Haemagglutination by Bovine Leukaemia Virus

Transcription

1 J. gen. Virol. (1982), 59, Printed in Great Britain 83 Key words: BL V/glycoprotein/haemagglutination/retrovirus Haemagglutination by Bovine Leukaemia Virus By HIROSHI SENTSUI,I*t RICHARD M. THORN, 1 YUJI KONO 2 AND JORGE F. FERRER 1 1 Section of Viral Oncology, Comparative Leukemia Studies Unit, School of Veterinary Medicine, University of Pennsylvania, New Bolton Center, Kennett Square, Pa , U.S.A. and 2 Equine Infectious Anemia Research Division, National Institute of A nimal Health, Yatabe, Ibaraki 305, Japan (Accepted 15 October 1981) SUMMARY Bovine leukaemia virus (BLV) was found to agglutinate mouse erythrocytes. Under optimal conditions, including the use of neuraminidase-treated erythrocytes, 200/tg/ml of BLV purified from the supernatant fluid of BLV-infected bat cells had haemagglutinating titres of about 512 units. BLV haemagglutination was drastically affected by ph and temperature; maximum agglutination occurred at ph 6 and 4 C. That the BLV haemagglutinin is a glycoprotein was suggested by the fact that trypsin, potassium periodate or neuraminidase, but not lipid solvents or phospholipase C, significantly reduced the haemagglutinating (HA) activity of purified BLV. Furthermore, purified BLV glycoprotein of mol. wt (gp51) had HA activity. The receptors for BLV on mouse erythrocytes were inactivated by proteolytic enzymes but not by sodium deoxycholate or potassium periodate. Neuraminidase treatment of erythrocytes increased their agglutinability fourfold. Haemagglutination is a relatively sensitive test for detecting BLV glycoprotein because 0.4 pg/ml of glycoprotein can be detected by this method. The ph and temperature sensitivity of the BLV HA reaction and specificity for mouse erythrocytes distinguish BLV from that of equine infectious anaemia virus and murine leukaemia virus, the other C type retroviruses known to have HA activity. INTRODUCTION Bovine ieukaemia virus (BLV) is regarded as the causative agent of the enzootic form of bovine lymphosarcoma (Ferrer, 1980). This virus has a number of characteristics not commonly found among C type retroviruses (Ferrer, 1972, 1980) and it may be the prototype of a new virus group. BLV has a major internal protein of mol. wt (p25) (Gilden et al., 1975; McDonald & Ferrer, 1976; McDonald et al., 1976) and a glycoprotein of mol. wt (gp51) which seems to be a surface component (Onuma et al., 1976; Devare & Stephenson, 1977). Electron microscope studies have shown that the surface of BLV has knob-like structures (Weiland & Uebersch~ir, 1976) similar to those found on viruses that agglutinate red blood cells (RBC). The only mammalian retroviruses known to have haemagglutinating (HA) activity are murine leukaemia virus (MuLV) (Sch~ifer & Szanto, 1969) and equine infectious anaemia virus (EIAV) (Sentsui & Kono, 1976). The haemagglutinins of these viruses are glycoproteins (Witter et al., 1973; Sentsui & Kono, 1981). This paper demonstrates that BLV agglutinates mouse erythrocytes and describes optimal conditions for BLV-induced haemagglutination. The characterizations of the BLV haemagglutinin and corresponding erythrocyte receptors are also described. "~ Present address: National Institute of Animal Health, Yatabe, Ibaraki 305, Japan /82/ $02.00 (~ 1982 SGM

2 84 H. SENTSUI AND OTHERS METHODS Source of virus and preparation of haemagglutinins. BLV was obtained from the supernatant fluid of cell line BLV-bat 2 clones, initiated by infecting a bat lung monolayer cell culture (Tb~Lu) with BLV (Graves & Ferrer, 1976; Diglio & Ferrer, 1976). This cell line produces BLV continuously and is free of contamination with the bovine viral diarrhoea virus and other adventitious agents. Supernatant fluids from cell line BLV-bat 2 clone s were collected at 12-h intervals, clarified by centrifugation at g for 20 min and concentrated about 100 times using the Pellicon Cassette System (Millipore). This virus suspension was further concentrated and purified by centrifugation on to a sucrose cushion as previously described (Gupta & Ferrer, 1978). To remove sucrose, the virus band at the interphase was diluted with 2 vol. 10 mi-tris, 100 mm-naci and 1 mm-edta solution and centrifuged for 150 min at g. The virus pellet was resuspended in 0.15 M-NaCI, adjusted to 200 #g/ml, unless otherwise indicated, and ultrasonicated (model W-200F Heat Systems; Ultrasonics, Inc., Plainview, N.Y., U.S.A.) for 