Journal of Medical Virology 37:39-47 (1992)
Detection of Multiple Epstein-BarrViral Strains in Allogeneic Bone Marrow Transplant Recipients J.W. Gratama, E.T. Lennette, B. Lonnqvist, M.A.P. Oosterveer, G. Klein, 0.Ringden, and I. Ernberg Department of Immunology, Daniel den Hoed Cancer Center, Rotterdam, The Netherlands (J.W.G., M.A.P.O.); Department of Immunohematology and Blood Bank ( J .W.G.), Leiden, The Netherlands; Virolab Inc., Berkeley, California (E.T.L.); Departments of Medicine (B.L.), Clinical Immunology and Transplantation Surgery (O.R.), Huddinge Sjukhus, Huddinge, Sweden; and Department of Tumor Biology, Karolinska Institute, Stockholm, Sweden (G.K.,I.E.) We have previously shown i n 3 allogeneic bonemarrow transplant (BMT) recipients that complete replacement of recipient marrow was associated with the elimination of the pretransplant Epstein-Barr virus (EBV) strain of the recipient. To study the kinetics of EBV elimination and reinfection in more detail, we have performed a longitudinal study of BMT recipients combining serology, virus isolation from mouthwashes and peripheral blood, and EBV strain characterization. Oropharyngeal EBV excretion was found to persist after the cytoreductive therapy prior to BMT, whereas EBV-carrying cells in the blood were detected only after 5 weeks following BMT. During the first month post-BMT, 2 different EBV strains could be isolated from sequential mouthwashes of 3 patients. The initial strains were found to persist up to 7, 21, and 29 days postBMT, whereas the subsequent strains appeared at 21, 42, and 34 days post-BMT, respectively. Thus, the original EBV strain may persist only for a limited time after BMT, and the oropharyngeal epithelium may be reinfected by a new EBV strain from the blood within 3 weeks. With respect to the coexistence of multiple EBV strains, 2 patterns were evident. From the day 62 mouthwash of 1 patient, 1 Type A and 1 Type B strain were isolated. From the day 180 mouthwash of a second patient, a dominant Type A strain was recovered, together with 6 ”variant” strains that differed from each other by only a single EBNA protein (EBNA 1). This pattern may be explained by viral recombinations during replication, which may form the basis for the vast polymorphism of EBV observed in unrelated individuals. o 1992 Wiiey-Liss, Inc.
KEY WORDS: serology, viral strain typing, EBV elimination kinetics, EBV reinfection 0 1992 WILEY-LISS, INC.
INTRODUCTION The combined size variations of Epstein-Barr virus (EBV)nuclear antigens EBNA 1,2,3,4, and 6 in irnrnunoblots of lymphoblastoid cell lines (LCL) have been used to distinguish between different EBV strains [Ernberg et al., 1986; Gratama et al., 1988, 1990a, 1990b;Rowe et al., 1989; Sculley et al., 1987,19901.We have used this “Ebnotyping” approach to distinguish between donor and recipient virus in bone-marrowgrafted patients [Gratama et al., 1988, 1990al. Allogeneic bone-marrow transplantation (BMT) aims to replace the recipient’s diseased hematopoietic tissue with that of the donor. Our studies, combining EBV serology with Ebnotyping, showed that repopulation of a seropositive recipient with bone marrow of a seronegative donor caused long-lasting seronegativity and made her susceptible to later infection with an exogenous EBV strain. In 2 other patients, repopulation with marrow from a seropositive donor was followed by the disappearance of the pretransplant EBV strain and the establishment of persistent infection with the exogenous virus, which could be traced back to the marrow donor in 1patient [Gratama et al., 1988,1990al. Graft failure in a fourth patient was associated with the persistence of recipient lymphocytes and the recipient’s pretransplant EBV strain [Gratama et al., 1990al. Taken together, these studies suggest that the hematopoietic compartment, and not the oropharyngeal epithelium, is the permanent reservoir of the virus. Observations on patients on long-term acyclovir treatment, which blocked oropharyngeal EBV replication but did not decrease the frequency of EBV-carrying lymphocytes in the peripheral blood [Yao et al., 19891,support the contention that the persistence of infected lymphocytes is independent of the oropharyngeal lytic infection. The disappearance of the recipient’s EBV strain and the appearance of the donor’s strain provides an interAccepted for publication April 30,1991. Address reprint requests to J a n W. Gratama, M.D., Department of Immunology, Daniel den Hoed Cancer Center, P.O. Box 5201, 3008 AE Rotterdam, The Netherlands.
