- Email: [email protected]
Purification of Cytoecetes phagocytophila, the causative agent of tick-borne fever, from Infected ovine blood
J. Comp.Path. 1991Vol. 105
Purification of Cytoecetesphagocytophila, the Causative Agent of Tick-Borne Fever, from Infected Ovine Blood Z. Woldehiwet, S. D. Carter and C. Dare University of Liverpool, Department of VeterinaryPathology, VeterinaryField Station, Leahurst, Neston, Wirral L64 7TE, U.K. Summary Cytoecetes phagocytophila, the causative agent of tick-borne fever, was successfully separated by Percoll and Renografin density gradient centrifugation and by cellular affinity chromatography, from the peripheral blood leucocytes of experimentally infected sheep. After centrifugation on Renografin or PercoU density gradients, infectious particles of C. phagocytophila banded at buoyant densities which ranged between 1.063 to 1"140. Rickettsiae separated by wheat germ lectin cellular affinity chromatography retained their morphology but often lost their infectivity. Cell-free C. phagocytophila remained infective to susceptible sheep for 6 months when kept at - 114°C in sucrose phosphate buffer with 10 per cent dimethylsulphoxide as a cryopreservatlve.
Introduction Tick-borne fever (TBF) is a febrile disease of sheep and cattle characterized by very high fever and severe leucopenia (Taylor, Holman and Gordon, 1941). The febrile period of TBF is often followed by other infections such as tickpyaemia, louping ill (MacLeod and Gordon, 1933), listeriosis (Gronstol and Ulvund, 1977) and pasteurellosis (Gilmour, Brodie and Holmes, 1982). The causative organism, Cytoecetesphagocytophila, is an obligate intracellular rickettsia which invades polymorphonuclear (PMN) cells and monocytes (Woldehiwet, 1987). During the peak period ofparasitaemia, both plasma and serum are infective (Foggie, 1951), but extracellular organisms have not been demonstrated. This has hindered progress in understanding the molecular biology and antigenic composition of the organism. Several methods have been used to purify similar organisms. For example, the Chlamydia spp. and several rickettsiae have been separated by density gradient centrifugation with Renografin (Howard, Orenstein and King, 1974; Dasch and Weiss, 1977) or Percoll (Tamura, Urakami and Tsuruhara, 1982; Neitz, Viljoen, Benzuidenhout, Oberem, Putterill, Verschoor, Visser and Vermeulen, 1986). Other Gram-negative bacteria have been separated by cellular affinity chromatography with wheat germ lectins (Viljoen, Vermeulen, Oberem, Prozesky, Verschoor, Bezuidenhout, Putterill, Visser and Neitz, 1985). In the present study, we successfully used Percoll and Renografin density gradient centri0021-9975/91/080431 + 08 $03.00/0
© 1991AcademicPressLimited
432
Z. Woldenhiwet et al.
fugation to purify viable C. phagocytophila from peripheral blood leucocytes of experimentally infected sheep. M a t e r i a l s and M e t h o d s
Cytoecetes phagocytophila A stabilite of the Old Sourhope Strain (Foster and Cameron~ 1970) was stored at - 1 1 4 " C as heparinized blood with 10 per cent dimethylsulphoxide ( D M S O as cryopreservative.
Experimental Infection One-year-old Suffolk cross-bred sheep, bred and maintained in tick-free condinons, were inoculated intravenously with 1'0 ml of a 10 per cent dilution of the stabilate in phosphate buffered saline (PBS) at pH 7"2. The sheep were kept under observation for up to 21 days and rectal temperatures and blood samples taken every day.
Collection of Blood Samples Peripheral blood samples were collected in EDTA-coated tubes and smears stained with Giemsa were examined for the presence of cytoplasmic inclusions in the granulocytes and monocytes and to establish the differential white cell count. The day the organisms were first detected was regarded as the first day of parasitaemia (Woldehiwet, 1987).
Preparation of Infected Leucocytes and Plasma Blood samples were collected during the peak of parasitaemia in heparin-coated containers as described previously (Woldehiwet, 1987). The plasma was separated by centrifugation. A volume of cold PBS equal to the volume of plasma was added to the packed cell fraction before adding cold, sterile distilled water to lyse the red blood cells. Isotonicity was restored by adding 2"7 per cent NaC1 solution in PBS as described by Carlson and Kaneko (1973). After washing three times with a sucrose phosphate glutamate solution (SPG) (Bovarnick, Miller and Snyder, 1950), the cell pellet was resuspended in SPG and the cells treated with one cycle ultrasonic vibration at maximum amplitude for 20 seconds (Vibra-Cell, Sonics & Materials Inc., Danbury, U.S.A.) to disintegrate the leucocytes and release the rickettsiae. The disintegrated cells were centrifuged at 1000 Xg for 30 min to clear debris. The saved plasma and the supernate of the disintegrated cells were separately centrifuged at 3000 x g at 4*C for 30 min. The supernates were again centrifuged at 80 000 x g for 30 min. The pellets were washed with SPG twice and finally resuspended in 1 to 2 ml ofSPG, R P M I 1640 medium or HEPES/saline buffer. These suspensions of cell-free rickettsiae were used for infectivity assays and to purify the rickettsiae further by Percoll and Renografin density centrifilgation or by affinity chromatography.
