Jan 19, 1993 - enhances antigen presentation by splenic macrophages to T-cells. ... antigen presentation and cytokine production in the initiation of the ...
Int. J. lramunopharmac., Vol. 15, No, 4, pp. 463-468, 1993. Printed in Great Britain.
0192-0561/93 $6.00 + .00 Pergamon Press Ltd. International Society for Immunopharmacology.
MELATONIN INCREASES ANTIGEN PRESENTATION AND AMPLIFIES SPECIFIC AND NON SPECIFIC SIGNALS FOR T-CELL PROLIFERATION CLAUDIO PIOLI, M. CRISTINA CAROLEO, GIUSEPPE NISTICO' and GINO DORIA* Division of Physics and Biomedical Sciences, ENEA C.R.E. Casaccia, Rome, and Department of Biology, University of Rome, Tor Vergata, Italy (Received 19 January 1993 and in final form 1 March 1993)
-Our preceding results have shown that melatonin administration to normal and immunodepressed mice increases significantly the antibody response. We also found that melatonin is able to restore the impaired T-helper cell activity in immunodepressed mice. The present study shows that melatonin enhances antigen presentation by splenic macrophages to T-cells. This effect is concomitant with an increase in the expression of MHC class II molecules and production of IL-1 and TNF-a. Considering the role of antigen presentation and cytokine production in the initiation of the immune response, the present findings provide evidence for relevant mechanisms that may account for the regulatory role of the pineal gland in immunoregulation.
Abstract
Several interactions among the nervous, endocrine and immune systems have been described in the past few years. It is well recognized that the pineal gland is an organ of neuroendocrine relevance in animals and man (Axelrod, Fraschini & Veto, 1982; Preslock, 1984) and that melatonin (MEL), its main neurohormone, affects the immune system (Maestroni, Conti & Pierpaoli, 1986, 1987; Lissoni et al., 1986). Since MEL secretion is modulated by daily l i g h t - d a r k alternation (Wurtman, Axelrod & Phillips, 1963) as well as by seasonal changes in enviromental light (Lewy, 1983), several studies have shown that abrogation of the cyclicity of MEL secretion, by evening administration of beta-blockers (RadosevicStasic, Jonic & Rukavina, 1983) or by permanent lighting leads to impairment of cellular and humoral immune responses in mice. Our preceding results (Caroleo, Frasca, Nistico' & Doria, 1992) have shown that MEL administration to normal and immunodepressed mice increases in vivo and in vitro antibody responses significantly. It was also found that MEL is able to restore the impaired T-helper cell activity in immunodepressed mice. These effects may result from MEL-induced activation of accessory ceils such as B-cells and
macrophages (M+). It is well established that the role of accessory cells in the initiation of the immune response is dependent on their capacity to generate both specific and non-specific signals recognized by T-cells. The specific signal is the antigenic peptide - MHC class II molecule complex displayed on the surface of accessory cells acting as antigen-presenting cells (APC) (Moller, 1987; Schwartz, 1985). The non specific signals are cytokines, such as IL-1 and TNF-t~, which are required for T-cell proliferation (Weaver & Unanue, 1990). In the present study MEL was found to enhance antigen presentation by M+ in concomitance with an increase in cytokine production and MHC class II expression. EXPERIMENTAL
PROCEDURES
Animals (C57BL/10 x DBA/2)FI male mice were used at the age of 2 months. Mice were housed in plastic cages, 4 animals per cage, and fed with standard diet and chlorinated water ad libitum. The animals were
*Author to whom correspondence should be addressed at: Division of Physics and Biomedical Sciences, ENEA C.R.E. Casaccia, C.P. 2400, 00100 Rome A.D., Italy. 463
464
C. PIOLI et al.
maintained on a 12 h light/12 h dark cycle from 6.00 a.m. to 6.00 p.m. and at constant temperature (20 +_ I°C).
MEL treatment MEL (Sigma M-5250, lot 108F0646), whose purity was greater than 99°7o as certified by Sigma (thin layer chromatography), was dissolved in ethanol and then diluted in pyrogen-free saline to reach the final concentration of 1% ethanol. Dosage and route of administration were 10 mg/kg in 0.2 ml pyrogenfree saline, injected s.c. daily at 4.00 p.m. for 4 or 5 consecutive days. Control mice received pyrogenfree saline alone.
