Veterinary Microbiology, 28 ( 1 9 9 1 ) 121-127 Elsevier Science Publishers B.V., A m s t e r d a m
121
Seroepidemiology of Bivens Arm virus infections of cattle in Florida, St Croix and Puerto Rico. T. Tuekam, E.C. Greiner, and E.P.J. Gibbs* Department of lnfectious Diseases, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610, USA (Accepted 10 January 1991 )
ABSTRACT Tuekam, T., Greiner, E.C. and Gibbs, E.P.J., 1991. Seroepidemiology of Bivens Arm virus infections of cattle in Florida, St Croix and Puerto Rico. Vet. Microbiol., 28:121-127. Bivens Arm virus (BAV) is a newly discovered rhabdovirus infecting cattle and water buffalo in Florida. The virus is classified as a member of the Tibrogargan group, members of which have hitherto been found only in Australasia. They are considered to be transmitted by Culicoides species. Bivens Arm virus was first isolated from Culicoides insights which suggests that BAV is also transmitted by this genus. A serological survey of two small groups of cattle raised in St. Croix and Puerto Rico, in the Caribbean, established that antibody to BAV, or a closely related virus, exists on both islands. A retrospective analysis of seroconversions to BAV in sentinel calves in Florida, relative to populations of potential Culicoides vectors, failed to demonstrate any statistically significanl correlation.
INTRODUCTION
During field studies in Florida in 1982 on the transmission of bluetongue virus among ruminants, an isolate of a new rhabdovirus was obtained from Culicoides insignis trapped in the vicinity of water buffalo (Bubalus bubalis) that had recently been imported from Trinidad (Gibbs et al., 1989). The isolate was named Bivens Arm and a close serological relationship was demonstrated between it and an Australian virus called Tibrogargan, which is also known to infect water buffalo (Cybinski et al., 1980; Gibbs et al., 1989; Calisher et al., 1989)**. In the 1970s, water buffalo had been imported on several occasions by air from Australia to Trinidad en route to South America (Borde, G.N., unpublished observations, 1988 ). While at the time of the isolation o f Bivens Arm virus (BAV) in Florida there was no evidence that these *Author to w h o m correspondence should be sent **In arbovirology, n a m e s o f nearby geographical features are frequently used to identify isolates; Bivens A r m is a small lake b o r d e r i n g the field where the buffalo were pastured and Tibrogargan is a peak within the Glasshouse range of m o u n t a i n s n o r t h o f Brisbane.
0378-1135/91/$03.50
© 1991-
Elsevier Science Publishers B.V.
122
T. T U E K A M ET AL.
Australian consignments of water buffalo were ever infected with Tibrogargan virus, the close association of both BAV and Tibrogargan viruses with water buffalo initially fueled speculation that the water buffalo were responsible for the introduction of the virus to Florida. However, a screen of sera, collected from Florida cattle prior to the importation in 1981, revealed that antibodies to BAV were present before this date, thus negating this hypothesis (Gibbs et al., 1989). Although neither Tibrogargan virus nor BAV has ever been associated with clinical disease, other vector-borne rhabdoviruses are important veterinary pathogens (eg. vesicular stomatitis and bovine ephemeral fever viruses ). Thus, we considered the economic importance of the rhabdoviridae justified additional study of the geographical distribution of BAV infections of cattle and the arthropods potentially involved in virus transmission. In this paper, we report the presence of antibody to BAV in cattle sera collected on the Caribbean islands of Puerto Rico and St Croix, and an examination of the seasonal association of seroconversion to BAV in Florida cattle with Culicoides populations.
MATERIALS AND METHODS
Geographical survey for serum antibody to Bivens Arm virus Sera. All cattle sera were available from collections stored at - 2 0 °C in the virology laboratory. Those originating from Puerto Rico and St Croix had been collected in connection with epidemiological studies on bluetongue in 1982 (Gibbs and Greiner, 1983a; Greiner et al., 1983). The sera were from animals of different ages and were heat inactivated by heating at 56 °C for 30 minutes before testing. Virus neutralization test. Sera were screened in Vero cells for neutralizing antibody to BAV by a conventional microneutralization test using approximately 100 tissue culture infective doses 50% of virus against varying dilutions of sera. The sera were initially screened for neutralizing antibody at 1 : 20 and later tested for determination of end-points by serial twofold dilutions. The titer was recorded as the reciprocal of the highest serum dilution that completely neutralized the virus challenge. Hyperimmune ascitic fluid produced in mice to the prototype virus was used as a known reference antibody in each test run (Gibbs et al., 1989). An antibody titer of 20 was considered positive.
