Antibody responses to Epstein-Barr virus-encoded

Journal of Medical Virology 68:370–377 (2002)

Antibody Responses to Epstein-Barr Virus-Encoded Latent Membrane Protein-1 (LMP1) and Expression of LMP1 in Juvenile Hodgkin’s Disease Pauline Meij,1 Marcel B.H.J. Vervoort,1 Elisabeth Bloemena,1 Tabitha E. Schouten,1 Cindy Schwartz,2 Seymour Grufferman,4 Richard F. Ambinder,3 and Jaap M. Middeldorp1* 1

Department of Pathology, VU Medical Center, Amsterdam, The Netherlands Department of Oncology and Pediatrics, Johns Hopkins University, School of Medicine, Baltimore, Maryland 3 Department of Oncology, Johns Hopkins University, School of Medicine, Baltimore, Maryland 4 Department of Family Medicine and Clinical Epidemiology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 2

A large group of juvenile Hodgkin’s disease patients (n ¼ 242, mean age 11.7 years, 75% [n ¼ 181] seropositive) was evaluated for antiEpstein-Barr virus (EBV) antibody responses and the presence of EBV-encoded EBER-RNA and latent membrane protein-1 (LMP1)-protein expression in the tumor. The molecular diversity of anti-EBV antibody responses in Hodgkin’s disease patients with EBV-positive and-negative tumors was studied by enzyme-linked immunosorbent assay (ELISA) and immunoblot. Using purified recombinant LMP1 protein as antigen, the presence of antibodies to LMP1 was related to expression of LMP1 in the tumor cells and specific EBV-serological patterns. Antibodies to LMP1 were detected in 30% of the EBV-seropositive Hodgkin’s disease patients. The presence of antibodies to LMP1 was not associated with a distinct anti-EBV antibody diversity profile (ELISA), but a significantly higher percentage of patients with antibodies to LMP1 had antibodies to ZEBRA and viral capsid antigen (VCA)p18 (Immunoblot). Significantly more patients with an EBV-positive tumor had detectable antibody responses to LMP1, but the presence of antibodies to LMP1 did not reflect the expression of LMP1 protein in the tumor cells. Interestingly, all patients with the strongest antibody responses to LMP1 had EBV-negative tumors, suggesting immunological selection in vivo. J. Med. Virol. 68:370–377, 2002. ß 2002 Wiley-Liss, Inc.

KEY WORDS: EBV; serology; immunological selection; immunity; oncogenesis

ß 2002 WILEY-LISS, INC.

INTRODUCTION Epstein-Barr virus (EBV) is a g-herpesvirus and the etiological agent of infectious mononucleosis. After primary infection, EBV establishes a lifelong persistent infection. More than 90% of the world population is infected with EBV. EBV is also associated with variety of lymphoid and epithelial malignancies, including Burkitt’s lymphoma, nasopharyngeal carcinoma, Bnon-Hodgkin lymphomas in immunocompromised patients, 40–90% of Hodgkin’s disease cases, T/natural killer (NK)-cell lymphomas, and 10% of gastric carcinomas. In the tumor cells of these malignancies, different EBV latent genes are expressed [Rickinson and Kieff, 1996]. In Hodgkin’s disease, a restricted set of latent genes is expressed, encoding the so-called latency type II antigens: Epstein-Barr nuclear antigen 1 (EBNA1), latent membrane proteins-1 and -2 (LMP1 and -2), two abundant small noncoding RNAs (EBER1 and -2), and BamHI-A rightward transcripts (BARTs) of unknown function [Herbst et al., 1991; Pallesen et al., 1991; Brooks et al., 1993; Niedobitek et al., 1997]. Hodgkin’s disease has a typical bimodal age distribution, with an early peak in children and young adults and in those >60 years. Individuals with a history of infectious mononucleosis and individuals with elevated antibody titers Grant sponsor: National Institutes of Health; Grant number: RO1 CA 47473; Grant sponsor: Pediatric Oncology Group; Grant sponsor: Children’s Cancer Group; Grant sponsor: National Cancer Institute; Grant numbers: CA30969, CA13539. *Correspondence to: Dr. Jaap M. Middeldorp, Department of Pathology, VU Medical Center, de Boelelaan 1117, 1081 HV Amsterdam, The Netherlands. E-mail: [email protected] Accepted 26 May 2002 DOI 10.1002/jmv.10213 Published online in Wiley InterScience (www.interscience.wiley.com)

