- Email: [email protected]
Routine detection of Epstein–Barr virus specific T-cells in the peripheral blood by flow cytometry
Journal of Immunological Methods 247 (2001) 35–47 www.elsevier.nl / locate / jim
Routine detection of Epstein–Barr virus specific T-cells in the peripheral blood by flow cytometry a b c c, Brian E. Crucian , Raymond P. Stowe , Duane L. Pierson , Clarence F. Sams * a Wyle Laboratories, Cell and Molecular Research Laboratories, Houston, TX, USA University of Texas Medical Branch, Department of Pathology, Galveston, TX, USA c NASA-Johnson Space Center, Life Sciences Research Laboratories, Mail Code SD3, Houston, TX 77058, USA b
Received 10 May 2000; received in revised form 1 November 2000; accepted 3 November 2000
Abstract The ability to detect cytomegalovirus-specific T-cells (CD4 1 ) in the peripheral blood by flow cytometry has been recently described by Picker et al. In this method, cells are incubated with viral antigen and responding (cytokine producing) T-cells are then identified by flow cytometry. To date, this technique has not been reliably used to detect Epstein–Barr virus (EBV)-specific T-cells primarily due to the superantigen / mitogenic properties of the virus which non-specifically activate T-cells. By modifying culture conditions under which the antigens are presented, we have overcome this limitation and developed an assay to detect and quantitate EBV-specific T-cells. The detection of cytokine producing T-cells by flow cytometry requires an extremely strong signal (such as culture in the presence of PMA and ionomycin). Our data indicate that in modified culture conditions (early removal of viral antigen) the non-specific activation of T-cells by EBV is reduced, but antigen presentation will continue uninhibited. Using this method, EBV-specific T-cells may be legitimately detected using flow cytometry. No reduction in the numbers of antigen-specific T-cells was observed by the early removal of target antigen when verified using cytomegalovirus antigen (a virus with no non-specific T-cell activation properties). In EBV-seropositive individuals, the phenotype of the EBV-specific cytokine producing T-cells was evaluated using four-color flow cytometry and found to be CD45 1 , CD3 1 , CD4 1 , CD45RA2 , CD69 1 , CD25 2 . This phenotype indicates the stimulation of circulating previously unactivated memory T-cells. No cytokine production was observed in CD4 1 T-cells from EBV-seronegative individuals, confirming the specificity of this assay. In addition, the use of four color cytometry (CD45, CD3, CD69, IFNg/ IL-2) allows the total quantitative assessment of EBV-specific T-cells while monitoring the interference of EBV non-specific mitogenic activity. This method may have significant utility for the monitoring of the immune response to latent virus infection / reactivation. 2001 Elsevier Science B.V. All rights reserved.
1. Introduction Epstein–Barr virus (EBV) causes infectious mononucleosis (IM) and is also closely linked to a *Corresponding author. Tel.: 11-281-483-7160; fax: 11-281483-2888. E-mail address: [email protected] (C.F. Sams).
variety of human cancers, including B cell lymphoma, Hodgkin’s disease and T-cell lymphoma (Henle et al., 1968; Zur Hausen et al., 1970; Hamilton-Dutoit et al., 1991; Herbst et al., 1991; Pallesen et al., 1991). Primary infection of EBV usually occurs asymptomatically during childhood, and greater than 90% of adults demonstrate previous exposure to EBV (Kieff, 1996). The virus infects
0022-1759 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0022-1759( 00 )00326-4
36
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
both B-cells and epithelial cells, but the primary viral target is the resting B cell (Kieff, 1996). EBV infection results in B-cell activation and polyclonal expansion which is limited by the immune system. Once acquired, the virus persists in the host throughout life and is carried in ‘‘latent’’ form in peripheral blood B-cells (Yao et al., 1985). There is now data suggesting that latent EBV may reactivate in response to psychological or other stress (Glaser et al., 1991, 1999). During acute IM, there is a marked lymphocytosis comprised primarily of T-cells expressing CD45RO, a memory T-cell marker (Miyawaki et al., 1991). A large-scale activation of non-specific T-cells also occurs, which has led to the belief that an EBV associated superantigen may be influencing the immune system during the establishment of infection (Sutkowski et al., 1996). The cytotoxic T lymphocyte (CTL) response, which includes both CD4 1 and CD8 1 T-cells, appears directed against a few epitopes that are mainly derived from EBV nuclear antigens 3, 4 and 6 (Campos-Lima et al., 1997). During latency, virus-specific memory T-cells have a critical role in controlling EBV reactivation and restricting proliferation of EBV-infected B-cells (Rickinson et al., 1984; Kieff, 1996). Recently a method for determining the peripheral frequency of viral specific memory T-cells in previously infected individuals by flow cytometry has been described (Waldrop et al., 1997; Maino and Picker, 1998; Suni et al., 1998). Using this method peripheral blood mononuclear cells (PBMCs) are cultured in the presence of viral antigens and certain costimulatory molecules. The viral antigens are ingested and processed by the antigen presenting cells (APCs) which then present the antigens to memory T-cells. This antigen presentation results in the selective activation of memory CD4 1 T-cells which produce inflammatory cytokines such as interleukin-2 (IL-2) or interferon-g (IFN-g). The detection of T-cell production of IL-2 and IFNg by flow cytometry may then be used as a marker for viral specific memory cells. This method presents many advantages over typical long-term cultures to detect antigen-specific T-cell activation. By performing rapid activation (after 6 h) an accurate assessment of viral specific T-cells may be achieved without the interfering effects of proliferation, cell death or bystander T-cell activation.
This method has been utilized to detect peripheral memory T-cells specific for Cytomegalovirus (CMV; Waldrop et al., 1997; Suni et al., 1998), Human Immunodeficiency Virus (HIV; Moss et al., 1999) and Lymphocytic Choriomeningitis Virus (LCMV; Varga and Welsh, 1998a,b). To date, this method has not been utilized to detect peripheral T-cells specific for EBV. This is primarily due to the non-specific activation of T-cells which is induced by EBV viral proteins. In this report, we describe efforts to modify to the cell culture system in an attempt to reduce the non-specific activation of T-cells by EBV while allowing viral antigen presentation to continue uninhibited. In addition, we describe the use of multiparameter flow cytometry to distinguish the nonspecific activation of T-cells by EBV from legitimate viral-specific memory T-cell activation.
2. Materials and methods
2.1. Blood donors For assay development, whole blood samples were obtained from adult donors into ACD anticoagulant vacutainers. The subjects had been screened for most major infectious diseases and found to be healthy. Informed consent was obtained from all blood donors after the nature of the studies had been explained to them. Following serology testing, all of the randomly selected adult donors were found to be EBV-seropositive. To obtain EBV-seronegative specimens pediatric whole blood samples were obtained from a clinical facility and screened for their serology status. Four confirmed seronegative pediatric specimens were obtained as well as an adult seronegative specimen kindly donated by Dr. Raymond Widen, Tampa General Hospital.
2.2. Mitogens, antigens and viral lysates Mitogens, antigens and viral lystates used to supplement cell culture were as follows: CMV infected cell extract (Advanced Biotechnologies Inc., Columbia, MD-ABI); P3HR1 EBV infected cell extract (ABI, Columbia, MD); vero uninfected control extract (ABI, Columbia, MD); human foreskin fibroblast uninfected control extract (ABI, Columbia, MD); staphylococcus enterotoxin A (Sigma, St.
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
Louis, MO); phorbol myristate acetate (Sigma, St. Louis, MO); ionomycin (Sigma, St. Louis, MO).
2.3. PBMC /purified T-cell isolation For PBMC isolation, 4 ml of whole anticoagulated blood was diluted 1:1 with sterile phosphate buffered saline (PBS) and subjected to centrifugation over a Ficoll-Hypaque density gradient (Pharmacia, Upsala Sweden). The isolated cells were washed three times in sterile PBS and resuspended in RPMI medium 1640 (Gibco BRL, Grand Island, NY) containing 10% fetal bovine serum (GIBCO) and 10 000 Units / ml penicillin and 1000 mg / ml streptomycin (GIBCO). The final concentration of cells was determined using a hemacytometer. A final concentration of approximately 1.0310 6 cells / ml was used in all experiments. Purified T-cells were isolated from PBMCs using commercially available T-cell purification columns (R&D Laboratories Inc., Marina Del Ray, CA) according to the manufacturers instructions. Monocyte depleted PBMC’s were created via CD14-magnetic bead separation (Dynal, Oslo, Norway) according to the manufacturers’ instructions.
2.4. Cell culture conditions for activation For PBMC stimulation with viral or uninfected cell lysates, PBMCs were cultured for 1 h in the presence of 2.5 mg of purified anti-CD28 (PharMingen, San Diego, CA) antibody and the indicated amount of viral extract. After 1 h (to allow antigen uptake by APC’s) cells were washed in sterile HBSS and resuspended in medium containing 3 mM monensin. To verify T-cell viability and reactivity, PBMCs were cultured in either medium containing 10 ng / ml PMA (Sigma, St. Louis, MO) and 1 mg / ml ionomycin (Sigma) or in medium containing 10 mg / ml staphylococcus enterotoxin A (Sigma). In all cases 3 mM monensin (Sigma) was added to the PBMC cultures for the final 4 h to inhibit extracellular transport of cytokines and allow intracellular accumulation.