SEROEP|DEMIOLOGY OF BIVENS ARM VIRUS
123
Potential Culicoides association with seasonal seroconversion o f cattle to Bivens Arm virus
Seroconversion in sentinel cattle herds in Florida. Viruses within the Tibrogargan virus group are considered to be biologically transmitted by Culicoides species. Considering that BAV was isolated from Culicoides, we considered it appropriate to focus our attention only on this genus of potential vectors. To indicate which species of Culicoides may be transmitting BAV in Florida, we examined groups of calves for seronconversion to the virus and related the patterns of seroconversion to the seasonal activity of Culicoides species at that site. The sera selected for examination were from groups of 20 calves sampled between 1981 and 1985 at monthly intervals during sentinel animal studies on the epidemiology of bluetongue virus in Florida. The sentinel calves were in herds located at four sites in South (Belle Glade ), South West (Ona), Mid- (Brooksville) and North-Central (Bell) Florida. The operation of the sentinel herd system has been described elsewhere (Gibbs et al., 1983b). The sera had been stored at - 20°C in the intervening period since collection; they were examined for antibody to BAV as described above.
Culicoides populations at the sentinel herds. The seasonal populations of Culicoides species at each of the sentinel herds were obtained from previously recorded data. The methodology used for trapping and identifying the Culicoides has been described elsewhere (Kramer et al, 1985; Greiner et al, 1985). Insect populations were based on collections made every two weeks throughout the year. The statistical correlation between the populations of Culicoides species detected at the sentinel herd site and the percentage of calves seroconverting in any one month was determined by Pearson's coefficient of determination (Ronald, 1982). This coefficient (r 2) measures the linear relationship between two factors (x and y). In this context, the number of insects trapped (i.e. of the same species) is represented by x and the percentage of seroconversion in the calves to BAV by y. Since we have not conducted any experimental infections of either cattle or water buffalo with BAV, we are uncertain of the precise time lag between infection and the onset of detectable serum antibody by the micro-neutralization test. For the purpose of this analysis, we have assumed that an animal exposed to a virus for the first time develops a humoral response within 5 to 20 days (Tyler and Cullor, 1989), thus seroconversions observed on sera collected each month can be attributed to the insect populations trapped during the previous month. An r 2 value of 0.5 was considered significant. Seroconversions in sentinel cattle were correlated to the insect populations of the same year.
124
7. TUEKAM ET AL.
RESULTS
Geographical survey for serum antibody to BA V. Serum antibody to BAV was present in the cattle sera collected from Puerto Rico ( 16/40, 40% positive) and St Croix ( 15/44, 34% positive). In general, the detectable antibody titers were low, ie. 20 or 40. Seasonal seroconversion in sentinel cattle to BA V in Florida. Serum antibody to BAV was detected in some calves at the first sampling at all four sentinel sites, at which time the calves were approximately three months of age (Table 1 ). At the sentinel sites at Ona and Bell in 1981, the majority of calves subsequently became seropositive. The number of calves seroconverting to BAV at Ona increased progressively from March to August, by which time 90% were positive. At Bell, the majority of the calves seroconverted within a shorter period from February to May. Although antibody to BAV was detected in cattle sampled in 1985 at the Brooksville and Belle Glade sentinel sites, no clear-cut patterns of seroconversion occurred and less than 35% of the calves ever had pre-existing antibody or seroconverted during the monitoring period. The antibody titers detected in the sera of the calves never exceeded 40 at any site. The relationship o f Culicoides populations to seroconversion to BA V. Only at the sentinel site at Bell was there any apparent temporal relationship seen between Culicoides populations and seroconversion of the calves to BAV. The seroconversions detected in the calves are presented in Fig. 1 in relation to the populations of Culicoides variipennis and Culieoides insignis
Z ~o I 0 0 0 ©
Sera ~ 1/20
a
/"
................................... 11
-20 -18
~
~ ,
-16
~
-14 //
© _~b :~
~ ~,~
~k,~///// / / / / / /
~o ©
121
Seroepidemiology of Bivens Arm virus infections of cattle in Florida, St Croix and Puerto Rico. T. Tuekam, E.C. Greiner, and E.P.J. Gibbs* Department of lnfectious Diseases, College of Veterinary Medicine, University of Florida, Gainesville, FL 32610, USA (Accepted 10 January 1991 )