Antibodies to LMP1 in Juvenile Hodgkin’s Disease Patients

to EBNA and viral capsid antigen (VCA) have an increased risk of developing Hodgkin’s disease [Mueller, 1987; Mueller et al., 1989; Hjalgrim et al., 2000]. LMP1, a transmembrane protein with six membranespanning segments, is expressed in abundance in the characteristic Hodgkin/Reed-Sternberg cells of Hodgkin’s disease. LMP1 is the most important transforming EBV protein, with clear oncogenic properties in both lymphoid and nonlymphoid cells [Kieff, 1996]. Most effects of LMP1 are mediated by the activation of NFkB via tumor necrosis factor receptor-associated factors (TRAF1-3), which are activated constitutively through the binding to the CTAR1 and CTAR2 elements in the Cterminal cytoplasmic domain of LMP1 [Devergne et al., 1996; Floettmann et al., 1998]. LMP1 mimics transmembrane signaling of cellular receptors such as CD40 but is considered a constant firing receptor [Uchida et al., 1999]. Furthermore, LMP1 may also modulate the immune response by induction of IL-10 expression in vitro [Nakagomi et al., 1994]. Recently it has been suggested that LMP1 (fragments) may have direct immunomodulating effects on T cells, rendering them anergic to antigenic and mitogenic stimuli [Dukers et al., 2000]. Cytotoxic T-cell responses to LMP1 have been described in healthy donors, although precursor frequencies are low [Khanna et al., 1998, 1999]. Antibody responses have been detected in subgroups of patients with EBV-associated diseases, namely in nasopharyngeal carcinoma, Hodgkin’s disease, infectious mononucleosis, and Burkitt’s lymphoma patients using a variety of techniques [Modrow and Wolf, 1986; Rowe et al., 1988; Chen et al., 1992; Xu et al., 2000]. In a previous study, using purified full-length recombinant LMP1 protein to analyze the antibody response to LMP1, LMP1-antibodies were found in 3/8 EBV-positive adult Hodgkin’s disease patients and in 3/40 nasopharyngeal carcinoma patients [Meij et al., 1999]. Antibodies to LMP1 were not found in healthy EBV carriers, infectious mononucleosis patients or patients with chronic EBV disease, despite the presence of a wide spectrum of antibodies to other EBV proteins in the latter [van Grunsven et al., 1993; Meij et al., 1999]. This study further defines the molecular fine-specificity of anti-EBV antibodies in a group of juvenile Hodgkin’s disease patients and focuses in more detail on the anti-LMP1 antibody response. Because antiLMP1 antibodies in Hodgkin’s disease patients have only been studied in relatively small groups, the EBV serology was examined more extensively analyzed in a large group of juvenile Hodgkin’s disease patients (n ¼ 242). It was determined whether the presence of EBV in the tumor is reflected by a specific EBV antibody pattern and whether antibodies to LMP1 in the circulation reflect the expression of LMP1 in the tumor. MATERIALS AND METHODS Patients and Control Sera Serum samples from juvenile patients with Hodgkin’s disease (n ¼ 242; mean age 11.7 years, range 3.2–