2.5. Fluorescent antibody staining Cell surface markers were stained first by resuspending the PBMC’s in 200 ml of PBS to which
37
0.5–1.0 mg of labeled mouse monoclonal antibodies to human CD69, CD4, CD25, CD3 or CD45 were added as indicated by the experiment and incubated at R.T. for 10 min. Antibody combinations and fluorochrome conjugates / emission wavelength are indicated in Table 1. The PBMC’s were then washed in PBS and fixed in 200 ml of 4.0% paraformaldehyde in PBS for 10 min, then washed again in PBS. To detect intracellular production of IFNg or IL-2 (following surface marker staining), the fixed PBMC’s were resuspended in 200 ml of permeabilization buffer, consisting of 5.0% non-fat dry milk and 0.5% saponin in PBS to which 0.5 mg of labeled mouse antibody to either IFNg or IL-2 (or both) was added. The cells were incubated at room temperature for 25 min and then washed in PBS containing saponin. The cells were then resuspended in 1.0% paraformaldehyde for analysis. The FITC conjugated IFNg and IL-2 antibodies were obtained from Becton Dickinson (Mountain View, CA) and all other antibodies were obtained from Beckman-Coulter (Miami, FL). The mean fluorescence intensity of the four color combinations were verified against single staining to ensure a lack of interference. In addition, the antibodies were verified in single, double and Table 1 Antibody panels used for flow cytometric analysis of EBV specific T-cells in peripheral blood specimens a Antibody panels for flow cytometry
Routine analysis Assay development
a
FL1
FL2
FL3
FL4
IFNg1IL-2 IFNg1IL-2 IFNg1IL-2 IFNg1IL-2
CD69 CD4 CD45RA CD25
CD3 CD3 CD3 CD3
CD45 CD45 CD45 CD45
Routine analysis indicates the antibody configuration suggested for routine evaluation of total EBV specific T-cells while monitoring non-specific activation. Assay development indicates the markers used to determine the phenotype of the EBV responding T-cells during initial experiments. FL1: fluorescence channel one (535 nm emission); FL2: fluorescence channel two (575 nm emission); FL3: fluorescence channel three (633 nm emission); FL4: fluorescence channel four (680 nm emission). The fluorochromes to which the antibodies were conjugated are as follows: FL1: FITC; FL2: PE; FL3: PE-Texas Red; FL4: PECyanin 5. Note that for FL1 detection of cytokines, a combination of antibodies to both IFNg and IL-2 (FITC conjugated) were used to detect production of either, or both cytokines in a single channel. FITC conjugated IFNg and IL-2 antibodies were obtained from Becton Dickinson (Mountain View, CA), all other antibodies were obtained from Beckman-Coulter (Miami, FL).
38
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
triple combinations to verify no artifactual interactions in the four color combination.
2.6. Flow cytometry analysis For data acquisition a Coulter XL flow cytometer was configured for four color analysis. Fluorescence emission channels (appropriate filter wavelengths) for the various labeled antibodies are indicated in Table 1. Instrument performance was verified by calibration with FlowSet fluorescent microspheres (Beckman Coulter). Appropriate electronic compensation for spectral overlap of the light emitted by the different fluorochromes was determined both by verification with a four-color compensation ‘kit’ (Beckman-Coulter) and by experimental internal verification with single-channel positive cell populations. The gating strategy devised for EBV T cell analysis is as follows: non-cellular and artifactual debris were excluded by gating CD45 1 cells from a CD45 vs. side scatter dot plot. From this population, T-cells were resolved by plotting CD3 vs. side scatter. At this point a clear resolution was visible between CD3 1 T-cells and non-CD3 1 cells. T-cells were then gated for analysis, and T-cell production of cytokines (IFNg1IL-2) were plotted vs. a secondary molecule of interest based on the experiment (either CD69, CD45RA, CD4 or CD25 — Table 1). The unique response of only the T-cell population was verified by plotting (from a CD45 1 lymphocyte gate) CD3 vs. cytokine production. Thresholds for positive were set using isotype matched control antibodies or internal cell populations negative for the antigens of interest. To quantitate viral specific T-cells (typically a rare-event population) we scanned a minimum of 100 000 T lymphocytes per analysis to ensure accurate percentages.
2.7. ELISA analysis The EBV serology status of the individual donors was assessed by serum or plasma ELISA by using commercial ELISA kits (Zeus Scientific Inc, Raritin, NJ) according to the manufacturers instructions. Circulating antibodies to EBV–VCA or EBV–EBNA were assayed and the serologic status was calculated as indicated by the manufacturer.
3. Results
3.1. Confirmation of flow cytometry assay for detection CMV-specific T-cells The method for detection of antigen-specific Tcells by flow cytometry has recently been described in a comprehensive fashion by Picker et al. (Waldrop et al., 1997; Suni et al., 1998). Prior to working with EBV viral antigens, for which a general method has not been described, we confirmed the basic method of Picker et. al. by utilizing CMV antigens. The CMV virus does not have superantigenic or mitogenic properties. We found that CMV specific T-cells were readily detectable in the peripheral blood of healthy CMV-seropositive donors using the reported method, and that peripheral blood cells from CMV-seronegative donors demonstrated no response. The responding T-cells expressed the early activation marker CD69, and were identified by their production of IFNg in response PBMC culture in the presence of CMV antigens (Fig. 1). In seropositve donors, the depletion of monocytes significantly abrogated the T-cell response to the viral antigens, indicating the APC dependent nature of the response (data not shown). The only deviation from the method described by Picker et al. was our use of a whole CMV infected cell viral lysate, suspended in a glycine buffer, which was obtained through commercial suppliers.
3.2. PBMC’ s cultured with EBV viral lysate or live virus indicate an EBV superantigen or mitogen property The literature has described the superantigen properties of the EBV virus, and the ability of EBV to activate T-cells in a non-specific fashion. Our initial experiments substituted either an infected cell viral lysate containing EBV proteins or purified live EBV for the CMV lysate in the method utilized above to observe the effects on T-cell activation. The expression of CD69 (as well as the production of IL-2 and IFNg) were assessed. We found that both the EBV viral lysate and the purified live EBV behaved similarly in that both agents did activate peripheral T-cells in a non-specific fashion (Fig. 2). In addition, the non-specific T-cell activation by both
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
39
Fig. 1. Detection of CMV specific T-cells in the peripheral blood of a CMV sero-positive healthy adult blood donor by three color flow cytometry. T-cells (CD3 1 ) have been gated and are plotted in these representative dot-plots. Interferon-g serves as an indicator for antigen specific T-cell activation, and is plotted with the early activation marker CD69. This experiment served to verify our ability to detect viral specific T-cells using the method recently described by Picker et al.
Fig. 2. T cell activation by EBV is nonspecific. Comparison of PBMC vs. purified T cell culture. Representative subject data demonstrating incubation of PBMC’s with live EBV virus or an EBV viral lysate (EBV antigens) induces broad-scale non-specific T-cell activation. 1.0310(6) PBMCs were cultured in the presence of 20 ml of control (uninfected cell) lysate or EBV viral lysate, 5 ml of highly purified live EBV virus, or 10 ng / ml PMA12 mg / ml ionomycin for 6 h. The cells were then stained for CD69 and CD3 1 T-cells were gated for analysis by two-color flow cytometry. Representative results obtained from one healthy adult donor. Note that the similar activation seen in purified T-cells (as compared to PBMC) indicates this activation is not APC dependent and is not the result of memory T-cells responding to specific viral antigen.
40
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
live EBV and the EBV lysate occurred when purified T-cells were substituted for PBMC’s. It was concluded that the EBV proteins may act on T-cells directly. Considering that a viral superantigen is supposed to activate (in the presence of MHC class II) only a particular Vb subset of T-cells, the broadscale activation of T-cells in the absence of APC’s seemed to indicate that the EBV proteins may have mitogenic properties as well as superantigen properties. It was concluded that the superantigenic / mitogenic properties of EBV, and the ability of EBV to activate T-cells in a non-specific fashion could prove to be a major impediment to the development of a routine, generalized flow cytometric assay for the detection of EBV specific T-cells.
3.3. Detection of EBV viral specific T-cells reactive to an EBV infected cell lysate It remained highly desirable to detect EBV-specific T-cells in the peripheral blood using the infected cell lysate. Utilization of the lysate would ensure that T-cells specific for all EBV proteins would be detected. In the method described by Picker et al. PBMC’s are co-cultured with viral antigens for 6 h to generate viral specific responses. Monensin is added to these cultures to allow cytokine accumula-
tion, but only after 1 h has passed to allow for antigen uptake and processing. We investigated the effects of removing the antigen from the culture medium after this 1-h antigen uptake phase. In theory, removal of antigen after 1 h should have no effect on the detection of antigen specific T-cells, since after 1 h the antigen has been internalized and is being processed by the APC’s. It was hypothesized that early removal of the antigen would abrogate the effects of an antigen having mitogenic properties that cause non-specific T-cell activation. An experiment was performed assessing CMV-specific T-cells (CMV antigens having no mitogenic or superantigen properties). This assay was performed incubating PBMC’s with CMV antigen1anti-CD28 for 6 h (monensin for final 5 h) or for 1 h with the antigen then removed (followed by 5 h monensin). Viral specific T-cells were then detected by cytokine flow cytometry. In this manner, CMV served as a control to validate that the early removal of viral antigens would not effect viral specific T-cell detection. Cytokine production was used to identify responding T-cells. It was found that antigen may indeed be removed after the initial 1 h incubation, prior to the 5 h monensin phase without loss sensitivity to viral specific T-cells (Fig. 3). This method was then applied to the detection of
Fig. 3. Detection of CMV specific T cells. Antigen exposure 1 hr vs. 6 hr. Early removal of viral antigen (after APC uptake) does not inhibit antigen presentation to T-cells. (A) PBMCs were incubated with CMV viral antigen for 1 h, followed by 5 h cytokine accumulation. The cells were then stained for CD3 in conjunction with intracellular IFNg1IL-2 for two-color analysis. Viral specific T-cells are identified as producing cytokine in response to incubation with viral antigen. A comparable result is seen in (B) where PBMC’s were incubated with CMV viral antigen and monensin for 6 h (representative of three experiments).
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
EBV specific T-cells. Monocyte depleted cultures were performed simultaneously to control for the APC dependent nature of the viral-specific T-cell response. The data indicated that EBV specific Tcells could be detected by this method with a reduction in non-specific T-cell activation (Fig. 4). The viral specific T-cell response was largely abrogated in the monocyte depleted cultures indicating the APC dependence of the response. In contrast, the responses to PMA1ionomycin were as great or greater in the monocyte-depleted cultures, confirming that the process of T-cell purification did not diminish the viability or responsiveness of the T-cell population (Fig. 4). These data may be contrasted with the data shown previously demonstrating that PBMC incubation with EBV proteins for a full 6 h resulted in monocyte independent non-specific T-cell activation (Fig. 2).
