- Email: [email protected]
Detection of platelet-bound antibodies in beagle dogs after artificial infection with Ehrlichia canis
Veterinary Immunology and Immunopathology 77 (2000) 145±150
Short communication
Detection of platelet-bound antibodies in beagle dogs after arti®cial infection with Ehrlichia canis Trevor Wanera,*, Igor Leykinb, Meir Shinitskyb, Ezra Sharabanic, Hillel Bucha, Avi Keysarya, Hylton Barkc, Shimon Harrusc a Israel Institute for Biological Research, PO Box 19, Ness Ziona, Israel Department of Membrane Research, Weitzmann Institute of Science, Rehovot, Israel c Koret School of Veterinary Medicine, Hebrew University of Jerusalem, PO Box 12, Rehovot 76100, Israel
b
Received 21 January 2000; received in revised form 31 May 2000; accepted 12 July 2000
Abstract Six dogs were infected with Ehrlichia canis by intravenous injection of heavily infected DH82 cells. All dogs developed typical signs of canine monocytic ehrlichiosis. Using ¯ow cytometric technology, platelet-bound IgG (PBIgG) were detected in 5 of the 6 dogs after experimental infection with E. canis over a period of 3±10 days post infection (PI). The ®rst detection of PBIgG was made as early as day 3 PI in 2 out of 6 dogs, and on day 5 PI in 1 dog. On day 7 PI, PBIgG was detected in 2 dogs, and on day 10 PI in 3 out of 6 dogs. This is the ®rst report documenting the presence of PBIgG following E. canis infection in dogs. This ®nding further supports the theory that the thrombocytopenia seen in canine monocytic ehrlichiosis has an immunological component and that exposure to an infectious agent, in this case the rickettsia E. canis, can trigger autoimmune mechanisms. Due to the heterogenous appearance of PBIgG among the infected dogs it was concluded that other non-immunological mechanisms are probably also involved in the pathogenesis of the thrombocytopenia seen in canine monocytic ehrlichiosis. # 2000 Elsevier Science B.V. All rights reserved. Keywords: Ehrlichia canis; Dog; Platelet-bound antibodies
Abbreviations: BPIgG, platelet-bound IgG antibodies; CME, canine monocytic ehrlichiosis; IFA, indirect immuno¯uorescence antibody assay; IgG, immunoglobulin G; MFU, mean ¯uorescent units; MPV, mean platelet volume; PI, post infection * Corresponding author. Tel.: 972-9-9381461; fax: 972-8-9401443. E-mail address: [email protected] (Trevor Waner). 0165-2427/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 2 4 2 7 ( 0 0 ) 0 0 2 2 5 - 7
146
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
1. Introduction Thrombocytopenia is the most prominent and consistent hematological change occurring in canine monocytic ehrlichiosis (CME) (Harrus et al., 1997). The pathogenesis of thrombocytopenia in CME is complex and poorly understood although it has been proposed to be in part immune mediated (Weisiger et al., 1975; Harrus et al., 1996). Arti®cial infection of dogs with Ehrlichia canis has been shown to induce the production of serum platelet-bindable IgG antibodies (Harrus et al., 1996). Furthermore, radiolabeled platelet survival studies have shown that platelets are removed at an accelerated rate in E. canis-infected dogs, and that platelet survival time decreases from 9 days to 4 days, 2±4 days after arti®cial infection with E. canis (Smith et al., 1975). Previous studies have relied on detecting serum platelet-bindable antibodies (Waner et al., 1995; Harrus et al., 1996). This study was undertaken to further elucidate the nature and course of platelet destruction during the incubation and early acute phase of arti®cial CME by studying the emergence of platelet-bound IgG antibodies (BPIgG) in dogs experimentally infected with E. canis. 