Separation of C. phagocytophila by Percoll Gradient Suspensions of the reconstituted pellet (rickettsiae) were carefully layered onto a tube filled with 50 per cent Percoll (Pharmacia Fine Chemicals, Uppsala, Sweden). After centrifugation for 15 min at 30 000 X g with maximal acceleration and half maximal braking in a fixed angle rotor in a Sorvall ultracentrifuge (Sorvall Instruments, Newtown, Connecticut, U.S.A.), fbur 9-ml fractions with buoyant densities of 1"139 to 1,098, 1.063 to 1'057, 1"057 to 1.051 and 1"051 to 1"018 were collected. Commercial density marker beads (Sigma Chemicals Co., Ltd, Poole, U.K.) were used to establish the buoyant densities of the various fractions. These fractions were assayed for infectivity and rickettsial purity by electron microscopy.
Purification of TBF Agent
433
Separation of C. phagocytophila by Renografin Density Gradient Rickettsial suspensions in SPG were layered on to a discontinuous Urografin (Urografin 370, Schering AG, Germany) gradient in SPG buffer with a final concentration of 10, 20 and 30 per cent Urografin. After the tubes were centrifuged at 45 000 x g for 1 h at 4*C, two visible bands were formed and three fractions, two of which contained the visible bands, were collected. One millilitre volumes containing the visible bands were weighed to establish their buoyant densities. The rickettsiae contained in these fractions were pelleted by centrifugation at 80 000 x g for 30 min, washed twice in SPG to remove Urografin and assayed for infectivity and rickettsial purity by electron microscopy.
Isolation of C. phagocytophila by AJ]inity Chromatography The method of purifying the related rickettsia Cowdria ruminantium by wheat germ lectin chromatography described by Viljoen et al. (1985) was followed. Briefly, a gel of wheat germ lectin Sepharose 6MB was generated in a 10 cm column with 100 ml of 0.1 M Tris-HC1, 0"5 ~ NaC1, 0'02 per cent (w/v) NaNa, pH 8"5, followed by 100 rnl of 0" 1 M sodium acetate, 0.5 M NaC1, 0.02 per cent NaNa, pH 4.5 and equilibrated with 100 ml of HEPES buffer containing 0"02 per cent NaN 3. After washing the column with 500 ml of HEPES buffer pH 7.4, to remove azide, 1 ml of rickettsial suspension was added and the column incubated for 2 h at room temperature. Non-adsorbed material was eluted with HEPES buffer. The column was pulsed with N-acetyl-D-glucosamine in HEPES (20 ml of HEPES with 2g of carbohydrate) at a flow of 40 ml per h to remove adherent rickettsiae. Elution of C. phagoeytophila was monitored by a single path UV monitor (Uvicord II, LKB, Stockholm, Sweden). The fi'actions containing proteins were pooled and used for assaying infectivity and purity by electron microscopy.
Infectivity of Cell:free C. phagocytophila For infectivity assays of the Percoll gradients, the sample was divided into four fractions, each of 0"1 ml, and four susceptible sheep were each injected with 0'1 ml. The Renografin gradient showed two clear bands. The gradient was therefore divided into three fractions and three susceptible sheep were each inoculated with 0" 1 ml, one for each fraction. Two susceptible sheep were inoculated with the concentrated material eluted from the lectin column and monitored for febrile reaction and parasitaemia. The infectivity assays were repeated in four separate experiments to purify rickettsiae by density gradient centrifugation and by affinity chromatography. Non-reacting animals were challenged with a stabilate of C. phagocytophila as described above. To ensure that the rickettsiae prepared from the plasma or the disintegrated leucocytes were infective before the purification processes, a susceptible sheep was inoculated with 0' 1 ml of the final suspension before purification by density gradient centrifugation or cellular affinity chromatography.
Viability of Cell-free C. phagocytophila The pelleted cell-free rickettsiae and purified infected leucocytes were separately resuspended in SPG with t0 .per cent DMSO and stored at - 114°C. One susceptible sheep was inoculated with a suspension of infected leucocytes or with cell-free rickettsiae after 1 and 2 weeks and after 6 months.
Electron Microsopic Studies Aliquots of Percoll and Renografin density gradient fi-actions were centrifuged at 80 000 x g for 30 rain in a swlng-out rotor in a Sorvall OTD-Combi ultracentrifuge to pellet the organisms. The sediments of each fraction were fixed in 2'5 per cent glutaraldehyde solution in PBS overnight, post.fixed in 2 per cent osmium tetroxide,
434
Z. W o l d e h i w e t e t al.
stained with 2 per cent uranyl acetate in 0"69 per cent maleic acid, dehydrated and embedded in polythene capsules in resin (Taab, Aldermaston, U.K.) polymerized at 60°C for 48 h, then stained with lead citrate and examined with an electron microscope (Philips, EM201). Results
Isolation of C. phagocytophila Materials prepared from disintegrated leucocytes obtained from infected sheep formed clearly visible bands in Percoll and Renografin gradients. After centrifugation on a Percoll density gradient, two faintly visible bands were formed. The lower band was formed at density of 1.098 to 1.120 g per cm a and the higher band was formed at a density of 1"018 to 1"033 g per cm a. Aliquots from the fraction with a buoyant density ranging from 1.139 to 1"063 had rickettsia-like structures (Fig. 1A). The structures had typical rickettsial cell membrane (CM) and outer membranes (OM) (Fig. 1B). After centrifugation on a 10 to 30 per cent discontinuous Renografin gradient, two clear bands were formed at densities of 1"14 g per cm a and 1.09 g per cm ~, respectively. Rickettsia-like structures were present in fractions which contained either of the bands (Fig. 2). These structures were similar to the intracellular forms of C. phagocytophila described earlier (Woldehiwet and Scott, 1982). Leucocyte preparations treated by lectin affinity chromatography also yielded similar rickettsia-like structures. When cell-free materials prepared from the plasma of infected blood were purified by Renografin density gradients, pure rickettsial structures were readily visible in fractions which contained a visible band at a buoyant density of 1"14 (Fig. 3A). These structures also had typical rickettsial CM and O M (Fig. 3B).
