Susceptibility of dogs to West Nile virus: A survey and pathogenicity trial

Susceptibility of dogs to West Nile virus: A survey and pathogenicity trial

J. Comp. Path. 1989 Vol. Susceptibility 100 of Dogs to West and Pathogenicity N. K. Blackburn*, Nile Virus: Trial F. Reyerst, W. L. Berry? A. ...

522KB Sizes 0 Downloads 15 Views

J. Comp.


1989 Vol.



of Dogs to West and Pathogenicity

N. K. Blackburn*,

Nile Virus: Trial

F. Reyerst, W. L. Berry? A. J. Shepherd*

A Survey and

*National Institute for Virology and Department of Virology, Universip of the Witwatersrand, Private Bag X4, Sandringham 2131, Republic of South Africa, and f Faculty of Veterinary Science, University of Pretoria, Private Bag X04, Onderstepoort 0110, Republic of South Africa Summary A serological survey of dogs from the highveld region of South Africa showed that 37 per cent (138 of 377) had neutralizing antibodies to West Nile (WN) virus and only 2.7 per cent (10 of 377) had antibodies to Sindbis virus. WN virus was isolated from one of the WN-antibody negative sera. Because these results suggested that dogs may play an important part in the epidemiology of WN virus, a pathogenicity trial was carried out. Two of three dogs infected with WN virus had a mild recurrent myopathy, but no other abnormalities were detected in the biochemical or haematological tests performed on any of the dogs. All three dogs developed antibodies but a low titre-viraemia was detected in only one dog. It was concluded that dogs do not play an important part in the epidemiology of WN virus but they may play a small part in the maintenance of the virus. Introduction

West Nile (WN) virus and, to a lesser extent, Sindbis (SIN) virus are the most common causes of arthropod-borne virus diseases of man in the highveld of South Africa (McIntosh, Dickinson, Serafini and De Sousa, 1962). Jupp and McIntosh (1970) established that Culex univittatus was the only important vector of these viruses in the region. This mosquito is primarily a bird feeder but in a study of blood meals, over a third of the C. univittatus tested had fed on mammals and a high proportion of these on dogs (Anderson, 1967). With this relationship between the main vector of WN and SIN viruses and dogs it was considered important to establish the susceptibility of dogs to these viruses. This article presents the results of a serological survey and a small pathogenicity trial on dogs experimentally infected with WN virus. Materials



j9 em

The first group consisted of 238 dog sera collected from various rural and peri-urban localities in the eastern and north-eastern Orange Free State (OFS), a highveld region (altitude circa 1500 to 2000m) in central South Africa. These sera were originally submitted to the laboratory for plague surveillance studies. The second group consisted of 139 dog sera from urban and peri-urban areas north of the city of Pretoria and were 002 I 9975/89/O

10059 + 08 $03.00/O

IC! 1989 Academic

Press Limited


N. K. Blackburn

et al.

obtained from patients admitted to the Faculty of‘ Veterinary Medicine, Universitv of Pretoria, Onderstepoort. Several of these patients had been diagnosed to have bec~ii infected with one of the following tick-borne agents: Babesin ~ani.\. Ehrlichicr caui.\ OI HPpato:oon cani.\. I’irus .rtrain.s The virus strains used in the serological tests were virus strain H442, Wesselsbron (WSL) virus strain The WN virus strain used to infect the dogs for the mousr passage 2, tissue culture passage 2 of a WN Haemagglutination


the South African WN prototype H177 and SIN virus, strain AR96. pathogenicity tests was OYI 174.5. virus isolated during this study

test (HI)

This test was based on the method described by Clarke and Casals (1958) which uses sucrose acetone-extracted virus antigen. Four haemagglutination units of antigen were used in all tests and 0.025ml vol in a microtiter system. The sera were acetoneextracted to remove nonspecific inhibitors and dilutions were made from 1 in 20 to 1 in 10240. The tests were incubated at 4°C overnight and read 1 h after the addition of the gander cells. End-points were recorded as the highest dilution in which there was complete inhibition of agglutination. All the sera were tested against WN, SIN and WSL antigens. .NeuLralization