2. Materials and methods Eight healthy adult beagle dogs (4 males and 4 females) were purchased from Marshall Farms (North Rose, NY). The dogs were about 8 months of age and tested seronegative to E. canis antibodies as determined by the indirect immuno¯uorescence antibody assay (IFA). Hematological and clinical chemistry test results were within reference ranges. The protocol for this study was approved by the ``Institute Animal Use Committee'' and the dogs were cared for according to the guidelines of the NIH for the ``Care and Use of Laboratory Animals'' (1997). Three male and 3 female beagle dogs were selected at random and injected intravenously with 5 ml of tissue culture growth medium containing 106 DH82 cells heavily infected (80%) with the Israel strain (#611) of E. canis (Keysary et al., 1996). One male and 1 female dog were injected with 5 ml of tissue culture medium only, and served as controls. Blood was collected from the jugular vein. Serum samples were stored at ÿ708C until assayed. Samples were tested for E. canis antibodies before infection and on days 3, 5, 7, 8 and 10 post infection (PI) by IFA as described previously (Ristic et al., 1972; Keysary et al., 1996). Complete blood counts were performed within 2 h of blood collection using a semi-automatic impedance cell counter Cellanalyzer CA 580 A (Medonic, Sweden). Platelet-bound IgG were detected by the ¯ow cytometric analysis as previously described (Lewis et al., 1995). For this purpose, blood was collected from all dogs prior to infection, and from control and infected dogs on days 3, 5, 7, and 10 PI. The ¯ow cytometric assay was carried out using the FACscan (Becton Dickinson Immunocytometry Systems, Mountain View, CA). Mean ¯uorescent units (MFU) greater than two standard deviations than values for the preinfected and control dogs combined were considered positive for PBIgG. The intensity of the ¯uorescence was graded as negative for results less than two standard deviations of
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
147
the controls; 1, 2±5 MFU above controls; 2, 6±10 MFU above controls and 3, 11±20 MFU above controls. Comparison between preinfection and postinfection hematological values was carried out using the paired Student's t-test. 3. Results and discussion All 6 beagle dogs injected intravenously with DH82 cells infected with the Israel strain of E. canis, consistently developed typical signs of acute ehrlichiosis. These included pyrexia (mean 40:4 0:3 C) seen at day 10 PI, accompanied by anorexia and generalized lymphadenomegaly. Five of the 6 dogs seroconverted on day 7 PI with titers ranging from 1:20 to 1:160, while the 6th dog (#748) showed antibodies to E. canis on day 8 PI (1:40). On day 10 PI the antibody titers of all 6 dogs had increased to 1:640 to 1:2560. Both control dogs were negative (<1:20) for E. canis antibodies throughout the study period. Thrombocytopenia was a consistent ®nding in all infected dogs. The individual platelet counts for all the dogs are presented in Table 1. By day 3 PI, mean platelet counts for the infected dogs had dropped by 26.4% from the preinfection values (202; 170 15; 000 platelets per ml versus mean preinfection counts of 274; 800 15; 330 platelets per ml) (P 0:0002) and on day 5 PI by 44.1% (P < 0:0001). This trend continued and the mean platelet counts for the 6 E. canis-infected dogs was 72; 330 6230 platelets per ml on day 10 PI, representing a decline of 73.7% from the mean preinfection values (P < 0:0001). Concurrently, platelets were seen to increase in size with mean platelet volumes (MPV) increasing from day 3 PI and reaching 11:5 0:55 ml3 on day 10 PI, compared to the preinfection MPV of 9:83 0:75 ml3 (P 0:0005). MFU for all the dogs throughout the study period are presented in Table 1. Five of the 6 dogs exhibited PBIgG at some stage throughout the 10 day experimental period, with only 1 dog presenting with PBIgG consistently throughout the study. At day 10 PI, 50% of the dogs had PBIgG. PBIgG were detected in 2 out of 6 dogs (#748, #152), 3 days PI and graded as 3 and 1 in intensity, respectively. On day 5 PI, dog #748 had the highest Table 1 Cytometric ¯ow results in mean platelet ¯uorescence units (MFU) and whole blood platelet counts (PLT) (1000 platelets per microliter) for beagle dogs infected with E. canis, from 1 day before infection to 10 days PI Dog no.