Infectivity of Cell-free C. phagocytophila All susceptible sheep infected with cryopreserved whole-blood or purified leucocytes containing C. phagocytophila reacted with fever and parasitaemia. Fresh, cell-free rickettsiae, prepared from both the plasma and disintegrated leucocytes, were also infective to susceptible sheep. After Percoll density gradient centrifugation of suspensions prepared from disintegrated leucocytes, all susceptible sheep inoculated with the fractions containing the lowest band, with a density of 1.139 to 1"063 g per cm a, developed fever and parasitaemia 4 to 6 days post-inoculation (PI). Fractions with densities lower than 1"063 g per cm 3 were not infective. After Renografin density gradient centrifugation, all sheep inoculated with the fractions containing the two bands (densities of 1.14g per cm a and 1.09g per cm a respectively) reacted with fever and parasitaemia 5 to 7 days PI. C. phagocytophila separated from disintegrated leucocytes by cellular affinity chromatography was infective to susceptible sheep on only one of four attempts.
Purification of TBF Agent
Fig. 1.
435
(A) Thin section of rickettsial structures purified from leucocytes infected with C. phagocytophila by Percoll gradient centri|hgation. × 33 000. (B) Percoll gradient purified rickettsia from leucocytes infected with C. phagocytophila showing cell membrane (CM) and outer membrane (OM). x 82 500.
Fig. 2. Thin section of rickettsial structures purified from leucocytes infected with C. phagoo,tophila by Renografin gradient centrifugatkm, x 33 000.
436
Fig. 3.
Z. W o l d e h i w e t et al.
(A) Thin section of rickettsial structures purified from the plasma of sheep infected with C. phagocytophila by R.enografin gradient eentrifilgation, x 30 000. (B) Renografin gradient purified rickettsia li'om the plasma of sheep infected with C. phagocytophila showing cell membrane (CM) and outer membrane (OM). x 75 000.
Viability of Cell-free C. phagocytophila Cell-free and cell-associated organisms remained infective to susceptible sheep when kept at - 1140C in SPG buffer with 10 per cent DMSO as a cryopreservative for 6 months, the longest period of storage examined.
Discussion Molecular and antigenic studies required pure and large numbers of the organisms to be studied. Antigenic and molecular studies of C. phago~,lophila have been hampered by the inability to obtain organisms free from host cells. Some of the obligate intracellular rickettsiae are difficult to purify because of their instability and fragility (Tamura et al., 1982). Density gradients of Renografin and PercolI have been found to be very useful steps in the purification of several obligate intracellular bacteria, including Chlamydia trachomatis (Howard et al., 1974), Rickettsia prowazekii (Dasch and Weiss, 1977) and Rickettsia lsutsugamushi (Tamura et al., 1982). C. phagocytophila is an obligate intracellular rickettsia belonging to the tribe Ehrlichiae in the family Rickettsiaceae. These tick-transmitted animal rickettsiae are difficult to cultivate in vitro and extremely labile, making their isolation problematic. In the present study, we demonstrated that pure and viable C. phagocytophila can be obtained from infected polymorphonuclear cells with Percoll and Renografin density gradients. We are not aware of any other report on the purification ofC. phagocytophila from infected cells but two groups of workers have success~hlly separated two
Purification of TBF Agent
437
members of the tribe Ehrlichiae by density gradient centrifugation. Neitz eL al. (1986) used Percoll density gradient centrifugation to isolate viable C. ruminantium from infected sheep brain and Dutta, Rice, Hughes, Savage and Myrup (1987) used Renografin density gradients to purify Ehrlichia risticii antigens for the enzyme-linked immunosorbent assay (ELISA). In the present study we found that purified C. phagocytophila remained viable if the Renografin gradients were prepared in SPG buffer while rickettsiae separated by Percoll gradients in either R P M I 1640 medium or SPG buffer were infective to susceptible sheep. No attempt was made to quantify infectivity or loss of infectivity since it would have required large numbers of susceptible sheep. Losses of infectivity in Renografin density gradients have been reported in R. tsutsugamushi (Tamura et al., 1982) which is regarded as being more fragile outside host cells than R. typhi and R. prowazekii which retain most of their infectivity after separation with Renografin (Dasch and Weiss, 1977). Although the plasma was found to be a good source of infective C. phagocytophila and the rickettsial structures isolated fi~om the plasma had no apparent cellular contamination, we cannot be certain that the rickettsial structures present in the plasma were all C. phagocytophila. Other rickettsiae, such as Eperythrozoon ovis, which intact the erythrocytes can be present free in the plasma (Kreier and Ristic, 1963; McKee, Ziegler and Giles, 1973). E. ovis has often been presented as a rickettsia causing inapparent infections in sheep which may complicate experimental work in sheep (Littlejohns, 1960). The first reported infection of sheep with E. ovis occurred after the intravenous inoculation of sheep with blood containing Cowdria ruminantium the causative agent ofheartwater (Neitz, Alexander and du Toit, 1934). Foggie and Nisbet (1964) also reported that blood stabilates of C. phagocytophila may contain E. ovis. However, in the present study, the erythrocytes of the blood sample from which the plasma was prepared did not appear to harbour E. ovis. Furthermore, Foggie and Nisbet (1964) had reported that, after experimental infection with both agents, sheep which were heavily infected with C. phagocytophila had little or no E. ovis on their erythrocytes. Nevertheless, for antigenic and molecular studies of C. phagocytophila, rickettsiae isolated from purified infected neutrophils should be used because they are less likely to be contaminated by E. ovis. Now that we have successfully demonstrated the means of obtaining viable C. phagocytophila free of host cells, it is intended to study the antigenic components of the organism which stimulate host immune responses. Acknowledgments
This work was supported by the Wellcome Trust. The technical support of A. Joplin and L. Hill is gratefully acknowledged. References
Bovarnick, M. R., Miller, J. C. and Snyder, J. C. (1950). The influence of certain salts, amino acids, sugars and proteins on the stability of Rickettsiae. Journal of Bacteriology, 59, 509-522. Carlson, G. P. and Kaneko, J.J. (1973). Isolation ofleukocytes from bovine peripheral blood. Proceedings of Experimental Biology and Medicine, 142, 853-856.