The NT test was carried out as described by Blackburn and Swanepoel (1980). The sera were diluted 1 in 8 to 1 in 1024 and lOOTCD,, of virus were added. After 1 h at 23”C, 4 x lo4 Vero cells were added to each well and the microtitre plates were incubated at 37°C. The results were finally read after 6 days incubation, the end-point being the highest dilution of serum showing no cytopathic effect. Where possible, all sera positive for WN and SIN antibodies by the HI test were tested by the NT test. Firus isolation Sera from the OFS which were negative by the HI test were inoculated intracerebrally (ic.) into a litter of 2- to 3-day-old mice. The mice were examined daily for 21 days, any dead or moribund mice were removed and tested for the presence of virus. Any isolations were confirmed as WN virus by NT test against WN prototype virus H442 mouse ascitic fluid. Pathogenicity


Experimental design, As there is no well-documented literature on the possible functional alterations which WN virus could cause in the dog, it was decided to include a number of haematological and biochemical tests in the design of the trial. The tests would monitor general inflammatory reactions, stress reactions, hepatopathy, nephropathy, myopathy and alteration in serum proteins as well as abnormalities in thrombocyte and erythrocyte kinetics. The tests were made daily for a period of 15 days from the day before infection with days - 1 and 0 considered as normal baselines. The measurements made over the 15-day period were compared with published normal values as well as the baselines established in the first two days. The WN virus infected group was also compared with the non-infected control animal. Experimental dogs. Four animals were selected which had no detectable antibodies to WN or WSL viruses. Three were infected with WN virus strain OM1745, 0.5ml of 1 x lo7 virus per ml each by the subcutaneous and intravenous routes, while the control dog was given 0.5ml of virus-free diluent by each of the same routes. The dogs were

West Nile Virus

in Dogs


bled on the day before infection and daily for 14 days, then at weeks 3 and 4. The blood was collected in potassium-EDTA for virus isolation and haematology; serum from clotted samples was collected for serological and biochemical tests. Clinical observations. All dogs were examined daily for demeanour, temperature, pulse, respiratory rate, appetite, water intake and general clinical status. Additionally, as WN virus is reported to be involved in central nervous system pathology in man (Pruzanski and Altman, 1962) and in the horse (Guillon, Oudar, Joubert and Hannoun, 1968), each dog was given a daily clinical neurological examination. Haematological 1. 2. 3. 4. 5. 6. 7.


tests. The following

Haemoglobin concentration (Hb): Coulter haemoglobinometer* with a 1 in 500 dilution in Isoton II* with Zaponin* as haemolysin. Erythrocyte count (RCC): Coulter model Fn electronic cell counter* with a I in 50000 dilution in Isotin II*. Haematocrit (Ht): as per RCC. Mean corpuscular volume (MCV) : as per RCC. Mean corpuscular haemoglobin content (MCHC): by calculation from Hb and Ht. Leucocyte count (WCC): Coulter model Fn electronic cell counter* with a 1 in 500 dilution in Isotin II*. Differential WCC (Diff. Count): microscopic enumeration and classification of 100 leucocytes in a CAM-Quick7 stained thin blood film by the Battlement technique {Lynch, Raphael, Mellor, Spare and Inwood, 1976) and classification into neutrophils, immature neutrophils, lymphocytes, monocytes and eosinophils. Thrombocyte count (Thr. C): Sysmex platelet counter model PL 110: with a 1 in 5 dilution of whole-EDTA blood in Cell Kit CD diluentf, centrifuging 4 minutes in a Sysmex Platelet Centrifuge (2000rpm)S with a further dilution of the supernatant, platelet-rich fraction at a 1 in 100 dilution in Cell-ent diluentf.

