VIRAL IMMUNOLOGYVolume 9, Number 1, 1996Mary Ann Lieben, Inc.Pp. 45-50

Decreased Levels of Circulating IFN-a and IncreasedsCD23 in Patients with Acute Infectious Mononucleosis

ANILA PRABHU,1 MARION WARWICK,24 and AMBIKA MATHUR3

ABSTRACT

Epstein-Barr virus (EBV) is the etiological agent for acute infectious mononucleosis (AIM).It is also associated with certain malignant disorders in individuals with immunodeficien-cies such as B cell lymphoproliferative disorder (BLPD). Our previous study with BLPD pa-tients had demonstrated significantly higher serum IL-4 and IgE levels and significantly de-creased IFN-a levels. These observations were consistent with the model of regulation of Bcell growth by T cell-derived cytokines, in which IL-4 promotes B cell growth and switch toIgE synthesis whereas IFN-a and IFN-y inhibit these IL-4-mediated effects. Since AIM isalso EBV associated, this study was designed to examine IL-4, IFN-a, IFN-y, IgE, and sol-uble CD23 (sCD23) serum levels in AIM patients. In this study we report for the first timethat in contrast to BLPD patients, AIM patients did not exhibit increased levels of IL-4 andIgE; however AIM patients do exhibit decreased IFN-a levels and, additionally, also exhibitsignificantly higher sCD23 levels. This could result in B cell activation and have implica-tions for the survival of the virus in B cells.

INTRODUCTION

Epstein-Barr vims (EBV) is a human herpesvirus with a DNA genome (8). The vims usually infects theoropharyngeal epithelium and B-lymphocytes and has been documented to be the causative agent of

certain benign diseases (8). The consequence of a primary infection with the vims usually is an asympto-matic infection during which period the vims is latent. Alternatively, the presence of the vims could resultin the manifestation of acute infectious mononucleosis (AIM). EBV is also associated with certain malig-nant disorders in individuals with inherited or acquired immunodeficiencies such as X-linked lymphopro-liferative syndrome, acquired immunodeficiency syndrome (AIDS), patients receiving immunosuppressivetherapies, or having partial T cell defects (reviewed in 8). B cell lymphoproliferative disorder (BLPD) isalso EBV associated and is a frequently fatal complication of immunodeficiency (2).In our previous study we had observed significantly higher levels of serum IgE and interleukin (IL)-4

and significantly lower levels of interferon (IFN)-a in BLPD patients as compared to healthy control indi-viduals (7). These observations were consistent with the well-established model of regulation of B cellgrowth by T cells and T cell-derived cytokines (reviewed in 14). The lymphokines IL-4, IFN-a, and IFN-

1 Department of Oral Science, 2Boynton Health Service, and departments of Oral Science and Pediatrics, Universityof Minnesota, Minneapolis, Minnesota 55455.

4Present address: University of Massachusetts, Boston, Massachusetts 01060.

45

PRABHU ET AL.

y play an important role in B cell growth and development; IL-4 promotes B cell proliferation and inducesB cells to switch to IgE synthesis (14), while IFN-a and IFN-y inhibit these effects of IL-4 (14), main-taining a balance between B cell proliferation and IgE synthesis. Since AIM is also caused by EBV, we in-vestigated whether a similar regulatory mechanism by T cells and cytokines was also observed in AIM pa-tients. Imbalances in the cytokine network have been documented for viral diseases and our interest was toobserve if changes in the cytokine pattern occurred in EBV-associated AIM.

Thus, in this study we have made simultaneous measurements of IFN-a, IFN-y, and IL-4 serum levelsin patients diagnosed for AIM and compared these levels wih EBV seropositive normal controls. AlthoughAIM is characterized by pronounced IgM antibody response (10), we have also examined the levels ofserum IgE as an additional marker for either functional excess of IL-4 and/or a relative deficiency of IFN-a. In addition, we also measured the levels of soluble CD23 (sCD23) in the circulation, since expressionof membrane CD23 (low affinity receptor for IgE) has been reported to increase in AIM (3), and CD23 isa ligand for CD21, which is the receptor for EBV on B cells (1).