371

16.4 years) were collected by Dr. R.F. Ambinder and Dr. S. Grufferman as a part of a large epidemiological study of Hodgkin’s disease in children (National Institutes of Health [NIH]). The specimens for this study were provided by the member institutions of the Pediatric Oncology Group and the Children’s Cancer Group, cooperative clinical trial groups supported by the National Cancer Institute (NCI) Grant CA30969 and CA13539. From these patients, race, sex, age at diagnosis, income and education of the parents at birth, Hodgkin’s disease type, and EBV status were recorded and will be described in detail elsewhere (R.F. Ambinder and S. Grufferman, manuscript in preparation). The EBV status of the tumor was determined by EBER1, -2 in situ hybridization (EBER-RISH) in 129 tumor samples as described previously [Ambinder and Mann, 1994]. Additional formalin-fixed and paraffin-embedded tumor material was obtained from 26 patients, to investigate the LMP1 expression in situ. Control sera were obtained from healthy EBV-positive laboratory donors (n ¼ 10) and healthy EBV-positive USA blood donors (n ¼ 18) kindly provided by Dr. Horwitz (Abbott Northwestern Hospital, Minneapolis, MN). Immunoblotting Nuclear extracts of early antigen (EA) and VCAinduced HH514.C16 cells [van Grunsven et al., 1993] and purified recombinant LMP1 [Meij et al., 2000] were used as a source of antigen for immunoblot assays, using standardized procedures as described by our group previously [Meij et al., 1999]. Nonspecific binding sites on the nitrocellulose blot strips were saturated with blocking buffer consisting of 5% nonfat milk powder and 5% fetal calf serum (FCS) in phosphate-buffered saline (PBS). Subsequently, monoclonal antibodies (at 1–2 mg/ ml) or human sera (diluted at 1:40–1:100) were added and incubated for 1 hr or overnight at room temperature. After washing with PBSt (0.05% Tween in PBS), specifically bound IgG was detected with horseradish peroxidase (HRP)-conjugated-anti-human IgG (Dako, Glosstrup, Denmark), diluted in blocking buffer. After washing in PBSt followed by washing in PBS, HRP activity was visualized with 0.06% 4-chloro-1-naphthol and 0.03% H2O2 in PBS. The diagnostic characteristics of EBV antibody profiling by immunoblot has been described in detail previously [Middeldorp and Herbrink, 1988; van Grunsven et al., 1994; Meij et al., 1999]. EBV-Specific ELISA For the measurement of EBNA1 IgG- and VCAspecific IgG and IgM antibodies, commercial ELISAs were used based on combinations of defined synthetic peptide epitopes derived from EBNA1 and VCA-p18 (Organon Teknika, Boxtel, The Netherlands). ELISAs were carried out according to the manufacturer’s instructions. For the detection of IgG and IgM antibodies directed against EAD, a two-step ELISA was used [Middeldorp, 1998]. EAD combipeptide 501, derived from the C-terminus of the EAD marker protein BMRF1,

372

Meij et al.

was coated overnight at 1 mg/ml in 0.1 M carbonate buffer (pH 9.6) on microtiter plates. Nonspecific binding sites were blocked subsequently for 1 hr with PBS/3% bovine serum albumin (BSA) at 378C. All further incubations were carried out for 1 hr at 378C, followed by four washes in PBSt. For the detection of EAD-specific IgG antibodies, human sera were diluted (1:100) in dilution buffer (PBS, 0.025% Tween, 1% BSA, 1% Triton X-100). For the analysis of the EAD-specific IgM response, human sera were first treated with GullSORP (Meridian Diagnostics, Cincinnati, OH). Bound IgG or IgM was detected with HRP-labeled second antibody (Dako) in dilution buffer (1:1,000). HRP activity was visualized using 3,30 -5,50 -tetramethyl benzidine (Organon Teknika), and the reaction was stopped by adding 1 M H2SO4. The optical density (OD) was determined at 450 nm, and cutoff values were defined as the mean OD450 value of three EBV-negative sera plus 3
SD. Immunohistochemical Analysis Immunohistochemistry was carried out on paraffinembedded tissue sections, which were deparaffinized with xylene. For the detection of LMP1, the S12 monoclonal antibody was used [Jiwa et al., 1995]. Sections were incubated with normal rabbit serum, followed by incubation with S12 for 1 hr at room temperature. Bound S12 was detected using a biotinylated rabbit antimouse antibody (Dako), followed by streptavidin-biotin HRP (Dako). The peroxidase activity was visualized with diaminobenzidine/H2O2 (Sigma, St. Louis, MO). Statistical Analysis Statistical evaluation of the differences between two groups (EBV-negative tumor/EBV-positive tumor and LMP1 antibody-negative/LMP1 antibody-positive) was carried out using the nonparametric Mann-Whitney Uranking test. All values were based on two-tailed statistical analysis. Differences were considered significant when P < 0.05. All statistical procedures were carried out with SPSS 10.0 software (SPSS, Chicago, IL). RESULTS EBV Serostatus In this study, the first step was to determine whether the Hodgkin’s disease patients were EBV seronegative or seropositive by testing for EBNA1 and VCA-specific antibody responses by ELISA. Detectable antibody responses to EBNA1 and VCA-p18 were regarded as evidence of a previous EBV infection. Since the patients were relatively young, we expected to find a higher percentage of EBV-negatives compared with the general population (