont 6t6 d6cel6s pendant la premiere semaine de la symptomatologie, et l'on observe des valeurs maximales de la DO durant les quatre premieres semaines; ces valeurs diminuent au cours des 3-9 mois. Tousles s6rums recueillis apr~s le 14e mois se sont montr~s n6gatifs en ELISA pour les IgM. La r~ponse anticorps pr~coce et l'int~r& d'un diagnostic sur un ~chantillon unique de s&um donnent /l I'ELISA lgM un avantage certain sur I'ELISA IgG. Le dosage des IgG par ELISA, compar6 au test de r6duct';on/neutralisation en plaque (PRNT), a 6t~ utile pour l'6tude s~ro-6pid6miologique d'une population normale de l'ile de Chypre. Sur les 183 s~rums examines, 14 (19 °/0) 6taient positifs en PRNT et 113 (62 %) l'~taient en ELISA. Certains s6rums PRNT n~gatifs 6taient fortement positifs en ELISA IgG ainsi qu'en irnmunofluorescence indirecte. Ces observations sugg&ent la presence/t Chypre d'un virus apparent6 au virus sicilien de la fi~vre/l phl~botome, mais qui n'a pas encore ~t~ isol_~. Mots-clds: Fi~vre ~ phl6botomes, Arbovirus, Phi~bovirus; IgG, IgM, ELISA, Virus sicilien, Chypre.
References
Doerf, R., Frank, K. & Taussig, S. (1909), Das Pappatacifieber, ein epidemisches Drei-Tage-Fieber im Adriatischen Kiistengebiete Oesterreich-Ungarns. Wien. Kiln. Wschr., 22, 609-610. Earley, E., Peralta, P.H. & Johnson, K.M. (1967), A plaque neutralization method for arboviruses. Proc. Soc. exp. Med., 125, 741-747. Eitrem, R., Niklasso~, B. & Weiland~ O. (!991)7 Sandfiy fever among Sw, dish tourists. Scand. J. infect. Dis. (in press). Eitrem, R., Vene, S. & Niklasson, B. (1990), Incidence of sandflv fever among, Swedish United Nations soldiers on Cyprus during 1985.Amer. J. trop. Med. Hyg., 43, 207-211. Hertig, M. & Sabin, A.B. 0964), Sandfly fever (Pappata-
394
R. E I T R E M E T A L .
ci, Phlebotomus, three-day-fever), in "Preventive medicine in World War II: volume 7" (Coates, J.B.) (pp. 109-174). Communicable diseases. US Government Printing Office, Washington, DC. Karabatsos, N. (1985), International catalogue of arboviruses including certain other viruses of vertebrates. Third ed. American Society of Tropical Medicine & Hygiene. San Antonio Texas. (Naples pp. 893-894; Sicilian pp. 895-896; Toscana pp. 1027-1028). Leoncini, F., Bartolozzi, D., Banchi, S., Balducci, M., Fratiglioni, L., CacioUi, S. & Rence, A. (1986), Toscana virus : a new phlebovirus, aetiological agent of acute inflammatory disease of the human central nervous system. G. Mal. infett., 38, 649-52. McCown, J.M., Brandt, W.E., Bancroft, W.H. & Russell, P.K. (1979), Dimethyl sulfoxide enhancement of Phiebotomus fever virus plaque formation. Amer. J. trop. Med. Hyg., 28, 733-739.
Niklasson, B. & Eitrem, R. (1985), Sandfly fever among Swedish UN troops in Cyprus. Lancet, I, 1212. Peters, C.J. & LeDuc, J.W. (1984), Bunyaviruses, phleboviruses and related viruses, in "Textbook of human virology" (Belshe, R.B.) (pp. 572-576). PSG Publ. Co., Littleton, MA. Pick, A. 0986), Zur Pathologie und Therapie einer eigentumlischen endemischen Krankheitsform. Wien. Med. Wschr., 36, 1141-45 and 1168-71. Sabin, A.B. (1951), Experimental studies on Phlebotomus (pappataci, sandfly) fever during World War II. Arch. ges. Virusforsch., 4, 367-410. Sundqvist, V. & Wahren, B. (1981), An interchangeable ELISA for cytomegalovirus antigen and antibody. J. Virol. Methods, 2, 301-312. Verani, P., Nicoletti, L. & Ciufoloni, M.G. (1984), Antigenic and biological characterization of Toscana virus, a new Phlebotomus fever group virus isolated in Italy. Acta. Virol., 28, 39-47.