40
esting opportunity to study the kinetics of establishment of EBV infection of B lymphocytes and oropharyngeal epithelium, and to study the “traffic” of the virus between these compartments. We report now on a prospective study of the appearance of EBV-carrying cells in the blood and changes in salivary EBV excretion patterns of allogeneic BMT recipients in relation to serology and viral strain typing.
PATIENTS AND METHODS Patients Thirty-two patients were entered into a longitudinal study, They received marrow grafts from a n identical twin (n = l ) , HLA-A,B,C,DR-identical siblings (n = 23), from other related donors (n = 6), or from HLA-matched unrelated donors (n = 2). The median age of the patients was 33 years (range, 2-58 years). The diagnostic indication for BMT was hematological malignancy in 29 patients and nonmalignant disease (severe aplastic anemia or metabolic disorders) in 3. Prophylaxis of acute graft-versus-host disease (GVHD) consisted of (1)T-cell depletion of the marrow grafts in 9 patients [Tollemar et al., 19891or (2) a combination of cyclosporin A and methotrexate in 22 patients [Storb et al., 19861. The recipient of identical-twin donor marrow did not receive any GVHD prophylaxis. Patients with hematological malignancies were prepared for BMT as follows: (1) those receiving T-cell-depleted grafts were treated with cyclophosphamide (60 mg/kg/ day x 2) followed by total-body irradiation (7.5 Gy in 1 session) coupled with total lymph-node irradiation (3 Gy in 2 sessions); (2) those receiving cyclosporin A and methotrexate were treated with cyclophosphamide (60 mg/kg/day x 2) followed by total-body irradiation (10 Gy in 1 session). The 3 patients with nonmalignant disorders were prepared for BMT using a combination of cyclophosphamide (2 g/m2/day x 4) and busulphan (4-5 mglkglday x 4). All patients were nursed in reversed isolation during the peritransplant period. Selective decontamination of their gastrointestinal tracts was attempted as described elsewhere [Ringden et al., 19861. Leukocytepoor erythrocyte and platelet concentrates were administered to maintain the hemoglobin level above 75 g/1 and the platelet count above 30 x 109/l.Thirty patients showed sustained engraftment and 2 patients died before marrow engraftment was documented. Antiviral drugs were given to 22 patients. Eleven patients received foscarvir as prophylaxis for active CMV infection. This drug was given as a continuous i.v. infusion from the start of the cytoreductive therapy until marrow engraftment. The dose was adjusted to reach plasma levels between 100 and 150 pg/l. This concentration inhibits the replication of most CMV strains in vitro. An additional patient received foscarvir therapeutically for proven active CMV infection. Five patients received acyclovir prophylactically and 6 therapeutically (250 mg/mz 2 or 3 times daily, respectively) against active HSV infection.
Gratama et al. Acute GVHD occurred in 23 patients: It was mild (Grade I) in 15 patients and moderate to severe (Grades 11-IV) in 8. Five patients developed limited chronic GVHD and 1extensive chronic GVHD. Fifteen patients died at a median time of 75 days post-BMT (range 10593 days): 9 from BMT-related causes and 6 from relapse of their original disease. Fifty-one patients were studied on a single occasion at a median of 727 days post-BMT (range 52-2562 days). They had received marrow grafts from a n identical twin (n = 11, HLA-A,B,C,DR-identical siblings (n = 46), or from other related donors (n = 4). Their median age was 21 years (range, 2-50 years). The diagnostic indication for BMT was hematological malignancy in 40 patients, severe aplastic anemia in 6, and other in 5. Chronic GVHD had developed in 15 patients: It was limited in 11and extensive in 4 patients. None of these patients received antiviral therapy a t the time of our study.