41
3.4. Routine detection and phenotype of viral specific T-cells Previous reports detecting viral specific T-cells monitored the production of IFNg, IL-2 or TNFa. Our goal was the development of an assay suited for a rapid quantitative assessment of all responding viral specific T-cells. To this end, cells producing any of these inflammatory cytokines, or any combination of them may be thought of as responding cells. We investigated the use of a pool of IFNg and IL-2 (both FITC labeled) in our cytometry assay. This allowed the detection of cells stimulated to produce either cytokine, or both cytokines simultaneously, in a single tube. An added benefit is the increased fluorescence levels of responding cells producing both cytokines. The use of a control lysate served as a monitor of constitutively activated T-
Fig. 4. Antigen specific T cell cytokine production in response to EBV is APC dependent. T cell production of IFNg1IL-2 in response to EBV antigens is antigen presenting cell dependent under modified culture conditions. Restriction of antigen exposure to 1 h allows the detection of legitimate EBV and CMV specific T-cells (although for EBV a superantigen effect cannot be ruled out by these results). The higher levels of cytokine production for the purified T-cells in response to PMA1ionomycin confirms their viability and reactivity. Representative two-color analysis results obtained from one healthy adult donor. The EBV and PMA / ionomycin results have been scaled as indicated to allow a graphical relative comparison of all data.
42 B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47 Fig. 5. Representative dot plots demonstrating the presence of EBV specific T-cells for an EBV sero-positive donor and the absence of EBV specific T-cells for an EBV sero-negative donor. Representative three-color analysis. Culture in the presence of EBV antigen (and mitogens / controls) was performed as described. To distinguish T-cell from non-T-cell cytokine production, lymphocytes were gated by CD45 expression and then T-cells (CD3 1 ‘Y axis’) were plotted vs. the cytokine combination of IFNg1IL-2 (‘X axis’). T-cell viability and ability to produce cytokines was assessed by responsiveness to PMA1ionomycin (monocyte independent) and SEA (monocyte dependent). Reactivity to a control lysate verified the absence of non-specific activation.
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
43
Fig. 6. Detection of EBV specific T-cells correlates with EBV serology status. The level of EBV specific T-cells was assessed by flow cytometry as described in three groups: a healthy adult donor group, a pediatric donor group obtained from a local hospital and an adult control group obtained from the same hospital. T-cell subsets producing IFNg1IL-2 in response to EBV antigen were observed in the EBV serology positive populations only. The adult hospital donor group served to verify that collection and any delays in transport did not effect T-cell viability or reactivity.
cells, which were essentially absent in most healthy donors (usually ,10 per 100 000 T-cells). Representative scatter plots demonstrating differences in EBV T-cell reactivity between an adult (EBV seropositive) clinical specimen and a pediatric (EBV sero-negative) clinical specimen which were collected and analyzed in parallel are presented in Fig. 5. In these scatter plots detection of cytokine production by PMA1ionomycin stimulated cultures served to confirm T-cell viability and ability to produce cytokines. Stimulation with the superantigen SEA was utilized to verify superantigen reactivity and monocyte presence, as well as T-cell–monocyte interactions. No cytokine reactivity is visible in response to EBV for the EBV-seronegative donor, whereas a well-defined cytokine reactivity is evident for the EBV-seropositive donor. In a separate experiment, we set up our cytometer for four-color analysis (IFNg1IL-2 / surface marker / CD3 / CD45) to assess the phenotype of the responding T-cells. The responding EBV-specific T-cells were found to primarily have a phenotype of IFNg/ IL-2 1 , CD45RA2 , CD69 1 , CD25 2 , CD3 1 , CD45 1 (data not shown). It was also found that if cultures were allowed to
progress to 24 h activation in the presence of antigen the viral reactive T-cells did express CD25. This finding and cellular phenotype is consistent with that of circulating previously unactivated memory Tcells.
3.5. Detection of EBV-specific T-cells corresponds with EBV serology status For routine clinical detection of total responding viral-specific T-cells, it was decided to utilize three color cytometry (IFNg1IL-2 / CD3 / CD45). This antibody combination would allow the detection of positively identified T-cells (CD3 1 ) responding to viral antigens. For rare event cytometry a clean analysis region is essential. We found the addition of CD45 allowed most cell debris to be gated out, resulting in a reasonably clean analysis even when collecting greater than 100 000 T-cell events. Initially, 14 healthy donors were screened for the presence of EBV specific T-cells using this assay, and the results were correlated with EBV antibody serology results. All the adult donors were found to have previous EBV exposure as determined by serology,
44
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
and also to have well-defined populations of EBVspecific T-cells as determined by flow cytometry (Fig. 6). To confirm the specificity of the assay it was necessary to obtain peripheral blood specimens from donors without previous EBV exposure (negative by EBV serology). Since greater than 95% of the adult population has been previously infected with EBV, obtaining these samples from a pool of healthy adult donors was not feasible. It was necessary to use clinical pediatric specimens, where the exposure rate is much lower. Pediatric blood samples were obtained and screened for EBV serology status. A lower infection rate was confirmed, and four EBV seronegative samples were obtained. To control for the unpredictable nature of such specimens, activation with SEA (monocyte dependent superantigen) and PMA1ionomycin (non-specific T-cell activation) were performed to verify the viability and reactivity of the peripheral T-cells, and their ability to be stimulated to produce cytokines. Pediatric clinical samples in which T-cell activation could not be demonstrated with SEA or PMA1ionomycin were rejected. In addition, serology-positive adult clinical specimens were always obtained in parallel with pediatric specimens to verify specimen transport did not affect the sample viability. The use of control antigens served to verify that no non-specific or constitutive T-cell activation was occurring. In addition to the four EBV sero-negative samples obtained, one EBV sero-negative adult sample from a known uninfected adult was obtained (kindly donated by Dr. Raymond Widen). The data revealed that in almost all the adult or pediatric serology-positive clinical samples, EBV-specific T-cells were detected by this method. In contrast, the four pediatric samples and one adult sample demonstrated to be EBV seronegative had no detectable EBV reactive T-cells by this method (Fig. 6). These data appear to confirm that the cells identified as EBV-specific T-cells by flow cytometry are indeed viral reactive T-cells, and not the result of a non-specific activation of T-cells by EBV superantigenic or mitogenic proteins.
3.6. Four color flow cytometry allows routine discrimination of EBV non-specific vs. antigenspecific T-cell activation Since human immunologic responses vary con-
siderably, T-cells producing cytokines in response to non-specific T-cell activation by EBV are problematic for this assay. It was desirable to develop a system to monitor non-specific T-cell activation in parallel with legitimate viral specific T-cell detection. An experiment was performed culturing PBMC’s with increasing amounts of EBV lysate (1 h incubation). T-cell activation was then assessed by cell surface CD69 and intracellular cytokine production simultaneously using four color flow cytometry (IFNg1IL-2 / CD69 / CD3 / CD45). Since the expression of the activation marker CD69 is easily inducible by most stimuli, CD69 expression was used to monitor non-specific activation in parallel with cytokine assessment for detecting viral specific T-cells. No non-specific activation was detected in PBMC cultures at minimal concentrations of EBV lysate. As EBV lysate concentrations were increased a dramatic onset of non-specific activation (CD69 expression) was observed (Fig. 7A). It was determined that 10–20 ml of EBV lysate per ml of culture was the maximal concentration of antigen that would completely avoid non-specific activation of T-cells. It is likely that the activation of T-cells progresses in sequence: first rapidly to the expression of surface CD69, and then later to the production of intracellular cytokines. When T-cells producing intracellular cytokines were assessed (simultaneously with CD69), it was found that a small population of responding T-cells could be detected producing cytokine following culture with concentrations of antigen lower than those required for broad-scale CD69 induction (Fig. 7A). A small but well-defined population of cytokine producing T-cells was visible following culture with 5 ml of lysate, and even more were detected following culture with 10 ml of lysate. Since these cells were detected as responding in the complete absence of large-scale non-specific T-cell activation as monitored by the induction of CD69, we conclude that these cells represent EBV specific T-cells. These cells were largely absent following culture with 1 ml of EBV lysate. At concentrations of lysate above 20 ml, even more T-cells were observed producing cytokines, however since large-scale expression of surface CD69 was evident, the interfering effect of non-specific activation was likely. A experiment of this type was also performed on a monocyte-depleted PBMC culture. The purpose of the both the purified T-cell experiment (Fig. 4) and
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
45
Fig. 7. Detection of viral specific T cells in conjunction with nonspecific T cell activation: PBMC vs. monocyte-depleted culture. Graphs demonstrating the broad scale induction of CD69 by EBV (d right Y-axis) and simultaneous detection of viral specific T-cells by cytokine expression (♦ left Y-axis) for increasing amounts of EBV antigen (X-axis). Representative data from an EBV sero-positive donor. Four-color analysis (IFNg1IL-2 / CD69 / CD3 / CD45) was performed allowing resolution of cells (CD45 1 ) vs. debris, followed by gating of T-cells (CD3 1 ) expressing either CD69 or cytokine. Viral specific T-cells are detected in the PBMC culture (A) mediated by the APC function of monocytes, and absent in the monocyte depleted culture (B). Although this experiment does not rule out a superantigen effect, the experiments with EBV sero-negative samples have indicated that the superantigen effect is not present in samples stimulated with 20 ml of antigen or less. The optimal amounts of lysate to avoid complete non-specific activation appear to be between 10 and 20 ml of EBV lysate per 1.0 ml of culture medium.
the monocyte depleted experiment (Fig. 7B) was to demonstrate that the cytokine production described (in response to low concentrations of EBV) was monocyte /APC dependent. In the monocyte depleted PBMC culture the ‘shoulder’ of cytokine producing T-cells that was observed in the PBMC culture at low concentrations of EBV lysate was absent. When the concentrations of EBV were increased in the monocyte-depleted culture the expression of CD69 and cytokines roughly tracked together, indicating that the initial cytokine producing T-cell population in the PBMC culture was indeed dependent on the presence of antigen-presenting cells (Fig. 7B). While it is possible that low numbers of CD14 2 APCs (i.e. dendritic cells) remain in the monocyte depleted culture, the loss of cytokine producing T-cells preceding large scale CD69 induction (and the purified T-cell experiments) argue for APC dependent specificity. These results, combined with the data obtained from the serology-negative donor population indicate that legitimate EBV specific T-cells are being detected with this assay.