2 min at 70% relative power. The concentration of BLV protein was estimated by the method of Lowry et al. (1951) with the use of bovine serum albumin as the standard. BLV p25 and gp51 were purified to homogeneity as previously reported (McDonald & Ferrer, 1976; Devare & Stephenson, 1977). Haemagglutinin test. A 50 #1 sample of 0-5 % erythrocytes which were suspended in a ph 6 phosphate buffer solution (0.117 M-KH2PO4, M-Na2HPO 4 and 0.15 M-NaCI) (PBS-pH 6) was added to microtitre V plates (Dynatech Laboratories) containing an equal volume of BLV serially diluted in 0.15 M-NaC1 containing 0.2 #g/ml gelatin. The mixture was incubated at 4 C for at least 2 h. HA units were calculated as the reciprocal of the highest virus haemagglutinin dilution showing complete haemagglutination. Treatment of BLV with chemicals and enzymes. A 50 al amount of BLV suspension (1 mg/ml) was mixed with an equal volume of the reagents examined. After incubation for the time and at the temperatures shown in Table 3, each mixture was diluted with 0.4 ml 0.15 M-NaC1 and assayed for HA activity. Trypsin (bovine pancreas, type 2, 2 x crystallized), protease (from Streptomyces griseus purified type VI), trypsin inhibitor (lyophilized from soybean), neuraminidase (from Clostridium perfringens type I, lyophilized) and phospholipase C (from Bacillus cereus) were all obtained from Sigma. Ultraviolet irradiation ofblv haemagglutinin. Samples (0,5 ml) of five separate BLV haemagglutinin preparations containing 128 HA units were placed in 30 mm Petri dishes and irradiated by a 15 W germicidal lamp at a distance of 30 cm. During exposure the dishes were shaken continuously. Samples were taken 2, 5, 10, 20 and 30 min after initiating irradiation and assayed for HA activity. Thermal stability ofbl V haemagglutinin. BLV haemagglutinin was incubated at 0 C (in ice), 37 C and 56 C for the times shown in Table 2. Immediately after incubation each sample was titrated for HA activity. Treatment of RBC with various reagents. Equal volumes of 5 % suspension of mouse RBC and various reagents, both diluted in 0.15 M-NaCI, were mixed together. The reagents, incubation times, temperature and final reagent concentrations are shown in Table 5. After treatment, the RBC were washed three times with 0-15 M-NaCI, resuspended in the PBS-pH 6, and tested for HA activity. Sera used in the haemagglutination inhibition (HI) tests. Sera were collected from sheep before and 10 weeks after inoculation with supernatant fluid of cell line BLV-bat 2 clone s during the first week of life and tested for BLV gp51 and p25 antibodies by radioimmunoassay (Gupta & Ferret, 1981). The characteristics of the rabbit antiserum to BLV p25 antigen have been described previously (McDonald & Ferrer, 1976). To remove non-specific HA inhibitors, 0.2 ml serum samples were mixed with 0.1 ml 0.01

3 Haemagglutination by BL V 85 M-potassium periodate (KIO4) in 0.15 M-NaCI solution, and incubated for 60 min at 37 C. The activity of KIO 4 was inhibited by adding 0.1 ml of 2% glycerine in 0.15 M-NaCI solution for 10 min at room temperature. The sample was heated at 56 C for 30 min, mixed with 0.4 ml 25 % kaolin solution in 0.15 M-NaCI, kept at room temperature for 20 min with occasional shaking, and centrifuged at 3000 g for 20 rain. To adsorb isohaemagglutinin, the serum sample was then mixed with 0.1 ml of washed and packed mouse RBC and incubated at 4 C for 20 rain with occasional shaking. Following adsorption, the sample was centrifuged for 20 min at 3000 g. The resulting supernatant fluid was used in the HI test at a 1 : 4 dilution. For the HI test, ml of the virus suspension (4 HA units) were mixed with ml of serum serially diluted with 0.15 M-NaC1 and 0-02% gelatin. After incubation for 60 min at 37 C, 0.05 ml of a 0-5% RBC suspension in PBS-pH 6 was added. The reciprocal of the highest serum dilution showing complete inhibition of haemagglutination was considered the HI titre. RESULTS Optimal conditions for BL V-induced haemagglutination Purified BLV was incubated at 4 C with horse, cattle, cat, rabbit, guinea-pig, hamster, mouse, goose or chicken RBC. Only mouse RBC were agglutinated. Haemagglutination was not induced by supernatant fluid from a BLV-free bat cell line. BLV agglutinated erythrocytes from all mouse strains tested (DDD, C57BL, ICR, C3H/He, ddy). In subsequent experiments we routinely used erythrocytes from the