EBV Serology Cryopreserved serum samples from the patients and their donors were tested for IgM and IgG antibodies against EBV viral capsid antigen (VCA), for IgG and IgA antibodies against early antigens (EA) of the diffuse (D) variety, and for IgG antibodies against EA of the restricted (R) variety using indirect immunofluorescence [Henle and Henle, 1966, 19701. Antibodies against the nuclear antigens EBNA 1, 2A, 2B, and 6 were determined using anticomplement immunofluorescence (ACIF) [Reedman and Klein, 19731. DG-75 cells transfected with the BamHI K fragment of B95-8 EBV DNA served as a source of EBNA 1. DG-75 cells transfected with the BamHI Y fragment of B95-8 EBV and RAT-2B cells transfected with the BamHI Y fragment of Ag-876 EBV were used for the EBNA 2A and 2B antibody assays, respectively. DG-75 cells transfected with the BamHI E fragment of B95-8 EBV served as a source of EBNA 6 . Expression of the relevant antigens was controlled by ACIF and Western blotting, and only transfected lines in which the majority of the cells expressed the antigens were used for the antibody assays. Raji cells expressing EBNA 1, 2A, 3, and 4, and EBV-negative BJA-B cells served as positive and negative controls, respectively. Smears were made by spreading 5 p1 of a suspension containing lo7 cells over 6 x 30 mm coverslips. The smears were allowed to dry rapidly in a n air current and were kept for 2 hours at room temperature prior to fixation in a precooled (- 2 0 C ) mixture of equal volumes of acetone and methanol for 1minute. After fixation, the smears were kept in plastic boxes at -20°C until used. All sera from each patient were tested on the same occasion to exclude the possible influence of interassay variability. EBV Isolation EBV was isolated either from the oropharynx by mouthwashing or by establishing a n EBV-carrying LCL from the peripheral blood. Mouthwashes were obtained by gargling for 1minute with 10 ml RPMI 1640
Multiple EBV Strains in BMT Recipients medium and were cryopreserved after removal of bacteria and debris, and the addition of 10% (v h ) fetal calf serum. For the lymphocyte transformation assays, 12 million cord blood mononuclear cells were incubated with 3-ml aliquots for 3-6 hours, washed, and cultured for 8-12 weeks in 96-well flat-bottomed microtiter plates [Gratama et al., 19881. To obtain spontaneously outgrowing lymphoblastoid cell lines (LCL), peripheral blood mononuclear cells (PBMC) were obtained from heparinized blood samples and cultured in 96-well flatbottomed microtiter plates for 8-12 weeks [Gratama et al., 19881. Between 1and 2 weeks after the initiation of the cultures, 100,000 fresh cord blood lymphocytes were added to each well in order to rescue EBV released from virus-carrying cells [Lewin et al., 19871. From these 2 types of cultures, wells with growing cells were harvested and the resulting LCL were expanded. From the 32 longitudinally studied patients, blood samples were obtained between 21 and 8 days prior to BMT, and at various time points between 30 and 180 days post-BMT. Mouthwashes were obtained between 21 and 8 days prior to BMT, once weekly between days 0 and 28 post-BMT, and concurrent with blood sampling thereafter. PBMC and mouthwashes were obtained from the marrow donors within 2 weeks prior to BMT. From the 51 patients studied on a single occasion, blood and mouthwashes were obtained between 100 and 2246 days post-BMT.