MATERIALS AND METHODS

Patients. Seven patients diagnosed as having EBV-induced infectious mononucleosis at Boynton Clinic,University of Minnesota, were examined in this study. The patients ranged from 19 to 27 years; the groupincluded 4 females and 3 males (summarized in Table 1). Healthy controls included 6 males and 6 femalesranging in age from 24 to 44 years, all of whom were EBV seropositive. Informed consent was obtainedfrom patients and controls, as per the human experimentation guidelines of the U.S. Department of Healthand Human Services and the University of Minnesota. Patient samples were obtained at the time of initialdiagnosis, before any treatment was initiated.Determination of serum IL-4, IFN-a, IFN-y, IgE, and sCD23 levels. Serum was collected from clot-

ted blood, aliquoted and frozen at—

70°C until required.IL-4 was measured in an ELISA assay using a kit from R&D Systems (Minneapolis, MN). IFN-a was

measured using an ELISA kit obtained from Biosource International (Camarillo, CA). The ELISA kit usedto measure IFN-y levels was obtained from Genzyme (Cambridge, MA). IgE levels were determined us-

ing a sandwich ELISA assay. Briefly, ELISA plates were coated with a monoclonal mouse antihuman IgEantibody (ICN, Lisle, IL). After the appropriate incubation and washing steps serum sample was added tothe plate, followed by the addition of biotinylated polyclonal goat antihuman IgE antibody (Vector labora-tories, Burlingame, CA). This step was followed by addition of avidin conjugated to alkaline phosphatase

Table 1. Levels of Cytokines, IgE, and sCD23 for Each AIM Patient

AIMpatientnumber Sex

Age(years)

IFN-a(pg/ml)

IgE(fig/ml)

sCD23(fig/liter)

1234567

FMFFFMM

22192027242420

7.5236.918.97.57.5

124.3395.6

6.1731.6112.1418.959.971.30.48

3.077.464.42.494.994.537.46

Control (mean ± SE) 317.3 ±23.1 29.25 ± 7.65 1.24 ±0.21

46

IFN-a AND sCD23 IN AIM PATIENTS

(Vector) and then substrate paranitrophenylphosphate (Sigma, St. Louis, MO) was added. Color develop-ment was measured on an ELISA reader at 405 nm. Human IgE myeloma (CalBiochem, San Diego, CA)was used as standard for reference. sCD23 levels were measured in an ELISA assay using a kit from BindingSite (Birmingham, UK).Statistical analysis. Statistical analysis of the data was done by Students' unpaired t test. For the pur-

pose of analysis, subject whose cytokine levels were below the lower limit of sensitivity of the assays wereassigned a cytokine value at the midpoint between 0 and the lower limit of detection for each assay. Datahave been expressed as mean ± SE.

RESULTS

Serum IL-4 levels. The lower limit of detection of the IL-4 ELISA assay was 30 pg/ml. Normal serumvalues of IL-4 are less than 30 pg/ml. Serum samples from AIM patients and from normal control indi-viduals had IL-4 levels less than the detection level of the kit.Serum IFN-y levels. The lower level of detection of the IFN-y kit was 100 pg/ml. We observed that

IFN-y levels of serum samples from AIM patients and normal healthy controls were in the normal range,since they were lower than the detection level of the ELISA assay used for detection of IFN-y.IFN-a levels. We observed that IFN-a levels of AIM patients were significantly lower (p < 0.001) than

the control group (Fig. 1). The lower level of detection for the IFN-a kit was 15 pg/ml. The mean con-

centration of IFN-a of healthy control individuals was 317.3 ± 23.06 pg/ml.IgE levels. As shown in Figure 1 no significant difference was observed between serum IgE levels of

AIM patients and the healthy control group. The lower detection limit of the IgE ELISA was 0.1 pg/ml.The mean concentration of IgE of healthy control individuals was 29.25 ± 7.65 pg/ml.sCD23 levels. sCD23 levels of AIM patients were significantly higher (p < 0.0001 ) compared to normal

control group (Fig. 1). The mean concentration of sCD23 in healthy controls was 1.24 ± 0.21 /ng/liter andthe lower detection level of the sCD23 kit used for the assay was 0.7 pg/liter. The data for each patientsample have been summarized in Table 1.

600

500

400

IFN-alpha

1 300

200

100

'S)

CONTROL AIM

sCD23

CONTROL AIM CONTROL AIM

FIG. 1. IFN-a, IgE, and sCD23 levels were determined as described in Materials and Methods. The circulating lev-els of IFN-a, IgE, and sCD23 of normal healthy control individuals (O) and acute infectious mononucleosis patients(•) along with the mean ± SE of each group is also shown. The lower level of detection for each assay is shown bya solid line.

47

PRABHU ET AL.