Days ÿ1 PI
152 32 628 712 267 748 136 Control 675 Control
3 PI
5 PI
7 PI
10 PI
MFU
PLT
MFU
PLT
MFU
PLT
MFU
PLT
MFU
PLT
0 0 0 0 0 0 0 0
310 241 215 261 231 227 318 208
1 0 0 0 0 3 0 0
260 196 157 207 204 189 301 220
0 0 0 0 0 1 0 0
217 146 124 142 155 134 274 202
0 0 1 0 0 3 0 0
145 112 112 79 120 96 300 188
0 0 0 1 2 2 0 0
84 67 66 49 82 86 316 201
148
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
MFU compared to the other infected dogs, although the intensity of the ¯uorescence had declined and at this stage was graded as 1. On day 7 PI 2 dogs (#628 and #748) showed the presence of PBIgG at MFU intensities of 1 and 3, respectively. On day 10 PI 3 dogs (#267, #712, and #748) were found to be positive for PBIgG, and graded in ¯uorescence intensity as 2, 1 and 2, respectively. The 2 control dogs' sera which were run concurrently with the infected dogs at each examination, did not show any signs of PBIgG at any of the ¯ow cytometric assays. It has been speculated that exposure to exogenous substances such as bacteria and viruses may trigger cases of immune mediated thrombocytopenia (Chang et al., 1998). Evidence for an immune mediated pathogenesis for the destruction of platelets in dogs infected with E. canis has up to now been based on the presence of serum bindable antiplatelet antibodies. This is the ®rst report documenting the presence of PBIgG associated with E. canis infection in dogs during the incubation period and acute phase of the disease. Serum antiplatelet-bindable antibody studies up to now have dealt with the period from day 7 PI and onwards (Harrus et al., 1996; Grindem et al., 1999). This study investigated the presence of PBIgG from day 3 PI to day 10 PI in arti®cially infected dogs. Flow cytometric assay demonstrated consistently high PBIgG in 1 dog (#748), from day 3 to 10 PI. Four other dogs were also shown to develop PBIgG during the course of the study, although the appearance of PBIgG was not consistent, and did not reach the high ¯uorescence intensity as observed in dog #748. These results substantiate previous observations that serum platelet-bindable antibodies are elevated in both experimental and natural cases of CME (Lewis et al., 1995; Waner et al., 1995; Harrus et al., 1996; Grindem et al., 1999). Our results should also be evaluated in the light of the fact that detection of platelet-bound antibodies has been found to be a more sensitive test for the diagnosis of immune mediated thrombocytopenia in dogs as compared to the detection of serum platelet-bindable antibodies (Lewis et al., 1995). The very early detection of PBIgG in 1 dog parallels the precocious decline in platelet numbers seen as early as 3 days PI. At this stage the mean platelet count for all the dogs had declined signi®cantly by 26% from the preinfection values. In a previous study we demonstrated an 11% decline in platelet numbers 3 days PI (Harrus et al., 1996). The difference in severity of decline of platelet numbers may be explained by the difference in load of infection, where a greater number of ehrlichial organisms were used in the present study. The appearance of antiplatelet antibodies at this early stage during the incubation of the disease has been proposed to be due to the possible alteration of the immune system by the E. canis infection resulting in the overproduction of natural antiplatelet antibodies with increased af®nity (Harrus et al., 1996). Further evidence for the premature platelet destruction after E. canis infection has been demonstrated in platelet survival studies where it has been shown that platelet survival time decreases from 9 days to 4 days, 2±4 days after arti®cial infection with E. canis (Smith et al., 1975). We have previously demonstrated the presence of serum platelet-bindable antibodies in arti®cially infected dogs as early as 7 days PI (Harrus et al., 1996). In the same study only 1 of the 6 dogs tested showed serum antiplatelet antibodies at this stage. Even on day 13 PI, only in 3 of 6 dogs could serum antiplatelet antibodies be demonstrated. It was previously postulated that binding of high-avidity antibodies to platelets in the very early
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
149