4. Discussion Epstein–Barr virus infection, although primarily a subclinical juvenile disease, can progress to severe infections or malignancy in certain types of immunocompromised individuals. The documented cycling of the virus through latency to lytic replication, and the ability of the immune system to control lytic infection and restore latency has been recently welldocumented. Current opinion holds that the T-cell Th1 / Th2 bias of the peripheral immune system also plays a significant role in the maintenance of viral latency and the prevention of reactivation. T-cell mediated immune responses, particularly CD8 1 cytotoxic responses, and B-cell-mediated humoral responses under the control of CD4 1 helper T-cells thus coordinate their efforts to control persistent efforts of EBV to reactivate and establish lytic replication. The development of a rapid flow cytometric assay to enumerate peripheral viral specific T-cells has made routine assessment a reality. The clinical value
46
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
of monitoring the peripheral levels of viral specific T-cells has recently been established in the medical literature. The application of these assays to the determination of peripheral EBV specific memory T-cells has been hampered by the mitogenic or superantigenic properties of the EBV viral proteins. Culture of PBMCs in the presence of EBV antigens results in non-specific activation of T-cells. This has primarily been determined by the expression of CD69. The ability to culture viral antigens with a patients peripheral mononuclear cells is critical to the assay, as it allows monocytes to present the viral antigens to viral-specific T-cells (which then activate) allowing their detection by cytokine flow cytometry. The larger-scale non-specific activation of T-cells by EBV essentially creates a problem by masking the lesser population of legitimate viral specific T-cells. In this paper our efforts to overcome this difficulty are documented. We have demonstrated that PBMC culture in the presence of viral antigens for the entire 6 h activation period is not required, since after 1 h the monocyte population has ingested sufficient viral antigen to present the antigens to specific T-cell populations. Removing the viral antigens after 1 h of culture and then allowing APC-mediated T-cell activation and cytokine accumulation to progress normally results in equal detection levels of viral specific T-cells as compared to 6-h cultures (documented for a non-superantigen / mitogenic virus). We have found that the removal of EBV antigen after 1 h largely reduces the non-specific activation of T-cells, while antigen-presentation (now mediated by antigen presenting cells, i.e. monocytes) continues uninhibited. The expression of CD69 is a relatively easy to trigger marker for T-cell activation. This is partially due to the fact that CD69 protein is present in the cellular cytoplasm and does not require new gene expression to produce cell surface protein. For this reason, CD69 expression is widely used in the medical literature as a marker for cellular activation. The assessment of activation by assessing cytokine production by flow cytometry, however is a harderto-stimulate marker of activation. Commonly used T-cell mitogens, such as PHA, do not result in rapid cytokine accumulation to levels detectable by flow cytometry. In contrast, more powerful cellular ac-
tivators such as PMA1ionomycin result in dramatic well-defined populations of cytokine (IL-2 or IFNg) producing T-cells as determined by flow cytometry. It is our belief that at low levels of exposure to EBV proteins, non-specific T-cell activation largely does not progress to cytokine production. Thus, cytokine production may be used as a marker for the detection of EBV specific T-cells, provided care is used to ensure the concentration of EBV antigen remains below the threshold for non-specific activation (and above the threshold required for antigen presentation). By using four-color flow cytometry (IFNg/ IL2, CD69, CD3, CD45) we found it was possible to: (1) positively identify T-cells for analysis; (2) provide a debris free analysis region (essential for rareevent cytometry); (3) identify viral specific T-cells by cytokine production; (4) monitor non-specific activation via expression of CD69. By performing analysis using several concentrations of EBV antigens the maximal concentration of antigen may be selected for detection that does not result in broad scale T-cell activation as monitored by CD69 (Fig. 7). While this technique can largely control for mitogenic effects induced by EBV, it cannot control for a superantigen effect. To prove that positive events are not the result of a superantigen effect, we have relied on the negative results obtained using this method for EBV seronegative clinical samples (Fig. 6). The positive results obtained from EBV seropositive samples and the negative results obtained from seronegative samples seem to confirm the detection of EBV specific T-cells. Given the ease and utility of the method described by Picker et al., we did not feel that an inability to detect EBV specific T-cells would be a problem. Clearly, in normal circumstances the coincubation of antigen, costimulatory molecules and PBMC’s will result in antigen-specific T-cell detection. The primary problem for EBV has continued to be the ‘contamination’ of the positive events with T-cells stimulated in a non-specific fashion by the mitogenic or superantigenic properties of the virus. Here, we believe a method has been described which allows the detection of EBV specific T-cells, while monitoring the advent of non-specific T-cell activation. Future studies with samples obtained from patients suffering from active infection, post-active infection (recovery), as well as adults suffering latent EBV
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
reactivation as a result of stress (or other factors) will hopefully demonstrate the clinical utility of this method as an assessment of the immune response to latent viral reactivation.
Acknowledgements The authors wish to thank Dr. Ray Widen, Tampa General Hospital for providing EBV seronegative whole blood samples and for critical review of this manuscript. We also wish to thank the St. John Hospital, Clinical Laboratory, Houston, Texas, for providing pediatric whole blood specimens.
References Campos-Lima, P.O., Levitsky, V., Imreh, M.P., Gavioli, R., Masucci, M.G., 1997. J. Exp. Med. 186, 83–89. Glaser, R., Friedman, S.B., Smyth, J., Ader, R., Bijur, P., Brunell, P., Cohen, N., Krilov, L.R., Lifrak, S.T., Stone, A., Toffler, P., 1999. The differential impact of training stress and final examination stress on herpesvirus latency at the United States Military Academy at West Point. Brain Behav. Immun. 13, 240–251. Glaser, R., Pearson, G.R., Jones, J.F., Hillhouse, J., Kennedy, S., Mao, H.Y., Kiecolt-Glaser, J.K., 1991. Stress-related activation of Epstein–Barr virus. Brain Behav. Immun. 5, 219–232. Hamilton-Dutoit, S.J., Pallesen, G., Franzmann, M.B., Karkov, J., Black, F., Skinhoj, P., Pedersen, C., 1991. AIDS-related lymphoma. Histopathology, immunophenotype, and association with Epstein–Barr virus as demonstrated by in situ nucleic acid hybridization. Am. J. Pathol. 138, 149–163. Henle, G., Henle, W., Diehl, V., 1968. Relation of Burkitt’s tumor-associated herpes-type virus to infectious mononucleosis. Proc. Natl. Acad. Sci. USA 59, 94–101. Herbst, H., Dallenbach, F., Hummel, M., Niedobitek, G., Pileri, S., Muller-Lantzsch, N., Stein, H., 1991. Epstein–Barr latent membrane protein expression in Hodgkin and Reed-Sternberg cells. Proc. Natl. Acad. Sci. USA 88, 4766–4770.
47
Kieff, E., 1996. Epstein-Barr virus and its replication. In: Fields, B.N., Knipe, D.M., Howley, P.M. (Eds.), Virology. 3rd ed., Vol. 2, Raven Press, New York, N.Y., pp. 2343–2396. Maino, V.C., Picker, L.J., 1998. Identification of funtional subsets by flow cytometry: intracellular detection of cytokine expression. Cytometry 34, 207–215. Miyawaki, T., Kasahara, Y., Kanegane, H. et al., 1991. Clin. Exp. Immunol. 83, 447–451. Moss, R.B., Wallace, M.R., Giermakowska, W.K., Webb, E., Savary, J., Chamberlin-Brandt, C., Theofan, G., Musil, R., Richieri, S.P., Jensen, F.C., Carlo, D.J., 1999. Phenotypic analysis of human immunodeficiency virus (HIV) type 1 cell-mediated immune responses after treatment with an HIV-1 immunogen. J. Infect. Dis. 180, 641–648. Pallesen, G., Hamilton-Dutoit, S.J., Rowe, M., Young, L.S., 1991. Expression of Epstein–Barr virus latent gene products in tumor cells of Hodgkin’s disease. Lancet 337, 320–322. Rickinson, A.B., Rowe, M., Hart, I.J., Yao, Q.Y., Henderson, L.E., Rabin, H., Epstein, M.A., 1984. Cell. Immunol. 87, 646–658. Suni, M.A., Picker, L.J., Maino, V.C., 1998. Detection of antigenspecific T-cell cytokine expression in whole blood by flow cytometry. J. Immunol. Methods 212, 89–98. Sutkowski, N., Palkama, T., Ciurli, C., Sekaly, R.P., ThorleyLawson, D.A., Huber, B.T., 1996. An Epstein–Barr virusassociated superantigen. J. Exp. Med. 184, 971–980. Varga, S.M., Welsh, R.M., 1998a. Detection of a high frequency of virus-specific CD4 1 T-cells during acute infection with lymphocytic choriomeningitis virus. J. Immunol. 161, 3215– 3218. Varga, S.M., Welsh, R.M., 1998b. Stability of virus-specific CD4 1 T-cell frequencies from acute infection into long term memory. J. Immunol. 161, 367–374. Waldrop, S.L., Pitcher, C.J., Peterson, D.M., Maino, V.C., Picker, L.J., 1997. Determination of antigen-specific memory / effector CD4 1 T-cell frequencies by flow cytometry: evidence for a novel, antigen-specific homeostatic mechanism in HIV-associated immunodeficiency. J. Clin. Invest. 99, 1739–1750. Yao, Q.Y., Rickinson, A.B., Epstein, M.A., 1985. Int. J. Cancer 35, 35–42. Zur Hausen, H., Schulte-Holthausen, H., Klein, G., Henle, W., Henle, G., Clifford, P., Santesson, L., 1970. EBV DNA in biopsies of Burkitt tumours and anaplastic carcinomas of the nasopharynx. Nature 228, 1956–1958.