438
Z. W o l d e h i w e t
et
al.
Dasch, G. A. and Weiss, E. (1977). Characterization of the Madrid E strain of Rickettsia prowazekii purified by Renografin density centrifugation. Infection and Immunity, 15, 280-286. Dutta, S. K., Rice, R. M., Hughes, T. D., Savage, P. K. and Myrup, A. C. (1987). Detection of serum antibodies against Ehrliehia risticii in potamic horse fever by enzyme-linked immunosorbent assay. Veterinary Immunology and Immunopathology, 12, 84-92. Foggie, A. (1951). Studies on the infectious agent of tick-borne fever in sheep. Journal of Pathology and Bacteriology, 63, 1-15. Foggie, A. and Nisbet, D. I. (1964). Studies on Eperythrozoon infections in sheep. Journal of ComparativePathology, 74, 45-61. Foster, W. N. M. and Cameron, A. E. (1970). Observations on ovine strains of tick-borne fever. Journal of ComparativePathology, 80, 429-436. GiImour, N . J . L . , Brodie, T. A. and Holmes, P. H. (1982). Tick-borne fever and pasteurellosis in sheep. VeterinaryRecord, 111, 512. Gronstol, H. and Ulvund, M.J. (1977). Listeric septicaemia in sheep associated with tick-borne fever. Acta VeterinariaScandinavica, 18, 575-577. Howard, L., Orenstein, N. S. and King, N. W. (1974). Purification on Renografin density gradients of Chlamydia trachomatis grown in the yolk sac of eggs. Applied Microbiology, 27, 102-106. Kreier, J. P. and Ristic, M. (1963). Morphologic antigenic and pathogenic characteristics of Eperythrozoon or# and Eperythrozoon wenyoni. American Journal of Veterinary Research, 24, 488-500. Littlejohns, I. R. (1960). Eperythrozoonosis in sheep. Australian VeterinaryJournal, 36, 261-265. MacLeod, J. and Gordon, W. S. (1933). Studies on tick-borne fever of sheep. I. Transmission by the tick Ixodes rieinus with a description of the disease produced. Parasitology, 25, 273-283. McKee, A. E., Ziegler, R. F. and Giles, R. C. X. (1973). Scanning and transmission electron microscopy of Haemobartonella canis and Eperythrozoonovis. AmericanJournal of Veterinary Research, 34, 1196-1201. Neltz, A. W. H., Viljoen, G. J., Benzuidenhout, J. D., Oberem, P. T., PutteriU, J. F., Verschoor, J. A., Visser, L. and Vermeulen, N. M. (1986). Isolation of Cowdria rumlnantium by means of Percoll density gradient centrifugation and detection by enzyme-linked immunosorbent assay. OnderstepoortJournal of VeterinaryResearch,53, 63-67. Neitz, W. O., Alexander, R. A. and du Toit, P.J. (1934). Eperythrozoonovis (sp. nov.) infection in sheep. OnderstepoortJournal of Veterinary Science, 3, 263-274. Tamura, A., Urakami, .H. and Tsuruhara, T. (1982). Purification of Rickettsia tsutsugamushi by Percoll density gradient centrifugation. Microbiology and Immunology, 28, 873-882. Taylor, A. W., Holman, H. H. and Gordon, W. S. (1941). Attempts to reproduce the pyaemia associated with tick-bite. VeterinaryRecord, 53, 339-344. Viljoen, G. J., Vermeulen, N. M. J., Oberem, P. T., Prozesky, L., Verschoor, J. A., Bezuidenhout, J. D., Putterill, J. F., Visser, L. and Neitz, A. W. H. (1985). Isolation ofCowdria ruminantium by cellular affinity chromatography and detection by an enzyme-linked immunosorbent assay. Onderstepoort Journal of Veterinary Research, 52, 227-232. Woldehiwet, Z. (1987). The effects of tick-borne fever on some functions of polymorphonuclear cells of sheep. Journal of ComparativePathology, 97, 481-485. Woldehiwet, Z. and Scott, G. R. (1982). Stages in the development of Cytoecetes phagocytophila, the causative agent of tick-borne fever. Journal of Comparative Pathology, 92, 469-474.