stages of the disease may result in circulating serum titters of antiplatelet antibody too low to be detected by the indirect assay method, and therefore the direct method was considered the most appropriate test method (Campbell et al., 1984). This study shows that this postulate may not be correct for all dogs with CME, and that both plateletbindable and platelet-bound antibodies can be detected in most dogs infected with E. canis early in the course of the disease. As 3 out of 6 dogs had PBIgG at day 10 PI, it is possible that in the course of time a greater percentage of dogs may have been found positive for PBIgG. Evidence of this may be found in our previous study where at 17 days PI, 5 of 6 dogs were positive for serum platelet-bindable antibodies (Harrus et al., 1996). Heterogeneity was noticed in the development of antiplatelet antibodies during the very early stages of CME. This may indicate that other mechanisms may also be responsible for the development of thrombocytopenia in CME. Proposed mechanisms include increased platelet consumption or sequestration (Smith et al., 1975); enhanced platelet sequestration and stasis due to a platelet migration inhibition factor (Ristic and Holland, 1993); splenic pooling in enlarged spleens, and increased platelet destruction by the spleen (Smith et al., 1974). In conclusion, this study substantiates that exposure to an infectious agent, in this case the rickettsia E. canis, can trigger autoimmune reactions to platelets. Furthermore, using a direct immunological assay technique, this study adds conclusive evidence showing that immunological reaction to platelets occurs at very early stage of CME. Due to the heterogeneous response in appearance of the platelet-bound antibodies, it was also concluded that other non-immune mechanisms may also be involved in the destruction of platelets during the early course of CME. Acknowledgements This work was supported by a grant from the Chief Scientist, Israel Ministry of Health, Jerusalem, Israel (contract number 3960/0). References Campbell, K.L., George, J.W., Greene, C.E., 1984. Application of enzyme-linked immunosorbent assay for the detection of platelet antibodies in dogs. Am. J. Vet. Res. 45, 2561±2564. Chang, Y.F., Novosel, V., Dubovi, E., Wong, S.J., Chu, F.K., Chang, C.F., Del Piero, F., Shin, S., Lein, D.H., 1998. Experimental infection of the human granulocytic ehrlichiosis agent in horses. Vet. Parasitol. 78, 137±145. Grindem, C.B., Breitschwerdt, E.B., Perkins, P.C., Cullins, L.D., Thomas, T.J., Hegarty, B., 1999. Plateletassociated immunoglobulin (antiplatelet antibody) in canine Rocky Mountain fever and ehrlichiosis. J. Am. Anim. Hosp. Assoc. 35, 56±61. Harrus, S., Waner, T., Weiss, D.J., Keysary, A., Bark, H., 1996. Kinetics of serum antiplatelet antibodies in experimental acute canine ehrlichiosis. Vet. Immunol. Immunopathol. 51, 13±20. Harrus, S., Waner, T., Bark, H., 1997. Canine monocytic ehrlichiosis: an update. Comp. Cont. Educ. Pract. Vet. 19, 431±447. Keysary, A., Waner, T., Rosner, M., Warner, C.K., Dawson, J.E., Zass, R., Biggie, K.L., Harrus, S., 1996. The ®rst isolation, in vitro propagation, and genetic characterization of Ehrlichia canis in Israel. Vet. Parasitol. 62, 331±340.
150
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
Lewis, D.C., McVey, D.S., Shuman, W.S., Muller, W.B., 1995. Development of characterization of a ¯ow cytometric assay for detection of platelet-bound immunoglobulin G in dogs. Am. J. Vet. Res. 56, 1555±1558. Lewis, D.C., Meyers, K.M., Callan, B., Bucheler, J., Giger, U., 1995. Detection of platelet-bound and serum platelet-bindable antibodies for diagnosis of idiopathic thrombocytopenic purpura in dogs. J. Am. Vet. Med. Assoc. 206, 47±52. Ristic, M., Holland, C.J., 1993. Canine ehrlichiosis. In: Woldehiwet, Z., Ristic, M. (Eds.), Rickettsial and Chlamydial Diseases of Domestic Animals. Pergamon Press, Oxford, pp. 169±186. Ristic, M., Huxsoll, D.L., Weisiger, R.M., Hildebrandt, P.K., Nyindo, M.B.A., 1972. Serological diagnosis of tropical canine pancytopenia by indirect immuno¯uorescence. Infect. Immun. 6, 226±231. Smith, R.D., Hooks, J.E., Huxsoll, D.L., Ristic, M., 1974. Canine ehrlichiosis (tropical canine pancytopenia): survival of phosphorus-32-labeled blood platelets in normal and infected dogs. Am. J. Vet. Res. 35, 269±273. Smith, R.D., Ristic, M., Huxsoll, D.L., Baylor, R.A., 1975. Platelet kinetics in canine ehrlichiosis: evidence for increased platelet destruction as the cause of thrombocytopenia. Infect. Immun. 11, 1216±1221. Waner, T., Harrus, S., Weiss, D.J., Bark, H., Keysary, A., 1995. Demonstration of serum antiplatelet antibodies in experimental acute canine ehrlichiosis. Vet. Immunol. Immunopathol. 48, 177±182. Weisiger, R.M., Ristic, M., Huxsoll, D.L., 1975. Kinetics of antibody response to Ehrlichia canis assayed by the indirect ¯uorescent antibody method. Am. J. Vet. Res. 36, 689±694.