Routine detection of Epstein–Barr virus specific T-cells in the peripheral blood by flow cytometry a b c c, Brian E. Crucian , Raymond P. Stowe , Duane L. Pierson , Clarence F. Sams * a Wyle Laboratories, Cell and Molecular Research Laboratories, Houston, TX, USA University of Texas Medical Branch, Department of Pathology, Galveston, TX, USA c NASA-Johnson Space Center, Life Sciences Research Laboratories, Mail Code SD3, Houston, TX 77058, USA b
Received 10 May 2000; received in revised form 1 November 2000; accepted 3 November 2000
Abstract The ability to detect cytomegalovirus-specific T-cells (CD4 1 ) in the peripheral blood by flow cytometry has been recently described by Picker et al. In this method, cells are incubated with viral antigen and responding (cytokine producing) T-cells are then identified by flow cytometry. To date, this technique has not been reliably used to detect Epstein–Barr virus (EBV)-specific T-cells primarily due to the superantigen / mitogenic properties of the virus which non-specifically activate T-cells. By modifying culture conditions under which the antigens are presented, we have overcome this limitation and developed an assay to detect and quantitate EBV-specific T-cells. The detection of cytokine producing T-cells by flow cytometry requires an extremely strong signal (such as culture in the presence of PMA and ionomycin). Our data indicate that in modified culture conditions (early removal of viral antigen) the non-specific activation of T-cells by EBV is reduced, but antigen presentation will continue uninhibited. Using this method, EBV-specific T-cells may be legitimately detected using flow cytometry. No reduction in the numbers of antigen-specific T-cells was observed by the early removal of target antigen when verified using cytomegalovirus antigen (a virus with no non-specific T-cell activation properties). In EBV-seropositive individuals, the phenotype of the EBV-specific cytokine producing T-cells was evaluated using four-color flow cytometry and found to be CD45 1 , CD3 1 , CD4 1 , CD45RA2 , CD69 1 , CD25 2 . This phenotype indicates the stimulation of circulating previously unactivated memory T-cells. No cytokine production was observed in CD4 1 T-cells from EBV-seronegative individuals, confirming the specificity of this assay. In addition, the use of four color cytometry (CD45, CD3, CD69, IFNg/ IL-2) allows the total quantitative assessment of EBV-specific T-cells while monitoring the interference of EBV non-specific mitogenic activity. This method may have significant utility for the monitoring of the immune response to latent virus infection / reactivation. 2001 Elsevier Science B.V. All rights reserved.
1. Introduction Epstein–Barr virus (EBV) causes infectious mononucleosis (IM) and is also closely linked to a *Corresponding author. Tel.: 11-281-483-7160; fax: 11-281483-2888. E-mail address: [email protected] (C.F. Sams).
variety of human cancers, including B cell lymphoma, Hodgkin’s disease and T-cell lymphoma (Henle et al., 1968; Zur Hausen et al., 1970; Hamilton-Dutoit et al., 1991; Herbst et al., 1991; Pallesen et al., 1991). Primary infection of EBV usually occurs asymptomatically during childhood, and greater than 90% of adults demonstrate previous exposure to EBV (Kieff, 1996). The virus infects
0022-1759 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0022-1759( 00 )00326-4
36
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
both B-cells and epithelial cells, but the primary viral target is the resting B cell (Kieff, 1996). EBV infection results in B-cell activation and polyclonal expansion which is limited by the immune system. Once acquired, the virus persists in the host throughout life and is carried in ‘‘latent’’ form in peripheral blood B-cells (Yao et al., 1985). There is now data suggesting that latent EBV may reactivate in response to psychological or other stress (Glaser et al., 1991, 1999). During acute IM, there is a marked lymphocytosis comprised primarily of T-cells expressing CD45RO, a memory T-cell marker (Miyawaki et al., 1991). A large-scale activation of non-specific T-cells also occurs, which has led to the belief that an EBV associated superantigen may be influencing the immune system during the establishment of infection (Sutkowski et al., 1996). The cytotoxic T lymphocyte (CTL) response, which includes both CD4 1 and CD8 1 T-cells, appears directed against a few epitopes that are mainly derived from EBV nuclear antigens 3, 4 and 6 (Campos-Lima et al., 1997). During latency, virus-specific memory T-cells have a critical role in controlling EBV reactivation and restricting proliferation of EBV-infected B-cells (Rickinson et al., 1984; Kieff, 1996). Recently a method for determining the peripheral frequency of viral specific memory T-cells in previously infected individuals by flow cytometry has been described (Waldrop et al., 1997; Maino and Picker, 1998; Suni et al., 1998). Using this method peripheral blood mononuclear cells (PBMCs) are cultured in the presence of viral antigens and certain costimulatory molecules. The viral antigens are ingested and processed by the antigen presenting cells (APCs) which then present the antigens to memory T-cells. This antigen presentation results in the selective activation of memory CD4 1 T-cells which produce inflammatory cytokines such as interleukin-2 (IL-2) or interferon-g (IFN-g). The detection of T-cell production of IL-2 and IFNg by flow cytometry may then be used as a marker for viral specific memory cells. This method presents many advantages over typical long-term cultures to detect antigen-specific T-cell activation. By performing rapid activation (after 6 h) an accurate assessment of viral specific T-cells may be achieved without the interfering effects of proliferation, cell death or bystander T-cell activation.
This method has been utilized to detect peripheral memory T-cells specific for Cytomegalovirus (CMV; Waldrop et al., 1997; Suni et al., 1998), Human Immunodeficiency Virus (HIV; Moss et al., 1999) and Lymphocytic Choriomeningitis Virus (LCMV; Varga and Welsh, 1998a,b). To date, this method has not been utilized to detect peripheral T-cells specific for EBV. This is primarily due to the non-specific activation of T-cells which is induced by EBV viral proteins. In this report, we describe efforts to modify to the cell culture system in an attempt to reduce the non-specific activation of T-cells by EBV while allowing viral antigen presentation to continue uninhibited. In addition, we describe the use of multiparameter flow cytometry to distinguish the nonspecific activation of T-cells by EBV from legitimate viral-specific memory T-cell activation.
2. Materials and methods
2.1. Blood donors For assay development, whole blood samples were obtained from adult donors into ACD anticoagulant vacutainers. The subjects had been screened for most major infectious diseases and found to be healthy. Informed consent was obtained from all blood donors after the nature of the studies had been explained to them. Following serology testing, all of the randomly selected adult donors were found to be EBV-seropositive. To obtain EBV-seronegative specimens pediatric whole blood samples were obtained from a clinical facility and screened for their serology status. Four confirmed seronegative pediatric specimens were obtained as well as an adult seronegative specimen kindly donated by Dr. Raymond Widen, Tampa General Hospital.
2.2. Mitogens, antigens and viral lysates Mitogens, antigens and viral lystates used to supplement cell culture were as follows: CMV infected cell extract (Advanced Biotechnologies Inc., Columbia, MD-ABI); P3HR1 EBV infected cell extract (ABI, Columbia, MD); vero uninfected control extract (ABI, Columbia, MD); human foreskin fibroblast uninfected control extract (ABI, Columbia, MD); staphylococcus enterotoxin A (Sigma, St.
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
Louis, MO); phorbol myristate acetate (Sigma, St. Louis, MO); ionomycin (Sigma, St. Louis, MO).
2.3. PBMC /purified T-cell isolation For PBMC isolation, 4 ml of whole anticoagulated blood was diluted 1:1 with sterile phosphate buffered saline (PBS) and subjected to centrifugation over a Ficoll-Hypaque density gradient (Pharmacia, Upsala Sweden). The isolated cells were washed three times in sterile PBS and resuspended in RPMI medium 1640 (Gibco BRL, Grand Island, NY) containing 10% fetal bovine serum (GIBCO) and 10 000 Units / ml penicillin and 1000 mg / ml streptomycin (GIBCO). The final concentration of cells was determined using a hemacytometer. A final concentration of approximately 1.0310 6 cells / ml was used in all experiments. Purified T-cells were isolated from PBMCs using commercially available T-cell purification columns (R&D Laboratories Inc., Marina Del Ray, CA) according to the manufacturers instructions. Monocyte depleted PBMC’s were created via CD14-magnetic bead separation (Dynal, Oslo, Norway) according to the manufacturers’ instructions.
2.4. Cell culture conditions for activation For PBMC stimulation with viral or uninfected cell lysates, PBMCs were cultured for 1 h in the presence of 2.5 mg of purified anti-CD28 (PharMingen, San Diego, CA) antibody and the indicated amount of viral extract. After 1 h (to allow antigen uptake by APC’s) cells were washed in sterile HBSS and resuspended in medium containing 3 mM monensin. To verify T-cell viability and reactivity, PBMCs were cultured in either medium containing 10 ng / ml PMA (Sigma, St. Louis, MO) and 1 mg / ml ionomycin (Sigma) or in medium containing 10 mg / ml staphylococcus enterotoxin A (Sigma). In all cases 3 mM monensin (Sigma) was added to the PBMC cultures for the final 4 h to inhibit extracellular transport of cytokines and allow intracellular accumulation.