ddy strain. HA activity was increased 8 to 16 times after purified virus material was treated with Triton X-100 or ultrasonicated (Table 4). Therefore, all virus suspensions were ultrasonicated before use as a haemagglutinin. The optimal concentration of erythrocytes for haemagglutination was 0.5 %. Non-specific haemagglutination and haemolysis were observed with erythrocytes stored for more than 3 days. The HA reaction was optimal at 4 C and at ph 6.0 to 6-2. The reaction decreased rapidly at higher temperatures (Table 1) or ph values (Fig. 1); ph values less than 6.0 caused non-specific agglutination or lysis. Characterization of virus haemagglutinin The HA activity of BLV haemagglutinin is stable for days at 4 C but only hours at 37 C and minutes at 56 C (Table 2). The HA activity was also significantly reduced by u.v. irradiation for 10 rain or longer (Fig. 2). Table 3 shows the results of experiments in which the effect of several reagents on BLV HA activity was examined. Some of the reagents could not easily be separated from BLV or inactivated without influencing HA activity. Therefore, before HA testing, the BLV-reagent mixture was diluted to a concentration at which the reagent would not affect RBC. To determine the effects of the reagents on RBC, they were mixed with 1 mg/ml bovine albumin, incubated, diluted and then mixed with mouse RBC. The reagent-albumin mixture did not cause haemagglutination or haemolysis. Furthermore, the HA titre was not affected if the mixture of haemagglutinin and reagents was diluted immediately without incubation. HA activity was not significantly affected by ether, chloroform, sodium deoxycholate, Triton X-100 or phospholipase C. Thus, the BLV agglutinin is not likely to have an active lipid component. In contrast, HA activity was inhibited by trypsin, KIO4 and neuraminidase. Haemagglutination with BL V gp51 As shown in Table 4, BLV gp51 purified to homogeneity had strong HA activity. In contrast, BLV p25 did not have any significant HA activity.

4 86 H. SENTSUI AND OTHERS I I I I I I , 1 I I I I < -r < "~0 ph values Fig. 1 ~ 64 "1" I 0 2 I I I I Time of u.v. irradiation (min) Fig. 2 Fig. 1. Effect of ph on the HA titre of BLV. The haemagglutination assay (see Methods) was done at the indicated ph values. Fig. 2. Effect of u.v. irradiation on HA activity. BLV (200 pg/ml) was irradiated by a 15 W germicidal lamp at a distance of 30 cm for the indicated periods. Table 1. Effect of temperature on the HA reaction* Incubation temperature ( C) HA units * HA reactions were done at ph 6. Table 2. Thermal stability of BL V haemagglutinin Duration at Duration at Duration at 56 C (min) HA units 37 C (h) HA units 4 C (days) HA units Characterization of the HA receptor on mouse erythrocytes Mouse RBC were treated with several enzymes and chemicals, washed and used for haemagglutination (Table 5). Treatment with agents known to inactivate proteins inhibited haemagglutinability of the erythrocytes, whereas reagents that inactivate lipids or carbohydrates had no effect. Thus, the reacting part of the RBC receptor for the BLV haemagglutinin seems to be consistent with a protein rather than carbohydrate or lipid components. Neuraminidase-treated RBC were much more sensitive than untreated cells. HI antibody in BL V-infected sheep HA activity was inhibited by sera from sheep which were infected with BLV, whereas sera from the same sheep before infection or anti-blv p25 rabbit serum (McDonald & Ferrer,

5 Haemagglutination by BL V Table 3. Haemagglutinating activity of BL V after treatment with various agents Treatment* ~k ( Agent Final concn. Condition HA units Control 60 min at 37 C 64 Trypsin~ 200 units/ml~ 60 min at 37 C 8 Formaldehyde 1.0% 60 min at 37 C 4 Formaldehyde 0.25 % 60 min at 37 C 8 Glutaraldehyde 0.3% 60 min at 37 C 4 KIO4 1 x 10-5 M 60 min at 37 C 4 Neuraminidase 0.2 units/ml 60 min at 37 o C 16 Phospholipase C 0.2 units/mill 60 min at 37 C 128 Sodium deoxycholate 0.25% 60 min at 37 C 64 Triton X % 30 min at 37 C 128 Ether 50% 60 min at 22 C 64 Chloroform 50% 60 min at 22 C * BLV was treated with the equal volume of indicated agent; the mixture was then diluted five times and used to agglutinate mouse erythroctyes. t Soybean trypsin inhibitor was added after treatment. The inhibitor alone did not influence HA activity. $ Determination made using Nct-benzoyl-L-arginine. Determination made using NAN-lactose. (I Determination made