Short communication
Detection of platelet-bound antibodies in beagle dogs after arti®cial infection with Ehrlichia canis Trevor Wanera,*, Igor Leykinb, Meir Shinitskyb, Ezra Sharabanic, Hillel Bucha, Avi Keysarya, Hylton Barkc, Shimon Harrusc a Israel Institute for Biological Research, PO Box 19, Ness Ziona, Israel Department of Membrane Research, Weitzmann Institute of Science, Rehovot, Israel c Koret School of Veterinary Medicine, Hebrew University of Jerusalem, PO Box 12, Rehovot 76100, Israel
b
Received 21 January 2000; received in revised form 31 May 2000; accepted 12 July 2000
Abstract Six dogs were infected with Ehrlichia canis by intravenous injection of heavily infected DH82 cells. All dogs developed typical signs of canine monocytic ehrlichiosis. Using ¯ow cytometric technology, platelet-bound IgG (PBIgG) were detected in 5 of the 6 dogs after experimental infection with E. canis over a period of 3±10 days post infection (PI). The ®rst detection of PBIgG was made as early as day 3 PI in 2 out of 6 dogs, and on day 5 PI in 1 dog. On day 7 PI, PBIgG was detected in 2 dogs, and on day 10 PI in 3 out of 6 dogs. This is the ®rst report documenting the presence of PBIgG following E. canis infection in dogs. This ®nding further supports the theory that the thrombocytopenia seen in canine monocytic ehrlichiosis has an immunological component and that exposure to an infectious agent, in this case the rickettsia E. canis, can trigger autoimmune mechanisms. Due to the heterogenous appearance of PBIgG among the infected dogs it was concluded that other non-immunological mechanisms are probably also involved in the pathogenesis of the thrombocytopenia seen in canine monocytic ehrlichiosis. # 2000 Elsevier Science B.V. All rights reserved. Keywords: Ehrlichia canis; Dog; Platelet-bound antibodies
Abbreviations: BPIgG, platelet-bound IgG antibodies; CME, canine monocytic ehrlichiosis; IFA, indirect immuno¯uorescence antibody assay; IgG, immunoglobulin G; MFU, mean ¯uorescent units; MPV, mean platelet volume; PI, post infection * Corresponding author. Tel.: 972-9-9381461; fax: 972-8-9401443. E-mail address: [email protected] (Trevor Waner). 0165-2427/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 2 4 2 7 ( 0 0 ) 0 0 2 2 5 - 7
146
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
1. Introduction Thrombocytopenia is the most prominent and consistent hematological change occurring in canine monocytic ehrlichiosis (CME) (Harrus et al., 1997). The pathogenesis of thrombocytopenia in CME is complex and poorly understood although it has been proposed to be in part immune mediated (Weisiger et al., 1975; Harrus et al., 1996). Arti®cial infection of dogs with Ehrlichia canis has been shown to induce the production of serum platelet-bindable IgG antibodies (Harrus et al., 1996). Furthermore, radiolabeled platelet survival studies have shown that platelets are removed at an accelerated rate in E. canis-infected dogs, and that platelet survival time decreases from 9 days to 4 days, 2±4 days after arti®cial infection with E. canis (Smith et al., 1975). Previous studies have relied on detecting serum platelet-bindable antibodies (Waner et al., 1995; Harrus et al., 1996). This study was undertaken to further elucidate the nature and course of platelet destruction during the incubation and early acute phase of arti®cial CME by studying the emergence of platelet-bound IgG antibodies (BPIgG) in dogs experimentally infected with E. canis. 