2.5. Fluorescent antibody staining Cell surface markers were stained first by resuspending the PBMC’s in 200 ml of PBS to which
37
0.5–1.0 mg of labeled mouse monoclonal antibodies to human CD69, CD4, CD25, CD3 or CD45 were added as indicated by the experiment and incubated at R.T. for 10 min. Antibody combinations and fluorochrome conjugates / emission wavelength are indicated in Table 1. The PBMC’s were then washed in PBS and fixed in 200 ml of 4.0% paraformaldehyde in PBS for 10 min, then washed again in PBS. To detect intracellular production of IFNg or IL-2 (following surface marker staining), the fixed PBMC’s were resuspended in 200 ml of permeabilization buffer, consisting of 5.0% non-fat dry milk and 0.5% saponin in PBS to which 0.5 mg of labeled mouse antibody to either IFNg or IL-2 (or both) was added. The cells were incubated at room temperature for 25 min and then washed in PBS containing saponin. The cells were then resuspended in 1.0% paraformaldehyde for analysis. The FITC conjugated IFNg and IL-2 antibodies were obtained from Becton Dickinson (Mountain View, CA) and all other antibodies were obtained from Beckman-Coulter (Miami, FL). The mean fluorescence intensity of the four color combinations were verified against single staining to ensure a lack of interference. In addition, the antibodies were verified in single, double and Table 1 Antibody panels used for flow cytometric analysis of EBV specific T-cells in peripheral blood specimens a Antibody panels for flow cytometry
Routine analysis Assay development
a
FL1
FL2
FL3
FL4
IFNg1IL-2 IFNg1IL-2 IFNg1IL-2 IFNg1IL-2
CD69 CD4 CD45RA CD25
CD3 CD3 CD3 CD3
CD45 CD45 CD45 CD45
Routine analysis indicates the antibody configuration suggested for routine evaluation of total EBV specific T-cells while monitoring non-specific activation. Assay development indicates the markers used to determine the phenotype of the EBV responding T-cells during initial experiments. FL1: fluorescence channel one (535 nm emission); FL2: fluorescence channel two (575 nm emission); FL3: fluorescence channel three (633 nm emission); FL4: fluorescence channel four (680 nm emission). The fluorochromes to which the antibodies were conjugated are as follows: FL1: FITC; FL2: PE; FL3: PE-Texas Red; FL4: PECyanin 5. Note that for FL1 detection of cytokines, a combination of antibodies to both IFNg and IL-2 (FITC conjugated) were used to detect production of either, or both cytokines in a single channel. FITC conjugated IFNg and IL-2 antibodies were obtained from Becton Dickinson (Mountain View, CA), all other antibodies were obtained from Beckman-Coulter (Miami, FL).
38
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
triple combinations to verify no artifactual interactions in the four color combination.
2.6. Flow cytometry analysis For data acquisition a Coulter XL flow cytometer was configured for four color analysis. Fluorescence emission channels (appropriate filter wavelengths) for the various labeled antibodies are indicated in Table 1. Instrument performance was verified by calibration with FlowSet fluorescent microspheres (Beckman Coulter). Appropriate electronic compensation for spectral overlap of the light emitted by the different fluorochromes was determined both by verification with a four-color compensation ‘kit’ (Beckman-Coulter) and by experimental internal verification with single-channel positive cell populations. The gating strategy devised for EBV T cell analysis is as follows: non-cellular and artifactual debris were excluded by gating CD45 1 cells from a CD45 vs. side scatter dot plot. From this population, T-cells were resolved by plotting CD3 vs. side scatter. At this point a clear resolution was visible between CD3 1 T-cells and non-CD3 1 cells. T-cells were then gated for analysis, and T-cell production of cytokines (IFNg1IL-2) were plotted vs. a secondary molecule of interest based on the experiment (either CD69, CD45RA, CD4 or CD25 — Table 1). The unique response of only the T-cell population was verified by plotting (from a CD45 1 lymphocyte gate) CD3 vs. cytokine production. Thresholds for positive were set using isotype matched control antibodies or internal cell populations negative for the antigens of interest. To quantitate viral specific T-cells (typically a rare-event population) we scanned a minimum of 100 000 T lymphocytes per analysis to ensure accurate percentages.
2.7. ELISA analysis The EBV serology status of the individual donors was assessed by serum or plasma ELISA by using commercial ELISA kits (Zeus Scientific Inc, Raritin, NJ) according to the manufacturers instructions. Circulating antibodies to EBV–VCA or EBV–EBNA were assayed and the serologic status was calculated as indicated by the manufacturer.
3. Results
3.1. Confirmation of flow cytometry assay for detection CMV-specific T-cells The method for detection of antigen-specific Tcells by flow cytometry has recently been described in a comprehensive fashion by Picker et al. (Waldrop et al., 1997; Suni et al., 1998). Prior to working with EBV viral antigens, for which a general method has not been described, we confirmed the basic method of Picker et. al. by utilizing CMV antigens. The CMV virus does not have superantigenic or mitogenic properties. We found that CMV specific T-cells were readily detectable in the peripheral blood of healthy CMV-seropositive donors using the reported method, and that peripheral blood cells from CMV-seronegative donors demonstrated no response. The responding T-cells expressed the early activation marker CD69, and were identified by their production of IFNg in response PBMC culture in the presence of CMV antigens (Fig. 1). In seropositve donors, the depletion of monocytes significantly abrogated the T-cell response to the viral antigens, indicating the APC dependent nature of the response (data not shown). The only deviation from the method described by Picker et al. was our use of a whole CMV infected cell viral lysate, suspended in a glycine buffer, which was obtained through commercial suppliers.
3.2. PBMC’ s cultured with EBV viral lysate or live virus indicate an EBV superantigen or mitogen property The literature has described the superantigen properties of the EBV virus, and the ability of EBV to activate T-cells in a non-specific fashion. Our initial experiments substituted either an infected cell viral lysate containing EBV proteins or purified live EBV for the CMV lysate in the method utilized above to observe the effects on T-cell activation. The expression of CD69 (as well as the production of IL-2 and IFNg) were assessed. We found that both the EBV viral lysate and the purified live EBV behaved similarly in that both agents did activate peripheral T-cells in a non-specific fashion (Fig. 2). In addition, the non-specific T-cell activation by both
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
39
Fig. 1. Detection of CMV specific T-cells in the peripheral blood of a CMV sero-positive healthy adult blood donor by three color flow cytometry. T-cells (CD3 1 ) have been gated and are plotted in these representative dot-plots. Interferon-g serves as an indicator for antigen specific T-cell activation, and is plotted with the early activation marker CD69. This experiment served to verify our ability to detect viral specific T-cells using the method recently described by Picker et al.
Fig. 2. T cell activation by EBV is nonspecific. Comparison of PBMC vs. purified T cell culture. Representative subject data demonstrating incubation of PBMC’s with live EBV virus or an EBV viral lysate (EBV antigens) induces broad-scale non-specific T-cell activation. 1.0310(6) PBMCs were cultured in the presence of 20 ml of control (uninfected cell) lysate or EBV viral lysate, 5 ml of highly purified live EBV virus, or 10 ng / ml PMA12 mg / ml ionomycin for 6 h. The cells were then stained for CD69 and CD3 1 T-cells were gated for analysis by two-color flow cytometry. Representative results obtained from one healthy adult donor. Note that the similar activation seen in purified T-cells (as compared to PBMC) indicates this activation is not APC dependent and is not the result of memory T-cells responding to specific viral antigen.
40
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
live EBV and the EBV lysate occurred when purified T-cells were substituted for PBMC’s. It was concluded that the EBV proteins may act on T-cells directly. Considering that a viral superantigen is supposed to activate (in the presence of MHC class II) only a particular Vb subset of T-cells, the broadscale activation of T-cells in the absence of APC’s seemed to indicate that the EBV proteins may have mitogenic properties as well as superantigen properties. It was concluded that the superantigenic / mitogenic properties of EBV, and the ability of EBV to activate T-cells in a non-specific fashion could prove to be a major impediment to the development of a routine, generalized flow cytometric assay for the detection of EBV specific T-cells.
3.3. Detection of EBV viral specific T-cells reactive to an EBV infected cell lysate It remained highly desirable to detect EBV-specific T-cells in the peripheral blood using the infected cell lysate. Utilization of the lysate would ensure that T-cells specific for all EBV proteins would be detected. In the method described by Picker et al. PBMC’s are co-cultured with viral antigens for 6 h to generate viral specific responses. Monensin is added to these cultures to allow cytokine accumula-
tion, but only after 1 h has passed to allow for antigen uptake and processing. We investigated the effects of removing the antigen from the culture medium after this 1-h antigen uptake phase. In theory, removal of antigen after 1 h should have no effect on the detection of antigen specific T-cells, since after 1 h the antigen has been internalized and is being processed by the APC’s. It was hypothesized that early removal of the antigen would abrogate the effects of an antigen having mitogenic properties that cause non-specific T-cell activation. An experiment was performed assessing CMV-specific T-cells (CMV antigens having no mitogenic or superantigen properties). This assay was performed incubating PBMC’s with CMV antigen1anti-CD28 for 6 h (monensin for final 5 h) or for 1 h with the antigen then removed (followed by 5 h monensin). Viral specific T-cells were then detected by cytokine flow cytometry. In this manner, CMV served as a control to validate that the early removal of viral antigens would not effect viral specific T-cell detection. Cytokine production was used to identify responding T-cells. It was found that antigen may indeed be removed after the initial 1 h incubation, prior to the 5 h monensin phase without loss sensitivity to viral specific T-cells (Fig. 3). This method was then applied to the detection of
Fig. 3. Detection of CMV specific T cells. Antigen exposure 1 hr vs. 6 hr. Early removal of viral antigen (after APC uptake) does not inhibit antigen presentation to T-cells. (A) PBMCs were incubated with CMV viral antigen for 1 h, followed by 5 h cytokine accumulation. The cells were then stained for CD3 in conjunction with intracellular IFNg1IL-2 for two-color analysis. Viral specific T-cells are identified as producing cytokine in response to incubation with viral antigen. A comparable result is seen in (B) where PBMC’s were incubated with CMV viral antigen and monensin for 6 h (representative of three experiments).
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
EBV specific T-cells. Monocyte depleted cultures were performed simultaneously to control for the APC dependent nature of the viral-specific T-cell response. The data indicated that EBV specific Tcells could be detected by this method with a reduction in non-specific T-cell activation (Fig. 4). The viral specific T-cell response was largely abrogated in the monocyte depleted cultures indicating the APC dependence of the response. In contrast, the responses to PMA1ionomycin were as great or greater in the monocyte-depleted cultures, confirming that the process of T-cell purification did not diminish the viability or responsiveness of the T-cell population (Fig. 4). These data may be contrasted with the data shown previously demonstrating that PBMC incubation with EBV proteins for a full 6 h resulted in monocyte independent non-specific T-cell activation (Fig. 2).