Biochemical Tests. All biochemical spectrophotometer, with a Setpoint 1. 2. 3. 4. 5. 6.


tets were conducted:

tests were conducted calibrator§.

on an RA 1000 automated

Total serum proteins (TSP): biuret end-point determination at 500nm, calibrator 64g per I, Technicon method TOl-1301-02. Albumin (ALB): bromo-cresyl-green end-point determination at 600nm, calibrator 36g per 1, Technicon method TOI-1377-02. Globulin (GLOB): by calculation from TSP minus ALB. Urea: urease, first order inverse rate reaction at 340 nm, calibrator 13.6 mmol per 1. Technicon method TOl-1821-56. Creatinine: alkaline picrate, first order rate reaction at 500nm, calibrator 778pmol per 1, Technicon method TOl-1304-53. i\lanine aminotransferase (E.C. (ALT‘I: pyruvate substrate zero order kinetic reaction at 340nm, reaction at 37°C. results expressed as at 25”(:, Trchnicon method TO1 - 1756-O 1. Alkaline phosphatase (E.C. (ALP): para-nitro phenol phosphate substrate, diethanolamine buffer zero order kinetic reaction at 405nm, reaction at 37”C, results expressed as at 25”C, Technicon method TOl-181 l-01.

*Coulvr Uxtronics I Pty) Ltd, Hi&ah, Florida, t C. A. Milsch (Pty) Ltd, Krugersdorp, R.S.A. $ TOA Medical Electronic Co. Ltd, Japan. $Terhnicon Instrument Corp., Tarrytown, NY.



62 8.

N. K. Blackburn

Creatine kinase (E.C. reaction at 340nm, reaction method TOl-1529-01.

et al.

(CK): creatine phosphate substrate zero order rate at 37”C, results expressed as at 25”C, Technicon


Of the 377 dog sera tested for arbovirus HI antibodies, 46 per cent (174 of 377) had WN HI titres of 1 in 20 or greater (Fig. 1)) 14.4 per cent (25 of 174) of which had titres in excess of 1 in 640. One hundred and sixty-two of the 174 WN HI positive sera were tested for WN NT antibodies; 85 per cent ( 138 of 162) had titres of 1 in 8 or greater (Fig. 1) and 22.5 per cent (31 of 138) of the titres were higher than 1 in 128, the higher titres being almost exclusively in the dog sera from the OFS. Fifteen per cent (57 of 377) of the sera had HI antibodies against WSL virus, but all of these sera had higher titres of HI and NT antibodies to WN. Ten per cent (36 of 377) had HI antibodies against SIN virus, 28 per cent (10 of 36) of which had SIN NT antibodies, all of which were lessthan 1 in 64. WN virus was isolated from one of 1lOWN HI negative j_< 1 in 20) dog sera. Pathogenicity None of the WN virus infected dogs exhibited any detectable clinical or neurological abnormality during the 14 days post-infection. With the exception of CK, the infected group did not differ significantly from published normal ranges (Tasker, 1978) for the haematological or biochemical measurements made, from baseline values established on days - 1 and 0 of the trial nor from the range of values observed in the non-infected


$20 640 ,280 2500 51 HI antibody titres



Reciprocal dag sera.






NT antibody






in 377

West Nile Virus

$ a 3


in Dogs


5 E 2 d




1 0



Serum trianglrs.

creating dog


kinase I; squares,


(CK) dog



activity 2; open














in experimental triangles, dog

dogs 3; cirrlcs,

in a West Nile control dog 4.




control dog. Two of the 3 dogs infected with WN virus (dogs 2 and 3) developed recurrent abnormally high peaks of CK activity during the 14 days after infection (see Fig. 2). One of the 3 dogs infected with WN virus developed a slight viraemia which was detectable from day 1 to day 7 (Table 1). All the infected dogs developed HI and NT antibodies against WN virus (Table l), the highest antibody titres occurring in the viraemic dog. Discussion It was established by McIntosh, Jupp, Dickinson, McGillivray and Sweetnam (1967) that C&x uniuittatus was the primary vector of WN and SIN viruses in the highveld region of South Africa, a region which includes the OFS and Pretoria. Anderson’s (1967) study of the host preferences of several species of mosquito including Cx. univittatus showed that 37 per cent (71 of 191) had fed on mammalian blood and of these 37 per cent (26 of 71) had fed on dogs. With West Nile virus being extremely prevalent in the highveld region (McIntosh et al., 1962), it was not surprising that WN NT antibodies should be found in such a high proportion of dogs, 37 per cent (138 of 377). The number of dogs with antibodies to SlN virus was no lessthan that found in other animals in the region. McIntosh et al. (1962) showed that O-9 per cent (9 of 1019j sheep, cattle and goats had NT antibodies to SIN virus as compared to 2.7 per cent (10 of 377) of the dogs in this study. The low infective rate and relatively low NT titres indicated that dogs do not play a part in the epidemiology of SIN virus.