ABSTRACT Tuekam, T., Greiner, E.C. and Gibbs, E.P.J., 1991. Seroepidemiology of Bivens Arm virus infections of cattle in Florida, St Croix and Puerto Rico. Vet. Microbiol., 28:121-127. Bivens Arm virus (BAV) is a newly discovered rhabdovirus infecting cattle and water buffalo in Florida. The virus is classified as a member of the Tibrogargan group, members of which have hitherto been found only in Australasia. They are considered to be transmitted by Culicoides species. Bivens Arm virus was first isolated from Culicoides insights which suggests that BAV is also transmitted by this genus. A serological survey of two small groups of cattle raised in St. Croix and Puerto Rico, in the Caribbean, established that antibody to BAV, or a closely related virus, exists on both islands. A retrospective analysis of seroconversions to BAV in sentinel calves in Florida, relative to populations of potential Culicoides vectors, failed to demonstrate any statistically significanl correlation.
INTRODUCTION
During field studies in Florida in 1982 on the transmission of bluetongue virus among ruminants, an isolate of a new rhabdovirus was obtained from Culicoides insignis trapped in the vicinity of water buffalo (Bubalus bubalis) that had recently been imported from Trinidad (Gibbs et al., 1989). The isolate was named Bivens Arm and a close serological relationship was demonstrated between it and an Australian virus called Tibrogargan, which is also known to infect water buffalo (Cybinski et al., 1980; Gibbs et al., 1989; Calisher et al., 1989)**. In the 1970s, water buffalo had been imported on several occasions by air from Australia to Trinidad en route to South America (Borde, G.N., unpublished observations, 1988 ). While at the time of the isolation o f Bivens Arm virus (BAV) in Florida there was no evidence that these *Author to w h o m correspondence should be sent **In arbovirology, n a m e s o f nearby geographical features are frequently used to identify isolates; Bivens A r m is a small lake b o r d e r i n g the field where the buffalo were pastured and Tibrogargan is a peak within the Glasshouse range of m o u n t a i n s n o r t h o f Brisbane.
0378-1135/91/$03.50
© 1991-
Elsevier Science Publishers B.V.
122
T. T U E K A M ET AL.
Australian consignments of water buffalo were ever infected with Tibrogargan virus, the close association of both BAV and Tibrogargan viruses with water buffalo initially fueled speculation that the water buffalo were responsible for the introduction of the virus to Florida. However, a screen of sera, collected from Florida cattle prior to the importation in 1981, revealed that antibodies to BAV were present before this date, thus negating this hypothesis (Gibbs et al., 1989). Although neither Tibrogargan virus nor BAV has ever been associated with clinical disease, other vector-borne rhabdoviruses are important veterinary pathogens (eg. vesicular stomatitis and bovine ephemeral fever viruses ). Thus, we considered the economic importance of the rhabdoviridae justified additional study of the geographical distribution of BAV infections of cattle and the arthropods potentially involved in virus transmission. In this paper, we report the presence of antibody to BAV in cattle sera collected on the Caribbean islands of Puerto Rico and St Croix, and an examination of the seasonal association of seroconversion to BAV in Florida cattle with Culicoides populations.
MATERIALS AND METHODS
Geographical survey for serum antibody to Bivens Arm virus Sera. All cattle sera were available from collections stored at - 2 0 °C in the virology laboratory. Those originating from Puerto Rico and St Croix had been collected in connection with epidemiological studies on bluetongue in 1982 (Gibbs and Greiner, 1983a; Greiner et al., 1983). The sera were from animals of different ages and were heat inactivated by heating at 56 °C for 30 minutes before testing. Virus neutralization test. Sera were screened in Vero cells for neutralizing antibody to BAV by a conventional microneutralization test using approximately 100 tissue culture infective doses 50% of virus against varying dilutions of sera. The sera were initially screened for neutralizing antibody at 1 : 20 and later tested for determination of end-points by serial twofold dilutions. The titer was recorded as the reciprocal of the highest serum dilution that completely neutralized the virus challenge. Hyperimmune ascitic fluid produced in mice to the prototype virus was used as a known reference antibody in each test run (Gibbs et al., 1989). An antibody titer of 20 was considered positive.