2. Materials and methods Eight healthy adult beagle dogs (4 males and 4 females) were purchased from Marshall Farms (North Rose, NY). The dogs were about 8 months of age and tested seronegative to E. canis antibodies as determined by the indirect immuno¯uorescence antibody assay (IFA). Hematological and clinical chemistry test results were within reference ranges. The protocol for this study was approved by the ``Institute Animal Use Committee'' and the dogs were cared for according to the guidelines of the NIH for the ``Care and Use of Laboratory Animals'' (1997). Three male and 3 female beagle dogs were selected at random and injected intravenously with 5 ml of tissue culture growth medium containing 106 DH82 cells heavily infected (80%) with the Israel strain (#611) of E. canis (Keysary et al., 1996). One male and 1 female dog were injected with 5 ml of tissue culture medium only, and served as controls. Blood was collected from the jugular vein. Serum samples were stored at ÿ708C until assayed. Samples were tested for E. canis antibodies before infection and on days 3, 5, 7, 8 and 10 post infection (PI) by IFA as described previously (Ristic et al., 1972; Keysary et al., 1996). Complete blood counts were performed within 2 h of blood collection using a semi-automatic impedance cell counter Cellanalyzer CA 580 A (Medonic, Sweden). Platelet-bound IgG were detected by the ¯ow cytometric analysis as previously described (Lewis et al., 1995). For this purpose, blood was collected from all dogs prior to infection, and from control and infected dogs on days 3, 5, 7, and 10 PI. The ¯ow cytometric assay was carried out using the FACscan (Becton Dickinson Immunocytometry Systems, Mountain View, CA). Mean ¯uorescent units (MFU) greater than two standard deviations than values for the preinfected and control dogs combined were considered positive for PBIgG. The intensity of the ¯uorescence was graded as negative for results less than two standard deviations of
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
147
the controls; 1, 2±5 MFU above controls; 2, 6±10 MFU above controls and 3, 11±20 MFU above controls. Comparison between preinfection and postinfection hematological values was carried out using the paired Student's t-test. 3. Results and discussion All 6 beagle dogs injected intravenously with DH82 cells infected with the Israel strain of E. canis, consistently developed typical signs of acute ehrlichiosis. These included pyrexia (mean 40:4 0:3 C) seen at day 10 PI, accompanied by anorexia and generalized lymphadenomegaly. Five of the 6 dogs seroconverted on day 7 PI with titers ranging from 1:20 to 1:160, while the 6th dog (#748) showed antibodies to E. canis on day 8 PI (1:40). On day 10 PI the antibody titers of all 6 dogs had increased to 1:640 to 1:2560. Both control dogs were negative (<1:20) for E. canis antibodies throughout the study period. Thrombocytopenia was a consistent ®nding in all infected dogs. The individual platelet counts for all the dogs are presented in Table 1. By day 3 PI, mean platelet counts for the infected dogs had dropped by 26.4% from the preinfection values (202; 170 15; 000 platelets per ml versus mean preinfection counts of 274; 800 15; 330 platelets per ml) (P 0:0002) and on day 5 PI by 44.1% (P < 0:0001). This trend continued and the mean platelet counts for the 6 E. canis-infected dogs was 72; 330 6230 platelets per ml on day 10 PI, representing a decline of 73.7% from the mean preinfection values (P < 0:0001). Concurrently, platelets were seen to increase in size with mean platelet volumes (MPV) increasing from day 3 PI and reaching 11:5 0:55 ml3 on day 10 PI, compared to the preinfection MPV of 9:83 0:75 ml3 (P 0:0005). MFU for all the dogs throughout the study period are presented in Table 1. Five of the 6 dogs exhibited PBIgG at some stage throughout the 10 day experimental period, with only 1 dog presenting with PBIgG consistently throughout the study. At day 10 PI, 50% of the dogs had PBIgG. PBIgG were detected in 2 out of 6 dogs (#748, #152), 3 days PI and graded as 3 and 1 in intensity, respectively. On day 5 PI, dog #748 had the highest Table 1 Cytometric ¯ow results in mean platelet ¯uorescence units (MFU) and whole blood platelet counts (PLT) (1000 platelets per microliter) for beagle dogs infected with E. canis, from 1 day before infection to 10 days PI Dog no.