N. K. Blackburn Table


inhibition infected

(HI), neutralization with log 10’ per


et al. 1 (NT) and West Nile

Do,< number I Dq





viraemia virus on

(V) test day 1



3 dogs

and test

2 .vr



3 .27



I 2 3 4 5 6 7 8 9 10 II 12 13 14 21 28 ’ Reciprocal




-40 80 80 80 80 160 80 160 160 160 160 160 80


40’ 160 320 160 320 320 320 320 320 320 HI titre;

8’ 8 8 16 32 64 64 32 32 64 32 64 ‘rrciprocal



-. 8 16 16 32 16 32 16 16 16 16 32 32 64


40 40 80 160 640 640 1280 1280

8 16 16 16 I6 16 16 32 64 64 128 128 256

2.8’ 1.5 2.4 1.7 I.7 1.5 1.5

‘Lo,qli, prr ml \VN virus

The isolation of WN virus from the serum of one of the dogs and the high antibody titres indicated that dogs may play a part in the maintenance of WN virus. This view was supported by the isolation of WN virus from a Huemuphysalis leachii tick taken from a dog in the OFS during studies on the epidemiology of Crimean-Congo haemorrhagic fever virus by one of the authors (A.J.S.). It was decided to carry out a pathogenicity trial with WN susceptible dogs to try to clarify the situation. Because of the high prevalence of WN antibodies in dogs and the relatively short supply of dogs for research purposes, only four dogs were available for this study. The lack of abnormality in 20 of the 21 measurements assessed in the trial suggests that WN virus does not cause major systemic inflammatory or stress reactions (WCC, Difl), thrombocyte kinetic changes (Thr.C), erythrocyte kinetic changes (RCC, Hb, Ht, MCV, MCHC), serum protein alterations (TSP, ALB, GLOB), hepatopathy or bile stasis (ALT, ALP) nor renal glomerular filtration rate changes (urea, creatinine). However, the pattern of‘ peaks observed in CK activity in 2 of the 3 infected dogs suggests a mild recurrent myopathy which does not appear to be associated with the administration of the inoculum vehicle as the control dog was given virus-free diluent on day 0, nor with an exercise-induced release as the dogs were confined. CK has a very short in vivo half life (Schmidt and Schmidt, 1976), of approximately 5 h. Consequently, the fluctuating pattern observed implies an intermittent release of the enzyme. This enzyme is found almost exclusively in the cytosol of muscle cells (Adolph and Lorenz, 1982) and thus the observed