SEROEP|DEMIOLOGY OF BIVENS ARM VIRUS
123
Potential Culicoides association with seasonal seroconversion o f cattle to Bivens Arm virus
Seroconversion in sentinel cattle herds in Florida. Viruses within the Tibrogargan virus group are considered to be biologically transmitted by Culicoides species. Considering that BAV was isolated from Culicoides, we considered it appropriate to focus our attention only on this genus of potential vectors. To indicate which species of Culicoides may be transmitting BAV in Florida, we examined groups of calves for seronconversion to the virus and related the patterns of seroconversion to the seasonal activity of Culicoides species at that site. The sera selected for examination were from groups of 20 calves sampled between 1981 and 1985 at monthly intervals during sentinel animal studies on the epidemiology of bluetongue virus in Florida. The sentinel calves were in herds located at four sites in South (Belle Glade ), South West (Ona), Mid- (Brooksville) and North-Central (Bell) Florida. The operation of the sentinel herd system has been described elsewhere (Gibbs et al., 1983b). The sera had been stored at - 20°C in the intervening period since collection; they were examined for antibody to BAV as described above.
Culicoides populations at the sentinel herds. The seasonal populations of Culicoides species at each of the sentinel herds were obtained from previously recorded data. The methodology used for trapping and identifying the Culicoides has been described elsewhere (Kramer et al, 1985; Greiner et al, 1985). Insect populations were based on collections made every two weeks throughout the year. The statistical correlation between the populations of Culicoides species detected at the sentinel herd site and the percentage of calves seroconverting in any one month was determined by Pearson's coefficient of determination (Ronald, 1982). This coefficient (r 2) measures the linear relationship between two factors (x and y). In this context, the number of insects trapped (i.e. of the same species) is represented by x and the percentage of seroconversion in the calves to BAV by y. Since we have not conducted any experimental infections of either cattle or water buffalo with BAV, we are uncertain of the precise time lag between infection and the onset of detectable serum antibody by the micro-neutralization test. For the purpose of this analysis, we have assumed that an animal exposed to a virus for the first time develops a humoral response within 5 to 20 days (Tyler and Cullor, 1989), thus seroconversions observed on sera collected each month can be attributed to the insect populations trapped during the previous month. An r 2 value of 0.5 was considered significant. Seroconversions in sentinel cattle were correlated to the insect populations of the same year.
124
7. TUEKAM ET AL.
RESULTS
Geographical survey for serum antibody to BA V. Serum antibody to BAV was present in the cattle sera collected from Puerto Rico ( 16/40, 40% positive) and St Croix ( 15/44, 34% positive). In general, the detectable antibody titers were low, ie. 20 or 40. Seasonal seroconversion in sentinel cattle to BA V in Florida. Serum antibody to BAV was detected in some calves at the first sampling at all four sentinel sites, at which time the calves were approximately three months of age (Table 1 ). At the sentinel sites at Ona and Bell in 1981, the majority of calves subsequently became seropositive. The number of calves seroconverting to BAV at Ona increased progressively from March to August, by which time 90% were positive. At Bell, the majority of the calves seroconverted within a shorter period from February to May. Although antibody to BAV was detected in cattle sampled in 1985 at the Brooksville and Belle Glade sentinel sites, no clear-cut patterns of seroconversion occurred and less than 35% of the calves ever had pre-existing antibody or seroconverted during the monitoring period. The antibody titers detected in the sera of the calves never exceeded 40 at any site. The relationship o f Culicoides populations to seroconversion to BA V. Only at the sentinel site at Bell was there any apparent temporal relationship seen between Culicoides populations and seroconversion of the calves to BAV. The seroconversions detected in the calves are presented in Fig. 1 in relation to the populations of Culicoides variipennis and Culieoides insignis
Z ~o I 0 0 0 ©
Sera ~ 1/20
a
/"
................................... 11
-20 -18
~
~ ,
-16
~
-14 //
© _~b :~
~ ~,~
~k,~///// / / / / / /
~o ©