Reeeived, May 17th, ~Accepted, June 28th,
1991 ] 1991
Purification of Cytoecetesphagocytophila, the Causative Agent of Tick-Borne Fever, from Infected Ovine Blood Z. Woldehiwet, S. D. Carter and C. Dare University of Liverpool, Department of VeterinaryPathology, VeterinaryField Station, Leahurst, Neston, Wirral L64 7TE, U.K. Summary Cytoecetes phagocytophila, the causative agent of tick-borne fever, was successfully separated by Percoll and Renografin density gradient centrifugation and by cellular affinity chromatography, from the peripheral blood leucocytes of experimentally infected sheep. After centrifugation on Renografin or PercoU density gradients, infectious particles of C. phagocytophila banded at buoyant densities which ranged between 1.063 to 1"140. Rickettsiae separated by wheat germ lectin cellular affinity chromatography retained their morphology but often lost their infectivity. Cell-free C. phagocytophila remained infective to susceptible sheep for 6 months when kept at - 114°C in sucrose phosphate buffer with 10 per cent dimethylsulphoxide as a cryopreservatlve.
Introduction Tick-borne fever (TBF) is a febrile disease of sheep and cattle characterized by very high fever and severe leucopenia (Taylor, Holman and Gordon, 1941). The febrile period of TBF is often followed by other infections such as tickpyaemia, louping ill (MacLeod and Gordon, 1933), listeriosis (Gronstol and Ulvund, 1977) and pasteurellosis (Gilmour, Brodie and Holmes, 1982). The causative organism, Cytoecetesphagocytophila, is an obligate intracellular rickettsia which invades polymorphonuclear (PMN) cells and monocytes (Woldehiwet, 1987). During the peak period ofparasitaemia, both plasma and serum are infective (Foggie, 1951), but extracellular organisms have not been demonstrated. This has hindered progress in understanding the molecular biology and antigenic composition of the organism. Several methods have been used to purify similar organisms. For example, the Chlamydia spp. and several rickettsiae have been separated by density gradient centrifugation with Renografin (Howard, Orenstein and King, 1974; Dasch and Weiss, 1977) or Percoll (Tamura, Urakami and Tsuruhara, 1982; Neitz, Viljoen, Benzuidenhout, Oberem, Putterill, Verschoor, Visser and Vermeulen, 1986). Other Gram-negative bacteria have been separated by cellular affinity chromatography with wheat germ lectins (Viljoen, Vermeulen, Oberem, Prozesky, Verschoor, Bezuidenhout, Putterill, Visser and Neitz, 1985). In the present study, we successfully used Percoll and Renografin density gradient centri0021-9975/91/080431 + 08 $03.00/0
© 1991AcademicPressLimited
432
Z. Woldenhiwet et al.
fugation to purify viable C. phagocytophila from peripheral blood leucocytes of experimentally infected sheep. M a t e r i a l s and M e t h o d s
Cytoecetes phagocytophila A stabilite of the Old Sourhope Strain (Foster and Cameron~ 1970) was stored at - 1 1 4 " C as heparinized blood with 10 per cent dimethylsulphoxide ( D M S O as cryopreservative.
Experimental Infection One-year-old Suffolk cross-bred sheep, bred and maintained in tick-free condinons, were inoculated intravenously with 1'0 ml of a 10 per cent dilution of the stabilate in phosphate buffered saline (PBS) at pH 7"2. The sheep were kept under observation for up to 21 days and rectal temperatures and blood samples taken every day.
Collection of Blood Samples Peripheral blood samples were collected in EDTA-coated tubes and smears stained with Giemsa were examined for the presence of cytoplasmic inclusions in the granulocytes and monocytes and to establish the differential white cell count. The day the organisms were first detected was regarded as the first day of parasitaemia (Woldehiwet, 1987).
Preparation of Infected Leucocytes and Plasma Blood samples were collected during the peak of parasitaemia in heparin-coated containers as described previously (Woldehiwet, 1987). The plasma was separated by centrifugation. A volume of cold PBS equal to the volume of plasma was added to the packed cell fraction before adding cold, sterile distilled water to lyse the red blood cells. Isotonicity was restored by adding 2"7 per cent NaC1 solution in PBS as described by Carlson and Kaneko (1973). After washing three times with a sucrose phosphate glutamate solution (SPG) (Bovarnick, Miller and Snyder, 1950), the cell pellet was resuspended in SPG and the cells treated with one cycle ultrasonic vibration at maximum amplitude for 20 seconds (Vibra-Cell, Sonics & Materials Inc., Danbury, U.S.A.) to disintegrate the leucocytes and release the rickettsiae. The disintegrated cells were centrifuged at 1000 Xg for 30 min to clear debris. The saved plasma and the supernate of the disintegrated cells were separately centrifuged at 3000 x g at 4*C for 30 min. The supernates were again centrifuged at 80 000 x g for 30 min. The pellets were washed with SPG twice and finally resuspended in 1 to 2 ml ofSPG, R P M I 1640 medium or HEPES/saline buffer. These suspensions of cell-free rickettsiae were used for infectivity assays and to purify the rickettsiae further by Percoll and Renografin density centrifilgation or by affinity chromatography.