using egg yolk lecithin. Table 4. HA activity of purified whole BL V, BL V gp51 and BL V p25 Sample Treatment Specific HA activity* BLV None 8 BLV Sonicationt 64 ~ 128 BLV Triton X- 100:~ 128 BLV gp ~ 640 BLV p25 <8 * HA units per 200 gg protein/ml. Ultrasonicated for 2 min at 70 % relative power. ~: Incubated for 30 min at 37 C with 0.1% Triton X Haemolysis was observed at dilutions less than 1 : 8 because of detergent present in the BLV p25 preparation. Table 5. Agglutinability of mouse erythrocytes by BL V after treatment with various agents Erythrocyte treatment* A Agent Final concn. Condition HA units None 3 h, room temperature 128 None 1 h, 37 C 128 Trypsin 100 units/ml~" 1 h, 37 C 4 Protease 2.0 units/ml:]: 1 h, 37 C 8 Formaldehyde 1.0% 3 h, room temperature 8 Formaldehyde 0.25 % 3 h, room temperature 32 Sodium deoxycholate 0.02% 1 h, 37 C 128 KIn SM 1 h, 37 C 128 Neuraminidase 0.2 units/ml I h, 37 C 512 * Mouse erythrocytes were treated with the indicated reagents, washed and used in the HA test. ~ Determination made using Nct-benzoyl-L-arginine. :[- Determination made using casein. Determination made using NAN-lactose.

6 88 H. SENTSUI AND OTHERS Table 6. Immunological specificity of BL V haemagglutination Pre-serum BLV antibody* Post serum BLV antibody* A. & r f -~ p25 gp51 HI titre+ p25 gp51 HI titret Sheep 827~: -- - < Sheep 828:1: - - < Sheep 830~: - - < Rabbit anti-p <4 * As detected by radioimmunoassays. t Reciprocal of serum dilution that showed definite inhibition of haemagglutination. These animals were injected with BLV derived from B LV bat 2 clone 6 supernatant fluid. The rabbit was immunized with purified BLV p ) did not inhibit HA activity (Table 6). Non-specific HA inhibitors or isohaemagglutinins were found in all sera tested, but the activity of these inhibitors was reduced to undetectable levels by treatment with KIO+ and kaolin, or adsorption with mouse RBC. DISCUSSION Our results show that BLV agglutinates mouse RBC under appropriate conditions. The other retroviruses known to have HA activity are MuLV and EIAV. BLV-induced haemagglutination increased as the ph decreased to 6.0. Mouse RBC are lysed or agglutinated at lower ph. A similar ph dependence has been described for togaviruses and rhabdoviruses (Howe & Lee, 1972), but not for MuLV or EIAV. Haemagglutination was remarkably reduced at 37 C (Table 1), but this was probably not due to heat inactivation of BLV haemagglutinin because incubation of BLV alone for 1 h at 37 C (Table 2) did not affect its subsequent HA activity at 4 C. Nor was 37 C inactivation due to virus-associated neuraminidase activity, as has been shown in myxoviruses (Howe & Lee, 1972), because BLV-agglutinated RBC were able to reagglutinate at 4 C after disaggregation at 37 C (not shown). The test that trypsin, KIO4 or neuraminidase significantly reduced the HA activity of BLV indicated that the viral haemagglutinin is a glycoprotein. Furthermore, purified BLV gp51 efficiently agglutinated mouse RBC. One of the chemical modifications produced in proteins by u.v. irradiation is disulphide bridge breakage (Smith & Hanawalt, 1969). Thus, since u.v. irradiation inhibits the HA activity of BLV, it appears that a disulphide bond is important for the activity of the virus haemagglutinin. HA activity of purified BLV particles was increased by Triton X-100 treatment or sonication (Table 4). This could be due to disaggregation of the virus cluster. Data on the effect of various treatments on the agglutinability of RBC by BLV suggest that the activity of the erythrocyte receptor is associated primarily with the proteins of the erythrocyte surface rather than a lipid or carbohydrate moiety (Table 5). The enhancement of HA activity by neuraminidase could be due to the unmasking of receptors or the reduction of RBC surface charge. Protein-reactive agents and neuraminidase have similar effects on the receptors on horse RBC for EIAV (Sentsui & Kono, 1981). The only other leukaemia virus known to have HA activity haemagglutinate is MuLV. In contrast to BLV, MuLV agglutinates sheep, mouse, chicken and goose RBC, and it does so only after treatment with neuraminidase and phospholipase C (Sch~fer & Szanto, 1969; Witter et al., 1973). The haemagglutinin of EIAV from horse leukocyte cultures has chemical properties similar to those of BLV. However, EIAV agglutinates horse RBC and it is not known to cause leukaemia or lymphosarcoma.