Immunoblotting Aliquots corresponding to 1million LCL dissolved in electrophoresis sample buffer were loaded per lane for electrophoresis in gels containing 7.5% polyacrylamide and sodium dodecyl sulfate [Gratama et al., 19881. The following molecular mass markers were used: myosin (200 kD), E. coli beta-galactosidase (116 kD), rabbit muscle phosphorylase b (97 kD), bovine serum albumin (66 kD), and hen egg-white ovalbumin (43 kD) (BioRad, Richmond, CA). The electroblotting procedure and the immunological detection of the EBNA proteins on the nitrocellulose filters was performed as described previously [Gratama et al., 19881.Serum from a patient with chronic lymphocytic leukemia (14) was used a t a 1:20 dilution for the detection of EBNA 1 ,2 ,3 ,4, and 6. This serum also reacted with EBNA 5, seen as multiple weak bands in the 22-90 kD region. The EBV antibody titers of this serum in immunofluorescence assays were VCA: IgA
Detection of Multiple Epstein-BarrViral Strains in Allogeneic Bone Marrow Transplant Recipients J.W. Gratama, E.T. Lennette, B. Lonnqvist, M.A.P. Oosterveer, G. Klein, 0.Ringden, and I. Ernberg Department of Immunology, Daniel den Hoed Cancer Center, Rotterdam, The Netherlands (J.W.G., M.A.P.O.); Department of Immunohematology and Blood Bank ( J .W.G.), Leiden, The Netherlands; Virolab Inc., Berkeley, California (E.T.L.); Departments of Medicine (B.L.), Clinical Immunology and Transplantation Surgery (O.R.), Huddinge Sjukhus, Huddinge, Sweden; and Department of Tumor Biology, Karolinska Institute, Stockholm, Sweden (G.K.,I.E.) We have previously shown i n 3 allogeneic bonemarrow transplant (BMT) recipients that complete replacement of recipient marrow was associated with the elimination of the pretransplant Epstein-Barr virus (EBV) strain of the recipient. To study the kinetics of EBV elimination and reinfection in more detail, we have performed a longitudinal study of BMT recipients combining serology, virus isolation from mouthwashes and peripheral blood, and EBV strain characterization. Oropharyngeal EBV excretion was found to persist after the cytoreductive therapy prior to BMT, whereas EBV-carrying cells in the blood were detected only after 5 weeks following BMT. During the first month post-BMT, 2 different EBV strains could be isolated from sequential mouthwashes of 3 patients. The initial strains were found to persist up to 7, 21, and 29 days postBMT, whereas the subsequent strains appeared at 21, 42, and 34 days post-BMT, respectively. Thus, the original EBV strain may persist only for a limited time after BMT, and the oropharyngeal epithelium may be reinfected by a new EBV strain from the blood within 3 weeks. With respect to the coexistence of multiple EBV strains, 2 patterns were evident. From the day 62 mouthwash of 1 patient, 1 Type A and 1 Type B strain were isolated. From the day 180 mouthwash of a second patient, a dominant Type A strain was recovered, together with 6 ”variant” strains that differed from each other by only a single EBNA protein (EBNA 1). This pattern may be explained by viral recombinations during replication, which may form the basis for the vast polymorphism of EBV observed in unrelated individuals. o 1992 Wiiey-Liss, Inc.
KEY WORDS: serology, viral strain typing, EBV elimination kinetics, EBV reinfection 0 1992 WILEY-LISS, INC.
INTRODUCTION The combined size variations of Epstein-Barr virus (EBV)nuclear antigens EBNA 1,2,3,4, and 6 in irnrnunoblots of lymphoblastoid cell lines (LCL) have been used to distinguish between different EBV strains [Ernberg et al., 1986; Gratama et al., 1988, 1990a, 1990b;Rowe et al., 1989; Sculley et al., 1987,19901.We have used this “Ebnotyping” approach to distinguish between donor and recipient virus in bone-marrowgrafted patients [Gratama et al., 1988, 1990al. Allogeneic bone-marrow transplantation (BMT) aims to replace the recipient’s diseased hematopoietic tissue with that of the donor. Our studies, combining EBV serology with Ebnotyping, showed that repopulation of a seropositive recipient with bone marrow of a seronegative donor caused long-lasting seronegativity and made her susceptible to later infection with an exogenous EBV strain. In 2 other patients, repopulation with marrow from a seropositive donor was followed by the disappearance of the pretransplant EBV strain and the establishment of persistent infection with the exogenous virus, which could be traced back to the marrow donor in 1patient [Gratama et al., 1988,1990al. Graft failure in a fourth patient was associated with the persistence of recipient lymphocytes and the recipient’s pretransplant EBV strain [Gratama et al., 1990al. Taken together, these studies suggest that the hematopoietic compartment, and not the oropharyngeal epithelium, is the permanent reservoir of the virus. Observations on patients on long-term acyclovir treatment, which blocked oropharyngeal EBV replication but did not decrease the frequency of EBV-carrying lymphocytes in the peripheral blood [Yao et al., 19891,support the contention that the persistence of infected lymphocytes is independent of the oropharyngeal lytic infection. The disappearance of the recipient’s EBV strain and the appearance of the donor’s strain provides an interAccepted for publication April 30,1991. Address reprint requests to J a n W. Gratama, M.D., Department of Immunology, Daniel den Hoed Cancer Center, P.O. Box 5201, 3008 AE Rotterdam, The Netherlands.