Separation of C. phagocytophila by Percoll Gradient Suspensions of the reconstituted pellet (rickettsiae) were carefully layered onto a tube filled with 50 per cent Percoll (Pharmacia Fine Chemicals, Uppsala, Sweden). After centrifugation for 15 min at 30 000 X g with maximal acceleration and half maximal braking in a fixed angle rotor in a Sorvall ultracentrifuge (Sorvall Instruments, Newtown, Connecticut, U.S.A.), fbur 9-ml fractions with buoyant densities of 1"139 to 1,098, 1.063 to 1'057, 1"057 to 1.051 and 1"051 to 1"018 were collected. Commercial density marker beads (Sigma Chemicals Co., Ltd, Poole, U.K.) were used to establish the buoyant densities of the various fractions. These fractions were assayed for infectivity and rickettsial purity by electron microscopy.
Purification of TBF Agent
433
Separation of C. phagocytophila by Renografin Density Gradient Rickettsial suspensions in SPG were layered on to a discontinuous Urografin (Urografin 370, Schering AG, Germany) gradient in SPG buffer with a final concentration of 10, 20 and 30 per cent Urografin. After the tubes were centrifuged at 45 000 x g for 1 h at 4*C, two visible bands were formed and three fractions, two of which contained the visible bands, were collected. One millilitre volumes containing the visible bands were weighed to establish their buoyant densities. The rickettsiae contained in these fractions were pelleted by centrifugation at 80 000 x g for 30 min, washed twice in SPG to remove Urografin and assayed for infectivity and rickettsial purity by electron microscopy.
Isolation of C. phagocytophila by AJ]inity Chromatography The method of purifying the related rickettsia Cowdria ruminantium by wheat germ lectin chromatography described by Viljoen et al. (1985) was followed. Briefly, a gel of wheat germ lectin Sepharose 6MB was generated in a 10 cm column with 100 ml of 0.1 M Tris-HC1, 0"5 ~ NaC1, 0'02 per cent (w/v) NaNa, pH 8"5, followed by 100 rnl of 0" 1 M sodium acetate, 0.5 M NaC1, 0.02 per cent NaNa, pH 4.5 and equilibrated with 100 ml of HEPES buffer containing 0"02 per cent NaN 3. After washing the column with 500 ml of HEPES buffer pH 7.4, to remove azide, 1 ml of rickettsial suspension was added and the column incubated for 2 h at room temperature. Non-adsorbed material was eluted with HEPES buffer. The column was pulsed with N-acetyl-D-glucosamine in HEPES (20 ml of HEPES with 2g of carbohydrate) at a flow of 40 ml per h to remove adherent rickettsiae. Elution of C. phagoeytophila was monitored by a single path UV monitor (Uvicord II, LKB, Stockholm, Sweden). The fi'actions containing proteins were pooled and used for assaying infectivity and purity by electron microscopy.
Infectivity of Cell:free C. phagocytophila For infectivity assays of the Percoll gradients, the sample was divided into four fractions, each of 0"1 ml, and four susceptible sheep were each injected with 0'1 ml. The Renografin gradient showed two clear bands. The gradient was therefore divided into three fractions and three susceptible sheep were each inoculated with 0" 1 ml, one for each fraction. Two susceptible sheep were inoculated with the concentrated material eluted from the lectin column and monitored for febrile reaction and parasitaemia. The infectivity assays were repeated in four separate experiments to purify rickettsiae by density gradient centrifugation and by affinity chromatography. Non-reacting animals were challenged with a stabilate of C. phagocytophila as described above. To ensure that the rickettsiae prepared from the plasma or the disintegrated leucocytes were infective before the purification processes, a susceptible sheep was inoculated with 0' 1 ml of the final suspension before purification by density gradient centrifugation or cellular affinity chromatography.
Viability of Cell-free C. phagocytophila The pelleted cell-free rickettsiae and purified infected leucocytes were separately resuspended in SPG with t0 .per cent DMSO and stored at - 114°C. One susceptible sheep was inoculated with a suspension of infected leucocytes or with cell-free rickettsiae after 1 and 2 weeks and after 6 months.
Electron Microsopic Studies Aliquots of Percoll and Renografin density gradient fi-actions were centrifuged at 80 000 x g for 30 rain in a swlng-out rotor in a Sorvall OTD-Combi ultracentrifuge to pellet the organisms. The sediments of each fraction were fixed in 2'5 per cent glutaraldehyde solution in PBS overnight, post.fixed in 2 per cent osmium tetroxide,
434
Z. W o l d e h i w e t e t al.
stained with 2 per cent uranyl acetate in 0"69 per cent maleic acid, dehydrated and embedded in polythene capsules in resin (Taab, Aldermaston, U.K.) polymerized at 60°C for 48 h, then stained with lead citrate and examined with an electron microscope (Philips, EM201). Results
Isolation of C. phagocytophila Materials prepared from disintegrated leucocytes obtained from infected sheep formed clearly visible bands in Percoll and Renografin gradients. After centrifugation on a Percoll density gradient, two faintly visible bands were formed. The lower band was formed at density of 1.098 to 1.120 g per cm a and the higher band was formed at a density of 1"018 to 1"033 g per cm a. Aliquots from the fraction with a buoyant density ranging from 1.139 to 1"063 had rickettsia-like structures (Fig. 1A). The structures had typical rickettsial cell membrane (CM) and outer membranes (OM) (Fig. 1B). After centrifugation on a 10 to 30 per cent discontinuous Renografin gradient, two clear bands were formed at densities of 1"14 g per cm a and 1.09 g per cm ~, respectively. Rickettsia-like structures were present in fractions which contained either of the bands (Fig. 2). These structures were similar to the intracellular forms of C. phagocytophila described earlier (Woldehiwet and Scott, 1982). Leucocyte preparations treated by lectin affinity chromatography also yielded similar rickettsia-like structures. When cell-free materials prepared from the plasma of infected blood were purified by Renografin density gradients, pure rickettsial structures were readily visible in fractions which contained a visible band at a buoyant density of 1"14 (Fig. 3A). These structures also had typical rickettsial CM and O M (Fig. 3B).