Preparation of antigen-pulsed M# Eighteen hours after the last injection, splenic cells from control or MEL-treated mice were suspended in medium (RPMI 1640, Sigma), exposed to osmotic shock to remove red blood cells, and then incubated (5 × 10 6 cells/ml) for 1 h at 37°C on plastic plates (Falcon 3003). After removing the non-adherent cells (5 washes), adherent cells (M+) were collected by gently scraping the plates with a rubber policeman. Md~ (5 × 106/ml) were pulsed with HEL (200/ag/ml) for 2 h at 37°C in polypropylene tubes (Falcon 2059). After pulsing, Md~ were washed 3 times and resuspended with medium supplemented with 10% fetal calf serum (Flow), 2 mM L-glutamine (Gibco), 5 × 10 -5 M 2-mercaptoethanol and 10/ag/ml gentamicin (Shering). The Md~ suspended in the supplemented medium (complete medium) were added to cultures at the indicated concentrations.
Lymphocyte priming For the antigen-specific T-cell proliferative assay (Corradin, Etlinger & Chiller, 1977), mice were injected s.c. at the base of the tail with HEL (Societa' Prodotti Antibiotici) 100/ag/mouse in 50/al saline emulsified with an equal part of CFA containing 1 mg/ml Mycobacterium tuberculosis strain H37 Ra (Difco), 8 days before culture.
Antigen specific T-cell proliferation assay Eight days after lymphocyte priming by HELCFA, inguinal lymph node cells (LNC) were distributed in microtiter plates (Falcon 3040). Each culture containing 4 × 105 cells/0.1 ml complete medium, received 0.1 ml complete medium alone (background control) or containing a given number of antigen-pulsed Md?. Cultures were incubated for
5 days at 37°C in 5°7o CO2 in humid air, and 18 h before harvest each culture was pulsed with 0.5 taCi tritiated thymidine (spec. act. 1.739 GBq/mmol, Amersham). The cells were then harvested by an automated cell harvester (Micromate 196 Cell Harvester, Packard Instruments) on a filter support and analyzed by Packard's Matrix 96 Direct Beta Counter without scintillation fluid. Results are expressed as arithmetic mean of counts/min from triplicate cultures minus background.
Production and titration of cytokines M~ (106/ml), prepared as described above from control or MEL-treated mice, were cultured in 48 well plates (Costar 3548) in complete medium alone or containing LPS (20/ag/ml, Difco). After 24 h, culture supernatants were collected, dialyzed with phosphate-buffered saline (PBS), filtered (0.2 lain, Sartorius) and stored at - 20°C until tests were performed. IL-I activity was titrated by a biological assay. Briefly, supernatants were 2-fold serially diluted in complete medium, and 100 ~l of each dilution were added to triplicate wells (Falcon 3040) containing 100/al of a cell suspension of thymocytes (106/well) from C3H/HeJ mice and P H A (M form, 1.2°/0 final concentration, Gibco). Triplicate control wells contained thymocytes and P H A only. After 3 days incubation at 37°C, all cultures received 0.5/aCi of tritiated thymidine and the radioactivity uptake was measured as describe above. TNF-a was titrated by an ELISA kit (Factor-test mTNF-a, Genzyme). A reference straight line obtained by plotting the absorbance versus the standard TNF-a concentrations was used to calculate the TNF concentrations and standard error (S.E.) of the experimental samples.
Flow cytometry Reagents to detect MHC class II molecules were anti-I-Ad (clone AMS-32.1) and anti-I-E (clone AMS-16.1) biotin-conjugated mAbs (Pharmingen). Biotin-conjugated mAbs were revealed by phycoerythrin-conjugated streptavidin (Becton-Dickinson). Splenic M+ purified by adherence were incubated with both biotin-conjugated mAbs for 30 min on ice, washed, and then resuspended with phycoerythrin-conjugated streptavidin. After 30 rain incubation on ice, cells were washed and used for flow cytometry. Samples of 10,000 viable cells were analyzed and fluorescence signals were collected in log mode.