DISCUSSION

In AIM although a humoral response is directed toward viral antigens of the lytic cycle it is generallyaccepted that it is the cell-mediated immunity that has a protective role in controlling the level of EBV-in-fected cells (5). It has been proposed that the disease is a consequence of the immune response generatedto control the virus and not an effect of viral replication. The role of cytokines in the generation of immuneresponse may result in the clearance of the virus. We have therefore examined serum cytokine levels ofhealthy controls and patients. Additionally, in a previous study we observed patients with BLPD, which isalso an EBV-associated immunodeficiency, had significantly higher levels of IL-4 and IgE and significantlylower levels of IFN-a (7). Since these results corresponded well with the model of regulation of B cellgrowth by T cells, we investigated whether a similar mechanism operated in AIM patients.It has been suggested on the basis of cytokine production that two subsets of T helper cell population

are present, although unlike the murine system, the distinction between the two subsets (TH1 and TH2) isnot as clearly demarcated in the human system (reviewed in 11). In addition to other cytokines, TH1 cellsproduce IL-2, IFN-a, and IFN-y, while TH2 cells produce IL-4 and IL-5 (11). IL-4 promotes B cell pro-liferation and switch to IgE synthesis. It also upregulates expression of membrane/surface CD23, the lowaffinity receptor for IgE on B lymphocytes (9). Membrane CD23 was first identified as an EBV-induced Bcell marker that undergoes autoproteolytic cleavage, giving rise to a soluble form of CD23 (12). Severalreports have suggested an IgE regulatory activity for sCD23 (12). Additionally, it has been demonstratedthat in AIM patients there is an increase in surface expression of CD23 levels in peripheral blood B cells,macrophages, and monocytes (3). We therefore examined the levels of IL-4, IgE, and sCD23 in AIM pa-tients. Since both IFN-a and IFN-y inhibit IL-4 activity we were interested in examining levels of IFN-aand -y in serum samples of AIM patients and control healthy individuals.

Our results indicate that although both BLPD and AIM are EBV associated, there are differences in thecytokine profiles between the two. Unlike BLPD, IL-4, IFN-y, and IgE levels of AIM patients were not

significantly different from healthy control group. A significant decrease was observed in IFN-a levels inAIM patients, which corresponds to the decrease observed in BLPD patients. It is, therefore, possible thatin AIM patients, cytokines or factors other than IL-4, IFN-a, and IFN-y regulate the IgE synthesis. This isthe first report of decreased IFN-a levels in AIM patients since in an earlier study (6) no detectable amountof IFN-a was present in either the healthy controls or in AIM patients during the onset of the infection. Wealso observed a significant increase in the sCD23 levels in AIM patients. It has been reported that there isan increase in the surface expression of CD23 on PBLs of AIM patients (3). The increase that we observein sCD23 levels supports this observation since CD23 undergoes autoproteolytic cleavage resulting in sCD23(12). sCD23 plays an important role in the B cell activation and differentiation (4). Thus our results indi-cate that the subsequent increase in sCD23 levels would result in B cell activation that could be advanta-geous to the survival of the virus in B cells.

We also observed that there were no significant differences in IFN-y levels of healthy controls and AIMpatients. Virelizier et al. have also reported decreased IFN-y levels in EBV infection (17). However, twoearlier studies (6,13) have reported increased IFN-y levels in serum from AIM patients. The differences inthe results could be probably due to the strain differences of the infecting EBV.

The open reading frame of BCRF-1 gene in EBV genome displays remarkable homology with humanIL-10. The protein coded by the BCRF-1, which is viral IL-10, shares a number of biological propertieswith human IL-10 (15). Since IL-10 inhibits IFN-y production by PBMCs (15), it is possible that the lowerlevels of IFN-y that we observe are due to the presence of IL-10 (viral or human). Our preliminary dataindicate that IL-10 levels in some AIM patients that we examined were significantly higher than those ofhealthy control individuals. Taga et al. (16) recently reported that 60% of patients with AIM have signifi-cantly higher levels of circulating vIL-lO/hIL-10 and 18% of the patients have significantly higher levelsof vIL-10 compared to healthy control individuals. We further plan to examine IL-10 levels of AIM pa-tients and distinguish whether the source of IL-10 is viral or human. Additionally, cytokine mRNA levelsof peripheral blood mononuclear cells of healthy controls and patients would also be examined to deter-mine if differences observed at the protein level are observed at the transcriptional level too.In conclusion, our data demonstrate a specific decrease in the circulating levels of the TH1 -derived cy-

IFN-a AND sCD23 IN AIM PATIENTS

tokine IFN-a and an increase in the circulating levels of sCD23, suggesting that a partial TH1 dysfunctionis observed with the development of EBV-associated AIM. This is in contrast to our study of EBV-associ-ated BLPD in which we have observed alterations in the levels of both TH1- and TH2-derived cytokines.The balance between the TH1 and TH2 subsets of the T helper cells and the importance of the cytokinepattern changes have been documented for viral diseases. In this study we report the novel finding that theTH1/TH2 imbalance in patients with AIM is not as profound as that seen in patients with BLPD, whichmay explain the fact that AIM is a self-limiting infection while BLPD is often fatal.