Infectivity of Cell-free C. phagocytophila All susceptible sheep infected with cryopreserved whole-blood or purified leucocytes containing C. phagocytophila reacted with fever and parasitaemia. Fresh, cell-free rickettsiae, prepared from both the plasma and disintegrated leucocytes, were also infective to susceptible sheep. After Percoll density gradient centrifugation of suspensions prepared from disintegrated leucocytes, all susceptible sheep inoculated with the fractions containing the lowest band, with a density of 1.139 to 1"063 g per cm a, developed fever and parasitaemia 4 to 6 days post-inoculation (PI). Fractions with densities lower than 1"063 g per cm 3 were not infective. After Renografin density gradient centrifugation, all sheep inoculated with the fractions containing the two bands (densities of 1.14g per cm a and 1.09g per cm a respectively) reacted with fever and parasitaemia 5 to 7 days PI. C. phagocytophila separated from disintegrated leucocytes by cellular affinity chromatography was infective to susceptible sheep on only one of four attempts.
Purification of TBF Agent
Fig. 1.
435
(A) Thin section of rickettsial structures purified from leucocytes infected with C. phagocytophila by Percoll gradient centri|hgation. × 33 000. (B) Percoll gradient purified rickettsia from leucocytes infected with C. phagocytophila showing cell membrane (CM) and outer membrane (OM). x 82 500.
Fig. 2. Thin section of rickettsial structures purified from leucocytes infected with C. phagoo,tophila by Renografin gradient centrifugatkm, x 33 000.
436
Fig. 3.
Z. W o l d e h i w e t et al.
(A) Thin section of rickettsial structures purified from the plasma of sheep infected with C. phagocytophila by R.enografin gradient eentrifilgation, x 30 000. (B) Renografin gradient purified rickettsia li'om the plasma of sheep infected with C. phagocytophila showing cell membrane (CM) and outer membrane (OM). x 75 000.
Viability of Cell-free C. phagocytophila Cell-free and cell-associated organisms remained infective to susceptible sheep when kept at - 1140C in SPG buffer with 10 per cent DMSO as a cryopreservative for 6 months, the longest period of storage examined.
Discussion Molecular and antigenic studies required pure and large numbers of the organisms to be studied. Antigenic and molecular studies of C. phago~,lophila have been hampered by the inability to obtain organisms free from host cells. Some of the obligate intracellular rickettsiae are difficult to purify because of their instability and fragility (Tamura et al., 1982). Density gradients of Renografin and PercolI have been found to be very useful steps in the purification of several obligate intracellular bacteria, including Chlamydia trachomatis (Howard et al., 1974), Rickettsia prowazekii (Dasch and Weiss, 1977) and Rickettsia lsutsugamushi (Tamura et al., 1982). C. phagocytophila is an obligate intracellular rickettsia belonging to the tribe Ehrlichiae in the family Rickettsiaceae. These tick-transmitted animal rickettsiae are difficult to cultivate in vitro and extremely labile, making their isolation problematic. In the present study, we demonstrated that pure and viable C. phagocytophila can be obtained from infected polymorphonuclear cells with Percoll and Renografin density gradients. We are not aware of any other report on the purification ofC. phagocytophila from infected cells but two groups of workers have success~hlly separated two
Purification of TBF Agent
437
members of the tribe Ehrlichiae by density gradient centrifugation. Neitz eL al. (1986) used Percoll density gradient centrifugation to isolate viable C. ruminantium from infected sheep brain and Dutta, Rice, Hughes, Savage and Myrup (1987) used Renografin density gradients to purify Ehrlichia risticii antigens for the enzyme-linked immunosorbent assay (ELISA). In the present study we found that purified C. phagocytophila remained viable if the Renografin gradients were prepared in SPG buffer while rickettsiae separated by Percoll gradients in either R P M I 1640 medium or SPG buffer were infective to susceptible sheep. No attempt was made to quantify infectivity or loss of infectivity since it would have required large numbers of susceptible sheep. Losses of infectivity in Renografin density gradients have been reported in R. tsutsugamushi (Tamura et al., 1982) which is regarded as being more fragile outside host cells than R. typhi and R. prowazekii which retain most of their infectivity after separation with Renografin (Dasch and Weiss, 1977). Although the plasma was found to be a good source of infective C. phagocytophila and the rickettsial structures isolated fi~om the plasma had no apparent cellular contamination, we cannot be certain that the rickettsial structures present in the plasma were all C. phagocytophila. Other rickettsiae, such as Eperythrozoon ovis, which intact the erythrocytes can be present free in the plasma (Kreier and Ristic, 1963; McKee, Ziegler and Giles, 1973). E. ovis has often been presented as a rickettsia causing inapparent infections in sheep which may complicate experimental work in sheep (Littlejohns, 1960). The first reported infection of sheep with E. ovis occurred after the intravenous inoculation of sheep with blood containing Cowdria ruminantium the causative agent ofheartwater (Neitz, Alexander and du Toit, 1934). Foggie and Nisbet (1964) also reported that blood stabilates of C. phagocytophila may contain E. ovis. However, in the present study, the erythrocytes of the blood sample from which the plasma was prepared did not appear to harbour E. ovis. Furthermore, Foggie and Nisbet (1964) had reported that, after experimental infection with both agents, sheep which were heavily infected with C. phagocytophila had little or no E. ovis on their erythrocytes. Nevertheless, for antigenic and molecular studies of C. phagocytophila, rickettsiae isolated from purified infected neutrophils should be used because they are less likely to be contaminated by E. ovis. Now that we have successfully demonstrated the means of obtaining viable C. phagocytophila free of host cells, it is intended to study the antigenic components of the organism which stimulate host immune responses. Acknowledgments
This work was supported by the Wellcome Trust. The technical support of A. Joplin and L. Hill is gratefully acknowledged. References
Bovarnick, M. R., Miller, J. C. and Snyder, J. C. (1950). The influence of certain salts, amino acids, sugars and proteins on the stability of Rickettsiae. Journal of Bacteriology, 59, 509-522. Carlson, G. P. and Kaneko, J.J. (1973). Isolation ofleukocytes from bovine peripheral blood. Proceedings of Experimental Biology and Medicine, 142, 853-856.