40
esting opportunity to study the kinetics of establishment of EBV infection of B lymphocytes and oropharyngeal epithelium, and to study the “traffic” of the virus between these compartments. We report now on a prospective study of the appearance of EBV-carrying cells in the blood and changes in salivary EBV excretion patterns of allogeneic BMT recipients in relation to serology and viral strain typing.
PATIENTS AND METHODS Patients Thirty-two patients were entered into a longitudinal study, They received marrow grafts from a n identical twin (n = l ) , HLA-A,B,C,DR-identical siblings (n = 23), from other related donors (n = 6), or from HLA-matched unrelated donors (n = 2). The median age of the patients was 33 years (range, 2-58 years). The diagnostic indication for BMT was hematological malignancy in 29 patients and nonmalignant disease (severe aplastic anemia or metabolic disorders) in 3. Prophylaxis of acute graft-versus-host disease (GVHD) consisted of (1)T-cell depletion of the marrow grafts in 9 patients [Tollemar et al., 19891or (2) a combination of cyclosporin A and methotrexate in 22 patients [Storb et al., 19861. The recipient of identical-twin donor marrow did not receive any GVHD prophylaxis. Patients with hematological malignancies were prepared for BMT as follows: (1) those receiving T-cell-depleted grafts were treated with cyclophosphamide (60 mg/kg/ day x 2) followed by total-body irradiation (7.5 Gy in 1 session) coupled with total lymph-node irradiation (3 Gy in 2 sessions); (2) those receiving cyclosporin A and methotrexate were treated with cyclophosphamide (60 mg/kg/day x 2) followed by total-body irradiation (10 Gy in 1 session). The 3 patients with nonmalignant disorders were prepared for BMT using a combination of cyclophosphamide (2 g/m2/day x 4) and busulphan (4-5 mglkglday x 4). All patients were nursed in reversed isolation during the peritransplant period. Selective decontamination of their gastrointestinal tracts was attempted as described elsewhere [Ringden et al., 19861. Leukocytepoor erythrocyte and platelet concentrates were administered to maintain the hemoglobin level above 75 g/1 and the platelet count above 30 x 109/l.Thirty patients showed sustained engraftment and 2 patients died before marrow engraftment was documented. Antiviral drugs were given to 22 patients. Eleven patients received foscarvir as prophylaxis for active CMV infection. This drug was given as a continuous i.v. infusion from the start of the cytoreductive therapy until marrow engraftment. The dose was adjusted to reach plasma levels between 100 and 150 pg/l. This concentration inhibits the replication of most CMV strains in vitro. An additional patient received foscarvir therapeutically for proven active CMV infection. Five patients received acyclovir prophylactically and 6 therapeutically (250 mg/mz 2 or 3 times daily, respectively) against active HSV infection.
Gratama et al. Acute GVHD occurred in 23 patients: It was mild (Grade I) in 15 patients and moderate to severe (Grades 11-IV) in 8. Five patients developed limited chronic GVHD and 1extensive chronic GVHD. Fifteen patients died at a median time of 75 days post-BMT (range 10593 days): 9 from BMT-related causes and 6 from relapse of their original disease. Fifty-one patients were studied on a single occasion at a median of 727 days post-BMT (range 52-2562 days). They had received marrow grafts from a n identical twin (n = 11, HLA-A,B,C,DR-identical siblings (n = 46), or from other related donors (n = 4). Their median age was 21 years (range, 2-50 years). The diagnostic indication for BMT was hematological malignancy in 40 patients, severe aplastic anemia in 6, and other in 5. Chronic GVHD had developed in 15 patients: It was limited in 11and extensive in 4 patients. None of these patients received antiviral therapy a t the time of our study.