Days ÿ1 PI
152 32 628 712 267 748 136 Control 675 Control
3 PI
5 PI
7 PI
10 PI
MFU
PLT
MFU
PLT
MFU
PLT
MFU
PLT
MFU
PLT
0 0 0 0 0 0 0 0
310 241 215 261 231 227 318 208
1 0 0 0 0 3 0 0
260 196 157 207 204 189 301 220
0 0 0 0 0 1 0 0
217 146 124 142 155 134 274 202
0 0 1 0 0 3 0 0
145 112 112 79 120 96 300 188
0 0 0 1 2 2 0 0
84 67 66 49 82 86 316 201
148
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
MFU compared to the other infected dogs, although the intensity of the ¯uorescence had declined and at this stage was graded as 1. On day 7 PI 2 dogs (#628 and #748) showed the presence of PBIgG at MFU intensities of 1 and 3, respectively. On day 10 PI 3 dogs (#267, #712, and #748) were found to be positive for PBIgG, and graded in ¯uorescence intensity as 2, 1 and 2, respectively. The 2 control dogs' sera which were run concurrently with the infected dogs at each examination, did not show any signs of PBIgG at any of the ¯ow cytometric assays. It has been speculated that exposure to exogenous substances such as bacteria and viruses may trigger cases of immune mediated thrombocytopenia (Chang et al., 1998). Evidence for an immune mediated pathogenesis for the destruction of platelets in dogs infected with E. canis has up to now been based on the presence of serum bindable antiplatelet antibodies. This is the ®rst report documenting the presence of PBIgG associated with E. canis infection in dogs during the incubation period and acute phase of the disease. Serum antiplatelet-bindable antibody studies up to now have dealt with the period from day 7 PI and onwards (Harrus et al., 1996; Grindem et al., 1999). This study investigated the presence of PBIgG from day 3 PI to day 10 PI in arti®cially infected dogs. Flow cytometric assay demonstrated consistently high PBIgG in 1 dog (#748), from day 3 to 10 PI. Four other dogs were also shown to develop PBIgG during the course of the study, although the appearance of PBIgG was not consistent, and did not reach the high ¯uorescence intensity as observed in dog #748. These results substantiate previous observations that serum platelet-bindable antibodies are elevated in both experimental and natural cases of CME (Lewis et al., 1995; Waner et al., 1995; Harrus et al., 1996; Grindem et al., 1999). Our results should also be evaluated in the light of the fact that detection of platelet-bound antibodies has been found to be a more sensitive test for the diagnosis of immune mediated thrombocytopenia in dogs as compared to the detection of serum platelet-bindable antibodies (Lewis et al., 1995). The very early detection of PBIgG in 1 dog parallels the precocious decline in platelet numbers seen as early as 3 days PI. At this stage the mean platelet count for all the dogs had declined signi®cantly by 26% from the preinfection values. In a previous study we demonstrated an 11% decline in platelet numbers 3 days PI (Harrus et al., 1996). The difference in severity of decline of platelet numbers may be explained by the difference in load of infection, where a greater number of ehrlichial organisms were used in the present study. The appearance of antiplatelet antibodies at this early stage during the incubation of the disease has been proposed to be due to the possible alteration of the immune system by the E. canis infection resulting in the overproduction of natural antiplatelet antibodies with increased af®nity (Harrus et al., 1996). Further evidence for the premature platelet destruction after E. canis infection has been demonstrated in platelet survival studies where it has been shown that platelet survival time decreases from 9 days to 4 days, 2±4 days after arti®cial infection with E. canis (Smith et al., 1975). We have previously demonstrated the presence of serum platelet-bindable antibodies in arti®cially infected dogs as early as 7 days PI (Harrus et al., 1996). In the same study only 1 of the 6 dogs tested showed serum antiplatelet antibodies at this stage. Even on day 13 PI, only in 3 of 6 dogs could serum antiplatelet antibodies be demonstrated. It was previously postulated that binding of high-avidity antibodies to platelets in the very early
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
149