438
Z. W o l d e h i w e t
et
al.
Dasch, G. A. and Weiss, E. (1977). Characterization of the Madrid E strain of Rickettsia prowazekii purified by Renografin density centrifugation. Infection and Immunity, 15, 280-286. Dutta, S. K., Rice, R. M., Hughes, T. D., Savage, P. K. and Myrup, A. C. (1987). Detection of serum antibodies against Ehrliehia risticii in potamic horse fever by enzyme-linked immunosorbent assay. Veterinary Immunology and Immunopathology, 12, 84-92. Foggie, A. (1951). Studies on the infectious agent of tick-borne fever in sheep. Journal of Pathology and Bacteriology, 63, 1-15. Foggie, A. and Nisbet, D. I. (1964). Studies on Eperythrozoon infections in sheep. Journal of ComparativePathology, 74, 45-61. Foster, W. N. M. and Cameron, A. E. (1970). Observations on ovine strains of tick-borne fever. Journal of ComparativePathology, 80, 429-436. GiImour, N . J . L . , Brodie, T. A. and Holmes, P. H. (1982). Tick-borne fever and pasteurellosis in sheep. VeterinaryRecord, 111, 512. Gronstol, H. and Ulvund, M.J. (1977). Listeric septicaemia in sheep associated with tick-borne fever. Acta VeterinariaScandinavica, 18, 575-577. Howard, L., Orenstein, N. S. and King, N. W. (1974). Purification on Renografin density gradients of Chlamydia trachomatis grown in the yolk sac of eggs. Applied Microbiology, 27, 102-106. Kreier, J. P. and Ristic, M. (1963). Morphologic antigenic and pathogenic characteristics of Eperythrozoon or# and Eperythrozoon wenyoni. American Journal of Veterinary Research, 24, 488-500. Littlejohns, I. R. (1960). Eperythrozoonosis in sheep. Australian VeterinaryJournal, 36, 261-265. MacLeod, J. and Gordon, W. S. (1933). Studies on tick-borne fever of sheep. I. Transmission by the tick Ixodes rieinus with a description of the disease produced. Parasitology, 25, 273-283. McKee, A. E., Ziegler, R. F. and Giles, R. C. X. (1973). Scanning and transmission electron microscopy of Haemobartonella canis and Eperythrozoonovis. AmericanJournal of Veterinary Research, 34, 1196-1201. Neltz, A. W. H., Viljoen, G. J., Benzuidenhout, J. D., Oberem, P. T., PutteriU, J. F., Verschoor, J. A., Visser, L. and Vermeulen, N. M. (1986). Isolation of Cowdria rumlnantium by means of Percoll density gradient centrifugation and detection by enzyme-linked immunosorbent assay. OnderstepoortJournal of VeterinaryResearch,53, 63-67. Neitz, W. O., Alexander, R. A. and du Toit, P.J. (1934). Eperythrozoonovis (sp. nov.) infection in sheep. OnderstepoortJournal of Veterinary Science, 3, 263-274. Tamura, A., Urakami, .H. and Tsuruhara, T. (1982). Purification of Rickettsia tsutsugamushi by Percoll density gradient centrifugation. Microbiology and Immunology, 28, 873-882. Taylor, A. W., Holman, H. H. and Gordon, W. S. (1941). Attempts to reproduce the pyaemia associated with tick-bite. VeterinaryRecord, 53, 339-344. Viljoen, G. J., Vermeulen, N. M. J., Oberem, P. T., Prozesky, L., Verschoor, J. A., Bezuidenhout, J. D., Putterill, J. F., Visser, L. and Neitz, A. W. H. (1985). Isolation ofCowdria ruminantium by cellular affinity chromatography and detection by an enzyme-linked immunosorbent assay. Onderstepoort Journal of Veterinary Research, 52, 227-232. Woldehiwet, Z. (1987). The effects of tick-borne fever on some functions of polymorphonuclear cells of sheep. Journal of ComparativePathology, 97, 481-485. Woldehiwet, Z. and Scott, G. R. (1982). Stages in the development of Cytoecetes phagocytophila, the causative agent of tick-borne fever. Journal of Comparative Pathology, 92, 469-474.
Reeeived, May 17th, ~Accepted, June 28th,
1991 ] 1991