EBV Serology Cryopreserved serum samples from the patients and their donors were tested for IgM and IgG antibodies against EBV viral capsid antigen (VCA), for IgG and IgA antibodies against early antigens (EA) of the diffuse (D) variety, and for IgG antibodies against EA of the restricted (R) variety using indirect immunofluorescence [Henle and Henle, 1966, 19701. Antibodies against the nuclear antigens EBNA 1, 2A, 2B, and 6 were determined using anticomplement immunofluorescence (ACIF) [Reedman and Klein, 19731. DG-75 cells transfected with the BamHI K fragment of B95-8 EBV DNA served as a source of EBNA 1. DG-75 cells transfected with the BamHI Y fragment of B95-8 EBV and RAT-2B cells transfected with the BamHI Y fragment of Ag-876 EBV were used for the EBNA 2A and 2B antibody assays, respectively. DG-75 cells transfected with the BamHI E fragment of B95-8 EBV served as a source of EBNA 6 . Expression of the relevant antigens was controlled by ACIF and Western blotting, and only transfected lines in which the majority of the cells expressed the antigens were used for the antibody assays. Raji cells expressing EBNA 1, 2A, 3, and 4, and EBV-negative BJA-B cells served as positive and negative controls, respectively. Smears were made by spreading 5 p1 of a suspension containing lo7 cells over 6 x 30 mm coverslips. The smears were allowed to dry rapidly in a n air current and were kept for 2 hours at room temperature prior to fixation in a precooled (- 2 0 C ) mixture of equal volumes of acetone and methanol for 1minute. After fixation, the smears were kept in plastic boxes at -20°C until used. All sera from each patient were tested on the same occasion to exclude the possible influence of interassay variability. EBV Isolation EBV was isolated either from the oropharynx by mouthwashing or by establishing a n EBV-carrying LCL from the peripheral blood. Mouthwashes were obtained by gargling for 1minute with 10 ml RPMI 1640
Multiple EBV Strains in BMT Recipients medium and were cryopreserved after removal of bacteria and debris, and the addition of 10% (v h ) fetal calf serum. For the lymphocyte transformation assays, 12 million cord blood mononuclear cells were incubated with 3-ml aliquots for 3-6 hours, washed, and cultured for 8-12 weeks in 96-well flat-bottomed microtiter plates [Gratama et al., 19881. To obtain spontaneously outgrowing lymphoblastoid cell lines (LCL), peripheral blood mononuclear cells (PBMC) were obtained from heparinized blood samples and cultured in 96-well flatbottomed microtiter plates for 8-12 weeks [Gratama et al., 19881. Between 1and 2 weeks after the initiation of the cultures, 100,000 fresh cord blood lymphocytes were added to each well in order to rescue EBV released from virus-carrying cells [Lewin et al., 19871. From these 2 types of cultures, wells with growing cells were harvested and the resulting LCL were expanded. From the 32 longitudinally studied patients, blood samples were obtained between 21 and 8 days prior to BMT, and at various time points between 30 and 180 days post-BMT. Mouthwashes were obtained between 21 and 8 days prior to BMT, once weekly between days 0 and 28 post-BMT, and concurrent with blood sampling thereafter. PBMC and mouthwashes were obtained from the marrow donors within 2 weeks prior to BMT. From the 51 patients studied on a single occasion, blood and mouthwashes were obtained between 100 and 2246 days post-BMT.
Immunoblotting Aliquots corresponding to 1million LCL dissolved in electrophoresis sample buffer were loaded per lane for electrophoresis in gels containing 7.5% polyacrylamide and sodium dodecyl sulfate [Gratama et al., 19881. The following molecular mass markers were used: myosin (200 kD), E. coli beta-galactosidase (116 kD), rabbit muscle phosphorylase b (97 kD), bovine serum albumin (66 kD), and hen egg-white ovalbumin (43 kD) (BioRad, Richmond, CA). The electroblotting procedure and the immunological detection of the EBNA proteins on the nitrocellulose filters was performed as described previously [Gratama et al., 19881.Serum from a patient with chronic lymphocytic leukemia (14) was used a t a 1:20 dilution for the detection of EBNA 1 ,2 ,3 ,4, and 6. This serum also reacted with EBNA 5, seen as multiple weak bands in the 22-90 kD region. The EBV antibody titers of this serum in immunofluorescence assays were VCA: IgA