ACKNOWLEDGMENTS

Supported in part by Emma B. Howe Foundation and NIH/NIDR P30DE09737.

REFERENCES

1. Aubry, J.P., S. Pochon, P. Graber, K.U. Janse, and J-Y. Bonnefoy. 1992. CD21 is a ligand for CD23 and regulatesIgE production. Nature (London) 358:505-507.

2. Block, AW. 1991. Cytogenetic analysis of B cell chronic lymphoproliferative disorder: New technology, com-plexities and promises. Prog. Clin. Biolog. Res. 368:145-173.

3. Fumkawa, S., T. Motohashi, T. Matsubara, T. Watanabe, T. Obara, and K. Yabuta. 1992. Increased expression ofCD23 on peripheral blood B cells and macrophages/monocytes during acute infectious mononucleosis. J. Infect.Dis. 166:691-692.

4. Gordon, J., L. Flores-Romo, J.A. Cairns, M.J. Millsum, P.J. Lane, GD. Johnson, and I.C.M. MacLennan. 1989.CD23: A multi-functional receptor/lymphokine? Immunol. Today 10:153-157.

5. Henle, W., and G. Henle. 1979. Seroepidemiology of the virus, Pp. 297-320, In M.A. Epstein and B.G Achong(eds.): The Epstein-Barr vims. Springer-Verlag, Berlin.

6. Linde, A., B. Andersson. S.B. Svenson, et al. 1992. Semm levels of lymphokines and soluble receptors in primaryEpstein-Barr virus infection and in patients with chronic fatigue syndrome. J. Infect. Dis. 165:994-1000.

7. Mathur, A., D.M. Kamat, A.H. Filipovich, M. Steinbuch, and R. Shapiro. 1994. Immunoregulatory in patients withEpstein-Barr vims-associated B cell lymphoproliferative disorders. Transplantation 57:1042-1045.

8. Moss, B.J., S.R. Burrows, R. Khanna, I.S. Misko, and T.B. Sculley. 1992. Immune surveillance against EBV. Sem.Immunol. 4:97-104.

9. Paul, W.E. 1991. Interleukin-4: A prototypic immunoregulatory lymphokine. Blood 77:1859-1870.

10. Pearson, GR., and J. Luka. 1986. Characterization of the virus-determined antigens, Pp. 48-73. In M.A. Epsteinand B.G. Achong (eds.): The Epstein-Barr virus. Recent advances. John Wiley, New York.

11. Romagnani, S. 1992. Human TH1 and TH2 subsets: Regulation of differentiation and role in protection and im-munopathology. Int. Arch. Allergy Immunol. 98:279-285.

12. Sarfati, M., E. Rector, M. Rubio-Trujillo, K. Wong, A.H. Sehan, and G. Delespesse. 1984. In vitro synthesis ofIgE by human lymphocytes. III. IgE potentiating activity of culture supematants from Epstein-Barr vims (EBV)transformed B cells. Immunology 53:207-214.

13. Schuster, V., M. Herold, H. Wächter, and G. Reibnegger. 1993. Serum concentrations of interferon gamma, inter-leukin-6 and Neopterin in patients with infectious mononucleosis and other Epstein-Barr vims-related lymphopro-liferative diseases. Infection 21:210-213.

14. Snapper, CM., and W.E. Paul. 1987. Interferon gamma and B cell stimulatory factor-1 reciprocally regulate Ig iso-type production. Science 236:944-947.

15. Spits, H., and R. de Waal Malefyt. 1992. Functional characterization of human IL-10. Int. Arch. Allergy Immunol.99:8-15.

49

PRABHU ET AL.

16. Taga, H., K. Taga, F. Wang, J. Chretien, and G. Tosato. 1995. Human and viral Interleukin-10 in acute Epstein-Barr virus-induced infectious moonucleosis. J. Infect. Dis. 171:1347-1350.

17. Virelizier, J.L., G. Lenoir, and C. Griscelli. 1978. Persistent Epstein-Barr virus infection in a child with hyper-gammaglobulinaemia and immunoblastic proliferation associated with a selective defect in immune interferon sere-

tion. Lancet ii:231-234.

Address reprint requests to:Ambika Mathur, Ph.D.

Department of Oral Science, 17-252, Moos Tower515 Delaware Street SEMinneapolis, MN 55455

50