41
3.4. Routine detection and phenotype of viral specific T-cells Previous reports detecting viral specific T-cells monitored the production of IFNg, IL-2 or TNFa. Our goal was the development of an assay suited for a rapid quantitative assessment of all responding viral specific T-cells. To this end, cells producing any of these inflammatory cytokines, or any combination of them may be thought of as responding cells. We investigated the use of a pool of IFNg and IL-2 (both FITC labeled) in our cytometry assay. This allowed the detection of cells stimulated to produce either cytokine, or both cytokines simultaneously, in a single tube. An added benefit is the increased fluorescence levels of responding cells producing both cytokines. The use of a control lysate served as a monitor of constitutively activated T-
Fig. 4. Antigen specific T cell cytokine production in response to EBV is APC dependent. T cell production of IFNg1IL-2 in response to EBV antigens is antigen presenting cell dependent under modified culture conditions. Restriction of antigen exposure to 1 h allows the detection of legitimate EBV and CMV specific T-cells (although for EBV a superantigen effect cannot be ruled out by these results). The higher levels of cytokine production for the purified T-cells in response to PMA1ionomycin confirms their viability and reactivity. Representative two-color analysis results obtained from one healthy adult donor. The EBV and PMA / ionomycin results have been scaled as indicated to allow a graphical relative comparison of all data.
42 B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47 Fig. 5. Representative dot plots demonstrating the presence of EBV specific T-cells for an EBV sero-positive donor and the absence of EBV specific T-cells for an EBV sero-negative donor. Representative three-color analysis. Culture in the presence of EBV antigen (and mitogens / controls) was performed as described. To distinguish T-cell from non-T-cell cytokine production, lymphocytes were gated by CD45 expression and then T-cells (CD3 1 ‘Y axis’) were plotted vs. the cytokine combination of IFNg1IL-2 (‘X axis’). T-cell viability and ability to produce cytokines was assessed by responsiveness to PMA1ionomycin (monocyte independent) and SEA (monocyte dependent). Reactivity to a control lysate verified the absence of non-specific activation.
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
43
Fig. 6. Detection of EBV specific T-cells correlates with EBV serology status. The level of EBV specific T-cells was assessed by flow cytometry as described in three groups: a healthy adult donor group, a pediatric donor group obtained from a local hospital and an adult control group obtained from the same hospital. T-cell subsets producing IFNg1IL-2 in response to EBV antigen were observed in the EBV serology positive populations only. The adult hospital donor group served to verify that collection and any delays in transport did not effect T-cell viability or reactivity.
cells, which were essentially absent in most healthy donors (usually ,10 per 100 000 T-cells). Representative scatter plots demonstrating differences in EBV T-cell reactivity between an adult (EBV seropositive) clinical specimen and a pediatric (EBV sero-negative) clinical specimen which were collected and analyzed in parallel are presented in Fig. 5. In these scatter plots detection of cytokine production by PMA1ionomycin stimulated cultures served to confirm T-cell viability and ability to produce cytokines. Stimulation with the superantigen SEA was utilized to verify superantigen reactivity and monocyte presence, as well as T-cell–monocyte interactions. No cytokine reactivity is visible in response to EBV for the EBV-seronegative donor, whereas a well-defined cytokine reactivity is evident for the EBV-seropositive donor. In a separate experiment, we set up our cytometer for four-color analysis (IFNg1IL-2 / surface marker / CD3 / CD45) to assess the phenotype of the responding T-cells. The responding EBV-specific T-cells were found to primarily have a phenotype of IFNg/ IL-2 1 , CD45RA2 , CD69 1 , CD25 2 , CD3 1 , CD45 1 (data not shown). It was also found that if cultures were allowed to
progress to 24 h activation in the presence of antigen the viral reactive T-cells did express CD25. This finding and cellular phenotype is consistent with that of circulating previously unactivated memory Tcells.
3.5. Detection of EBV-specific T-cells corresponds with EBV serology status For routine clinical detection of total responding viral-specific T-cells, it was decided to utilize three color cytometry (IFNg1IL-2 / CD3 / CD45). This antibody combination would allow the detection of positively identified T-cells (CD3 1 ) responding to viral antigens. For rare event cytometry a clean analysis region is essential. We found the addition of CD45 allowed most cell debris to be gated out, resulting in a reasonably clean analysis even when collecting greater than 100 000 T-cell events. Initially, 14 healthy donors were screened for the presence of EBV specific T-cells using this assay, and the results were correlated with EBV antibody serology results. All the adult donors were found to have previous EBV exposure as determined by serology,
44
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
and also to have well-defined populations of EBVspecific T-cells as determined by flow cytometry (Fig. 6). To confirm the specificity of the assay it was necessary to obtain peripheral blood specimens from donors without previous EBV exposure (negative by EBV serology). Since greater than 95% of the adult population has been previously infected with EBV, obtaining these samples from a pool of healthy adult donors was not feasible. It was necessary to use clinical pediatric specimens, where the exposure rate is much lower. Pediatric blood samples were obtained and screened for EBV serology status. A lower infection rate was confirmed, and four EBV seronegative samples were obtained. To control for the unpredictable nature of such specimens, activation with SEA (monocyte dependent superantigen) and PMA1ionomycin (non-specific T-cell activation) were performed to verify the viability and reactivity of the peripheral T-cells, and their ability to be stimulated to produce cytokines. Pediatric clinical samples in which T-cell activation could not be demonstrated with SEA or PMA1ionomycin were rejected. In addition, serology-positive adult clinical specimens were always obtained in parallel with pediatric specimens to verify specimen transport did not affect the sample viability. The use of control antigens served to verify that no non-specific or constitutive T-cell activation was occurring. In addition to the four EBV sero-negative samples obtained, one EBV sero-negative adult sample from a known uninfected adult was obtained (kindly donated by Dr. Raymond Widen). The data revealed that in almost all the adult or pediatric serology-positive clinical samples, EBV-specific T-cells were detected by this method. In contrast, the four pediatric samples and one adult sample demonstrated to be EBV seronegative had no detectable EBV reactive T-cells by this method (Fig. 6). These data appear to confirm that the cells identified as EBV-specific T-cells by flow cytometry are indeed viral reactive T-cells, and not the result of a non-specific activation of T-cells by EBV superantigenic or mitogenic proteins.
3.6. Four color flow cytometry allows routine discrimination of EBV non-specific vs. antigenspecific T-cell activation Since human immunologic responses vary con-
siderably, T-cells producing cytokines in response to non-specific T-cell activation by EBV are problematic for this assay. It was desirable to develop a system to monitor non-specific T-cell activation in parallel with legitimate viral specific T-cell detection. An experiment was performed culturing PBMC’s with increasing amounts of EBV lysate (1 h incubation). T-cell activation was then assessed by cell surface CD69 and intracellular cytokine production simultaneously using four color flow cytometry (IFNg1IL-2 / CD69 / CD3 / CD45). Since the expression of the activation marker CD69 is easily inducible by most stimuli, CD69 expression was used to monitor non-specific activation in parallel with cytokine assessment for detecting viral specific T-cells. No non-specific activation was detected in PBMC cultures at minimal concentrations of EBV lysate. As EBV lysate concentrations were increased a dramatic onset of non-specific activation (CD69 expression) was observed (Fig. 7A). It was determined that 10–20 ml of EBV lysate per ml of culture was the maximal concentration of antigen that would completely avoid non-specific activation of T-cells. It is likely that the activation of T-cells progresses in sequence: first rapidly to the expression of surface CD69, and then later to the production of intracellular cytokines. When T-cells producing intracellular cytokines were assessed (simultaneously with CD69), it was found that a small population of responding T-cells could be detected producing cytokine following culture with concentrations of antigen lower than those required for broad-scale CD69 induction (Fig. 7A). A small but well-defined population of cytokine producing T-cells was visible following culture with 5 ml of lysate, and even more were detected following culture with 10 ml of lysate. Since these cells were detected as responding in the complete absence of large-scale non-specific T-cell activation as monitored by the induction of CD69, we conclude that these cells represent EBV specific T-cells. These cells were largely absent following culture with 1 ml of EBV lysate. At concentrations of lysate above 20 ml, even more T-cells were observed producing cytokines, however since large-scale expression of surface CD69 was evident, the interfering effect of non-specific activation was likely. A experiment of this type was also performed on a monocyte-depleted PBMC culture. The purpose of the both the purified T-cell experiment (Fig. 4) and
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
45
Fig. 7. Detection of viral specific T cells in conjunction with nonspecific T cell activation: PBMC vs. monocyte-depleted culture. Graphs demonstrating the broad scale induction of CD69 by EBV (d right Y-axis) and simultaneous detection of viral specific T-cells by cytokine expression (♦ left Y-axis) for increasing amounts of EBV antigen (X-axis). Representative data from an EBV sero-positive donor. Four-color analysis (IFNg1IL-2 / CD69 / CD3 / CD45) was performed allowing resolution of cells (CD45 1 ) vs. debris, followed by gating of T-cells (CD3 1 ) expressing either CD69 or cytokine. Viral specific T-cells are detected in the PBMC culture (A) mediated by the APC function of monocytes, and absent in the monocyte depleted culture (B). Although this experiment does not rule out a superantigen effect, the experiments with EBV sero-negative samples have indicated that the superantigen effect is not present in samples stimulated with 20 ml of antigen or less. The optimal amounts of lysate to avoid complete non-specific activation appear to be between 10 and 20 ml of EBV lysate per 1.0 ml of culture medium.
the monocyte depleted experiment (Fig. 7B) was to demonstrate that the cytokine production described (in response to low concentrations of EBV) was monocyte /APC dependent. In the monocyte depleted PBMC culture the ‘shoulder’ of cytokine producing T-cells that was observed in the PBMC culture at low concentrations of EBV lysate was absent. When the concentrations of EBV were increased in the monocyte-depleted culture the expression of CD69 and cytokines roughly tracked together, indicating that the initial cytokine producing T-cell population in the PBMC culture was indeed dependent on the presence of antigen-presenting cells (Fig. 7B). While it is possible that low numbers of CD14 2 APCs (i.e. dendritic cells) remain in the monocyte depleted culture, the loss of cytokine producing T-cells preceding large scale CD69 induction (and the purified T-cell experiments) argue for APC dependent specificity. These results, combined with the data obtained from the serology-negative donor population indicate that legitimate EBV specific T-cells are being detected with this assay.