7 Haemagglutination by BL V 89 Specific haemagglutinin inhibition by sheep sera containing antibodies to BLV was used to determine the immunological specificity of BLV haemagglutinin (Table 6). It is not likely that the inhibition was due to antibodies against other bovine viruses because the sheep were exposed only to BLV from bat clone cell culture fluid. Although these sera had non-specific HA inhibitors or isohaemagglutinins, these factors were removed by treatment with KIO4, kaolin and adsorption with mouse RBC. Cattle sera also had isohaemagglutinins against mouse RBC but they could not be removed by the same procedures. It remains to be determined whether HI antibodies have the same specificity as the gp antibodies. Although the radioimmunoassay is a more sensitive technique than the haemagglutination test, the HA and HI tests, which are simple, faster and which do not require a radioisotope, have a possibility of being used to detect BLV and HI antibody after further technical improvement. The secretarial help of Mrs Betty Thompson is gratefully acknowledged. This work was supported by National Cancer Institute grant CA REFERENCES DEVARE, S. G. & STEPHENSON, J. R. (1977). Biochemical and immunological characterization of the major envelope glycoprotein of bovine leukemia virus. Journal of Virology 23, OlGLIO, C. A. & FERRER, J. F. (1976). Induction of syncytia by the bovine C-type leukemia virus. Cancer Research 36, FERRER, J. F. (1972). Antigenic comparison of bovine type C virus with murine and feline leukemia virus. Cancer Research 32, FERRER, J. F. (1980). Bovine lymphosarcoma. Advances in Veterinary Science & Comparative Medicine 24, GILDEN, R. V., LONG, C. W., HANSON, M., TONI, R., CHARMAN, H. P. & OROSZLAN, S. (1975). Characteristics of the major internal protein and RNA-dependent DNA polymerase of bovine leukaemia virus. Journal of General Virology 29, GRAVES, D. C. & FERRER, J. F. (1976). In vitro transmission and propagation of the bovine leukemia virus in monolayer cell cultures. Cancer Research 36, GUPTA, P. & FERRER, J. F. (1978). Detection of a precursor of bovine leukemia virus structural protein in purified virions. Annales de Recherches Veterinaires 9, GUPTA, P. & FERRER, J. F. (198 I). Diagnosis of BLV infection in cattle; a critical evaluation of various serological and direct methods. International Journal of Cancer 28, HOWE, C. & LEE, T. (1972). Virus-erythrocyte interactions. Advances in Virus Research 17, LOWRY, O. H., ROSEBROUGH, N. J., FARR, A. L. & RANDALL, R. J. (1951). Protein measurement with the Folin phenol reagent. Journal of Biological Chemistry 193, McDONALD, H. C. & FERRER, J. F. (1976). Detection, quantitation and characterization of the major internal antigen of the bovine leukemia virus by radioimmunoassay. Journal of the National Cancer Institute 57, McDONALD, H. C., GRAVES, D. C. & FERRER, J. F. (1976). Isolation and characterization of an antigen of the bovine C-type virus. Cancer Research 36, ONUMA, M., OLSON, C. & DRISCOLL, O. M. (1976). Properties of two isolated antigens associated with bovine leukemia virus infection. Journal of the National Cancer Institute 57, SCH.AFER, W. & SZANTO, J. (1969). Studies on mouse leukemia virus. I1. Nachweis eines virus spezifiscben hamagglutinins. Zeitschriftfiir Natursforschung 24B, SENTSUI, H. & KONO, Y. (1976). Hemagglutination by equine infectious anemia virus. Infection and Immunity 14, l. SENTSUI, U. & KONO, V. (1981). Hemagglutination by several strains of equine infectious anemia virus. Archives of Virology 67, SMITH, K. C. & HANAWALT, P. C. (1969). Molecular Photobiology Inactivation and Recovery, pp New York & London: Academic Press. WEILANO, F. & UEBERSCH~R, S. (1976). Ultrastructural comparison of bovine leukemia virus (BLV) with C-type particles of other species. Archives of Virology 52, WITTER, R., FRANK, H., MOENNIG, V.. HUNSMANN, G., LANGE, J. & SCH.~FER, W. (1973). Properties of mouse leukemia virus. IV. Hemagglutination assay and characterization of hemagglutinating surface components. Virology 54, (Received 20 July 1981)