West Nile Virus

in Dogs


pattern is highly suggestive of a myopathy. The enzyme also occurs in the central nervous system but will enter the plasma only if there is a severe breakdown of the blood-brain barrier (Adolph and Lorenz, 1982). In the absence of any abnormal clinical signs suggesting a central nervous system involvement, it can be concluded that the observed alteration in CK activity was not related to pathology of the central nervous system. Some 3 weeks after the trial, dog 3 was found to be clinically abnormal and, upon further investigation and subsequent necropsy, was found to be suffering from primary adreno-cortical hyperplasia (Canine Cushing’s disease). This dog had exhibited high-normal to abnormal ALP activity, low-normal urea and creatinine concentrations and an intermittent eosinopenia during the trial. These findings are all consistent with Cushing’s disease (Lorenz, 1982) and suggest that dog 3 was, at the time of the trial, in the early stages of this disease. This disease is typified by high blood cortisol concentrations and a resultant immune suppression (Lorenz, 1982). It is, therefore, of interest to note that this particular dog was the one which exhibited a viraemia and one of the two dogs thought to have developed a virus-induced myopathy. The maximal virus titre in the viraemic dog detected on day 1 was probably due in part to the original inoculum which had a particularly high titre. It was a little surprising though, that no virus was detected in the two other dogs at this early stage. The viraemia, although very low, would be sufficient to infect a small proportion of Cx univittatus feeding on the dog, as their 50 per cent infectivity threshold for WN virus is log,,2.1 per ml (Jupp, 1976). Serological titres were very similar in dogs 1 and 2 which had no detectable viraemia, whereas dog 3, with viraemia, had antibody titres at least 4-fold higher in the NT and HI tests. Only a small proportion of the survey dogs had HI and NT antibody titres equal to or higher than those of dog 3, namely 6.6 per cent (‘25 in 377) and 8.2 per cent (31 in 377), respectively. Even if all these high titres are the result of viraemia, which is highly unlikely, it can be assumed from these results that dogs do not play a major role in the epidemiology of West Nile virus. The dog may be incidentally involved in the maintenance of the virus through the mosquito or tick. The one viraemic dog which had Cushing’s disease suggests that the immune status of a dog at the time of infection may be important.

References Anderson, D. (1967). Ecological studies on Sindbis and West Nile viruses in South Africa. III. Host preferences of mosquitoes as determined by the precipitin test. South African Journal of Medical Science, 32, 34-39. Adolph, L. and Lorenz, R. (1982). Enzyme Diagnosis in Diseases of the Heart, Liver and Pancreas. S. Karber, Basel. of flavivirus infections Blackburn, N. K. and Swanepoel, R. (1980). A n investigation of cattle in Zimbabwe/Rhodesia with particular reference to Wesselsbron virus. Journal of Hygiene, 85, 1-33. Clarke, D. H. and Casals, J. (1958). Techniques for haemagglutination and haemagglutination inhibition with arthropod-borne viruses. American Journal of Tropical Medicine and Hygiene, 7, 561-573.


N. K. Blackburn

et al.

C. (1968). Histological lesions of Guillon, J. C., Oudar, J, Joubert, L. and Hannoun, the nervous system in West Nile infection in the horse. Annales de f’fnstitut Pasteur, 114, 539-550. Jupp, P. G. (1976). The susceptibility of four South African species of Culex to West Nile and Sindbis viruses by two different infecting methods. Mosquito News, 36, 166-173. Jupp, P. G. and McIntosh, B. M. (1970). Q uantitative experiments on the vector caoabilitv of Culex fculexi univittatus Theobald with West Nile and Sindbis viruses. Joirnal oj Medical kntomkogy, 7, 371-373. Lynch, M. J., Raphael, S. S., Mellor, L. D., Spare, P. D. and Inwood, M. J. H. (1976). Medical Laboratory Technology, volume II, 3rd Edit. W.B. Saunders, Philadelphia. Lorenz, M. D. (1982). Diagnosis and medical management of canine Cushing’s disease: a study of 57 consecutive cases. Journal of the Amerkan Animal Hospital Association, 18, 707-7 16. McIntosh, B. M., Dickinson, D. B., Serafini, E. T. and De Sousa, J. (1962). Antibodies against certain arboviruses in sera from human beings and domestic animals from the South African highveld. South Af rican Journal of Medical Science, 27, 87-94. McIntosh, B. M., Jupp, P. G., Dickinson, D. B., McGillivray, G. M. and Sweetnam, J. (1967). Ecological studies on Sindbis and West Nile viruses in South Africa. 1. Viral activity as revealed by infection of mosquitoes and sentinel fowls. South African Journal of Medical Science, 32, 1-14. Pruzanski, W. and Altman, R. (1962). Encephalitis due to West Nile fever virus. World Neurology, 3, 524-528. Schmidt, E. and Schmidt, F. W. (1976). Brief Guide to Practical Enzyme Diagnosis, 2nd Edit. Boehringer, Manheim GmbH. Tasker, J. B. (1978). Reference values for clinical chemistry using the Coulter chemistry system. Cornell Veterinarian, 68, 460-479. [Received for publication,


30th, 19871