4. Discussion Epstein–Barr virus infection, although primarily a subclinical juvenile disease, can progress to severe infections or malignancy in certain types of immunocompromised individuals. The documented cycling of the virus through latency to lytic replication, and the ability of the immune system to control lytic infection and restore latency has been recently welldocumented. Current opinion holds that the T-cell Th1 / Th2 bias of the peripheral immune system also plays a significant role in the maintenance of viral latency and the prevention of reactivation. T-cell mediated immune responses, particularly CD8 1 cytotoxic responses, and B-cell-mediated humoral responses under the control of CD4 1 helper T-cells thus coordinate their efforts to control persistent efforts of EBV to reactivate and establish lytic replication. The development of a rapid flow cytometric assay to enumerate peripheral viral specific T-cells has made routine assessment a reality. The clinical value
46
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
of monitoring the peripheral levels of viral specific T-cells has recently been established in the medical literature. The application of these assays to the determination of peripheral EBV specific memory T-cells has been hampered by the mitogenic or superantigenic properties of the EBV viral proteins. Culture of PBMCs in the presence of EBV antigens results in non-specific activation of T-cells. This has primarily been determined by the expression of CD69. The ability to culture viral antigens with a patients peripheral mononuclear cells is critical to the assay, as it allows monocytes to present the viral antigens to viral-specific T-cells (which then activate) allowing their detection by cytokine flow cytometry. The larger-scale non-specific activation of T-cells by EBV essentially creates a problem by masking the lesser population of legitimate viral specific T-cells. In this paper our efforts to overcome this difficulty are documented. We have demonstrated that PBMC culture in the presence of viral antigens for the entire 6 h activation period is not required, since after 1 h the monocyte population has ingested sufficient viral antigen to present the antigens to specific T-cell populations. Removing the viral antigens after 1 h of culture and then allowing APC-mediated T-cell activation and cytokine accumulation to progress normally results in equal detection levels of viral specific T-cells as compared to 6-h cultures (documented for a non-superantigen / mitogenic virus). We have found that the removal of EBV antigen after 1 h largely reduces the non-specific activation of T-cells, while antigen-presentation (now mediated by antigen presenting cells, i.e. monocytes) continues uninhibited. The expression of CD69 is a relatively easy to trigger marker for T-cell activation. This is partially due to the fact that CD69 protein is present in the cellular cytoplasm and does not require new gene expression to produce cell surface protein. For this reason, CD69 expression is widely used in the medical literature as a marker for cellular activation. The assessment of activation by assessing cytokine production by flow cytometry, however is a harderto-stimulate marker of activation. Commonly used T-cell mitogens, such as PHA, do not result in rapid cytokine accumulation to levels detectable by flow cytometry. In contrast, more powerful cellular ac-
tivators such as PMA1ionomycin result in dramatic well-defined populations of cytokine (IL-2 or IFNg) producing T-cells as determined by flow cytometry. It is our belief that at low levels of exposure to EBV proteins, non-specific T-cell activation largely does not progress to cytokine production. Thus, cytokine production may be used as a marker for the detection of EBV specific T-cells, provided care is used to ensure the concentration of EBV antigen remains below the threshold for non-specific activation (and above the threshold required for antigen presentation). By using four-color flow cytometry (IFNg/ IL2, CD69, CD3, CD45) we found it was possible to: (1) positively identify T-cells for analysis; (2) provide a debris free analysis region (essential for rareevent cytometry); (3) identify viral specific T-cells by cytokine production; (4) monitor non-specific activation via expression of CD69. By performing analysis using several concentrations of EBV antigens the maximal concentration of antigen may be selected for detection that does not result in broad scale T-cell activation as monitored by CD69 (Fig. 7). While this technique can largely control for mitogenic effects induced by EBV, it cannot control for a superantigen effect. To prove that positive events are not the result of a superantigen effect, we have relied on the negative results obtained using this method for EBV seronegative clinical samples (Fig. 6). The positive results obtained from EBV seropositive samples and the negative results obtained from seronegative samples seem to confirm the detection of EBV specific T-cells. Given the ease and utility of the method described by Picker et al., we did not feel that an inability to detect EBV specific T-cells would be a problem. Clearly, in normal circumstances the coincubation of antigen, costimulatory molecules and PBMC’s will result in antigen-specific T-cell detection. The primary problem for EBV has continued to be the ‘contamination’ of the positive events with T-cells stimulated in a non-specific fashion by the mitogenic or superantigenic properties of the virus. Here, we believe a method has been described which allows the detection of EBV specific T-cells, while monitoring the advent of non-specific T-cell activation. Future studies with samples obtained from patients suffering from active infection, post-active infection (recovery), as well as adults suffering latent EBV
B.E. Crucian et al. / Journal of Immunological Methods 247 (2001) 35 – 47
reactivation as a result of stress (or other factors) will hopefully demonstrate the clinical utility of this method as an assessment of the immune response to latent viral reactivation.
Acknowledgements The authors wish to thank Dr. Ray Widen, Tampa General Hospital for providing EBV seronegative whole blood samples and for critical review of this manuscript. We also wish to thank the St. John Hospital, Clinical Laboratory, Houston, Texas, for providing pediatric whole blood specimens.
References Campos-Lima, P.O., Levitsky, V., Imreh, M.P., Gavioli, R., Masucci, M.G., 1997. J. Exp. Med. 186, 83–89. Glaser, R., Friedman, S.B., Smyth, J., Ader, R., Bijur, P., Brunell, P., Cohen, N., Krilov, L.R., Lifrak, S.T., Stone, A., Toffler, P., 1999. The differential impact of training stress and final examination stress on herpesvirus latency at the United States Military Academy at West Point. Brain Behav. Immun. 13, 240–251. Glaser, R., Pearson, G.R., Jones, J.F., Hillhouse, J., Kennedy, S., Mao, H.Y., Kiecolt-Glaser, J.K., 1991. Stress-related activation of Epstein–Barr virus. Brain Behav. Immun. 5, 219–232. Hamilton-Dutoit, S.J., Pallesen, G., Franzmann, M.B., Karkov, J., Black, F., Skinhoj, P., Pedersen, C., 1991. AIDS-related lymphoma. Histopathology, immunophenotype, and association with Epstein–Barr virus as demonstrated by in situ nucleic acid hybridization. Am. J. Pathol. 138, 149–163. Henle, G., Henle, W., Diehl, V., 1968. Relation of Burkitt’s tumor-associated herpes-type virus to infectious mononucleosis. Proc. Natl. Acad. Sci. USA 59, 94–101. Herbst, H., Dallenbach, F., Hummel, M., Niedobitek, G., Pileri, S., Muller-Lantzsch, N., Stein, H., 1991. Epstein–Barr latent membrane protein expression in Hodgkin and Reed-Sternberg cells. Proc. Natl. Acad. Sci. USA 88, 4766–4770.
47
Kieff, E., 1996. Epstein-Barr virus and its replication. In: Fields, B.N., Knipe, D.M., Howley, P.M. (Eds.), Virology. 3rd ed., Vol. 2, Raven Press, New York, N.Y., pp. 2343–2396. Maino, V.C., Picker, L.J., 1998. Identification of funtional subsets by flow cytometry: intracellular detection of cytokine expression. Cytometry 34, 207–215. Miyawaki, T., Kasahara, Y., Kanegane, H. et al., 1991. Clin. Exp. Immunol. 83, 447–451. Moss, R.B., Wallace, M.R., Giermakowska, W.K., Webb, E., Savary, J., Chamberlin-Brandt, C., Theofan, G., Musil, R., Richieri, S.P., Jensen, F.C., Carlo, D.J., 1999. Phenotypic analysis of human immunodeficiency virus (HIV) type 1 cell-mediated immune responses after treatment with an HIV-1 immunogen. J. Infect. Dis. 180, 641–648. Pallesen, G., Hamilton-Dutoit, S.J., Rowe, M., Young, L.S., 1991. Expression of Epstein–Barr virus latent gene products in tumor cells of Hodgkin’s disease. Lancet 337, 320–322. Rickinson, A.B., Rowe, M., Hart, I.J., Yao, Q.Y., Henderson, L.E., Rabin, H., Epstein, M.A., 1984. Cell. Immunol. 87, 646–658. Suni, M.A., Picker, L.J., Maino, V.C., 1998. Detection of antigenspecific T-cell cytokine expression in whole blood by flow cytometry. J. Immunol. Methods 212, 89–98. Sutkowski, N., Palkama, T., Ciurli, C., Sekaly, R.P., ThorleyLawson, D.A., Huber, B.T., 1996. An Epstein–Barr virusassociated superantigen. J. Exp. Med. 184, 971–980. Varga, S.M., Welsh, R.M., 1998a. Detection of a high frequency of virus-specific CD4 1 T-cells during acute infection with lymphocytic choriomeningitis virus. J. Immunol. 161, 3215– 3218. Varga, S.M., Welsh, R.M., 1998b. Stability of virus-specific CD4 1 T-cell frequencies from acute infection into long term memory. J. Immunol. 161, 367–374. Waldrop, S.L., Pitcher, C.J., Peterson, D.M., Maino, V.C., Picker, L.J., 1997. Determination of antigen-specific memory / effector CD4 1 T-cell frequencies by flow cytometry: evidence for a novel, antigen-specific homeostatic mechanism in HIV-associated immunodeficiency. J. Clin. Invest. 99, 1739–1750. Yao, Q.Y., Rickinson, A.B., Epstein, M.A., 1985. Int. J. Cancer 35, 35–42. Zur Hausen, H., Schulte-Holthausen, H., Klein, G., Henle, W., Henle, G., Clifford, P., Santesson, L., 1970. EBV DNA in biopsies of Burkitt tumours and anaplastic carcinomas of the nasopharynx. Nature 228, 1956–1958.