stages of the disease may result in circulating serum titters of antiplatelet antibody too low to be detected by the indirect assay method, and therefore the direct method was considered the most appropriate test method (Campbell et al., 1984). This study shows that this postulate may not be correct for all dogs with CME, and that both plateletbindable and platelet-bound antibodies can be detected in most dogs infected with E. canis early in the course of the disease. As 3 out of 6 dogs had PBIgG at day 10 PI, it is possible that in the course of time a greater percentage of dogs may have been found positive for PBIgG. Evidence of this may be found in our previous study where at 17 days PI, 5 of 6 dogs were positive for serum platelet-bindable antibodies (Harrus et al., 1996). Heterogeneity was noticed in the development of antiplatelet antibodies during the very early stages of CME. This may indicate that other mechanisms may also be responsible for the development of thrombocytopenia in CME. Proposed mechanisms include increased platelet consumption or sequestration (Smith et al., 1975); enhanced platelet sequestration and stasis due to a platelet migration inhibition factor (Ristic and Holland, 1993); splenic pooling in enlarged spleens, and increased platelet destruction by the spleen (Smith et al., 1974). In conclusion, this study substantiates that exposure to an infectious agent, in this case the rickettsia E. canis, can trigger autoimmune reactions to platelets. Furthermore, using a direct immunological assay technique, this study adds conclusive evidence showing that immunological reaction to platelets occurs at very early stage of CME. Due to the heterogeneous response in appearance of the platelet-bound antibodies, it was also concluded that other non-immune mechanisms may also be involved in the destruction of platelets during the early course of CME. Acknowledgements This work was supported by a grant from the Chief Scientist, Israel Ministry of Health, Jerusalem, Israel (contract number 3960/0). References Campbell, K.L., George, J.W., Greene, C.E., 1984. Application of enzyme-linked immunosorbent assay for the detection of platelet antibodies in dogs. Am. J. Vet. Res. 45, 2561±2564. Chang, Y.F., Novosel, V., Dubovi, E., Wong, S.J., Chu, F.K., Chang, C.F., Del Piero, F., Shin, S., Lein, D.H., 1998. Experimental infection of the human granulocytic ehrlichiosis agent in horses. Vet. Parasitol. 78, 137±145. Grindem, C.B., Breitschwerdt, E.B., Perkins, P.C., Cullins, L.D., Thomas, T.J., Hegarty, B., 1999. Plateletassociated immunoglobulin (antiplatelet antibody) in canine Rocky Mountain fever and ehrlichiosis. J. Am. Anim. Hosp. Assoc. 35, 56±61. Harrus, S., Waner, T., Weiss, D.J., Keysary, A., Bark, H., 1996. Kinetics of serum antiplatelet antibodies in experimental acute canine ehrlichiosis. Vet. Immunol. Immunopathol. 51, 13±20. Harrus, S., Waner, T., Bark, H., 1997. Canine monocytic ehrlichiosis: an update. Comp. Cont. Educ. Pract. Vet. 19, 431±447. Keysary, A., Waner, T., Rosner, M., Warner, C.K., Dawson, J.E., Zass, R., Biggie, K.L., Harrus, S., 1996. The ®rst isolation, in vitro propagation, and genetic characterization of Ehrlichia canis in Israel. Vet. Parasitol. 62, 331±340.
150
Trevor Waner et al. / Veterinary Immunology and Immunopathology 77 (2000) 145±150
Lewis, D.C., McVey, D.S., Shuman, W.S., Muller, W.B., 1995. Development of characterization of a ¯ow cytometric assay for detection of platelet-bound immunoglobulin G in dogs. Am. J. Vet. Res. 56, 1555±1558. Lewis, D.C., Meyers, K.M., Callan, B., Bucheler, J., Giger, U., 1995. Detection of platelet-bound and serum platelet-bindable antibodies for diagnosis of idiopathic thrombocytopenic purpura in dogs. J. Am. Vet. Med. Assoc. 206, 47±52. Ristic, M., Holland, C.J., 1993. Canine ehrlichiosis. In: Woldehiwet, Z., Ristic, M. (Eds.), Rickettsial and Chlamydial Diseases of Domestic Animals. Pergamon Press, Oxford, pp. 169±186. Ristic, M., Huxsoll, D.L., Weisiger, R.M., Hildebrandt, P.K., Nyindo, M.B.A., 1972. Serological diagnosis of tropical canine pancytopenia by indirect immuno¯uorescence. Infect. Immun. 6, 226±231. Smith, R.D., Hooks, J.E., Huxsoll, D.L., Ristic, M., 1974. Canine ehrlichiosis (tropical canine pancytopenia): survival of phosphorus-32-labeled blood platelets in normal and infected dogs. Am. J. Vet. Res. 35, 269±273. Smith, R.D., Ristic, M., Huxsoll, D.L., Baylor, R.A., 1975. Platelet kinetics in canine ehrlichiosis: evidence for increased platelet destruction as the cause of thrombocytopenia. Infect. Immun. 11, 1216±1221. Waner, T., Harrus, S., Weiss, D.J., Bark, H., Keysary, A., 1995. Demonstration of serum antiplatelet antibodies in experimental acute canine ehrlichiosis. Vet. Immunol. Immunopathol. 48, 177±182. Weisiger, R.M., Ristic, M., Huxsoll, D.L., 1975. Kinetics of antibody response to Ehrlichia canis assayed by the indirect ¯uorescent antibody method. Am. J. Vet. Res. 36, 689±694.