Melatonin Increases Antigen Presentation
465
3000
,°° l 2O0O .=_o
1000
10 "
tD
1
0 0
100
1000
Macrophages / culture ( xlO 3 )
Fig. 1. Effect of MEL on antigen presentation. Antigen specific T-ceil proliferation of LNC from HEL/CFAprimed mice was induced by graded numbers of HELpulsed M~ from MEL-treated (closed circles) or from control mice (open circles). RESULTS
Antigen presentation Figure 1 shows the proliferation of LNC from mice primed with H E L - C F A as induced by graded numbers of HEL-pulsed Mdp. M~ from MEL-treated mice are more effective than Md? from control mice in stimulating LNC proliferation. Analysis of variance of the linear regressions fitted by the least squares method revealed no statistically significant deviation from parallelism. By interpolating the number of HEL-pulsed Md~ that induces the same proliferation, the relative efficiency was calculated and found to be 8-fold greater for Md~ from MELtreated mice. These findings have been confirmed in 2 other similar experiments. IL-1 production Figure 2 shows the proliferation, expressed as stimulation index, of thymocytes from C 3 H / H e J mice as induced by P H A and several dilutions of M~ culture supernatants containing IL-1. The culture supernatants of Mdp from MEL-treated mice display greater activity than the ones from control mice at all dilutions tested. Analysis of variance of the linear regressions fitted by the least squares method revealed no statistically significant deviation from parallelism. Calculation of the relative efficiency of the supernatants from the 2 groups indicates that MEL treatment yielded a 2.2-fold increase in IL-I production by M~ cultured with LPS. A similar enhancement of IL-1 production by MEL was found in other 4 out of 5 experiments. If M~ are not
1
2
3
4
5
Reciprocal o! supernatanl dilution ( log 2 )
Fig. 2. Effect of MEL on IL-1 production. Several dilutions of the supernatants of cultured Md~ from MEL-treated (closed circles; maximal proliferation: 3333 counts/min) or control (open circles; maximal proliferation: 1394 counts/ rain) mice were used to stimulate the proliferation of C3H/HeJ thymocytes. Proliferation is expressed as stimulation index: ratio between the counts/rain of each experimental sample (supernatant and PHA) and the counts/rain of the control sample (medium and PHA; 129 counts/min). stimulated in vitro by LPS, IL-I is not detectable in both MEL-treated and control groups (data not shown).
TNF-a production The culture supernatants assayed for IL-I were also examined for the presence of TNF-a by ELISA. The results of 3 independent experiments are shown in Table 1 as TNF-a concentration in the supernatants. The reported values are lower than those usually found when the assay is performed on peritoneal M~ obtained from mice injected with thioglycolate (Gifford & Lohman-Matthes, 1987). It should be pointed out that the TNF-a concentrations found in the present study are not sufficient to affect thymocyte proliferation in the biological assay used to detect IL-I (Ranges, Zlotnik, Espevik, Dinarello, Cerami & Palladino, 1988). MEL treatment produced a variable effect increasing TNF-a production by 1 . 2 - 4 . 8 fold. Expression of MHC class H molecules Mdp from MEL-treated or control mice were analyzed for the expression of MHC class II molecules by indirect immunofluorescence. M+ were labeled with anti-I-A and anti-I-E biotin-conjugated mAbs which were revealed by phycoerythrin-conjugated streptavidin. Cell number versus log fluorescence intensity is shown in Fig. 3. Md~ incubated with
466
C. PIOLI et al. Table 1. Effect of MEL on TNF-a production by splenic macrophages Treatment
TNFa (pg/ml _+S.E.)
1
saline MEL
101 + 3 121 _+3
1.2
2
saline MEL
61 _+2 292 _ 9
4.8
3
saline MEL
55 _+2 136 ___4
2.5
Experiment
A
MEL/Saline
B
.,'.,
m
/,
! \
"6 /
l
i
z
Log fluorescence intensity
Fig. 3. Effect of MEL on MHC class II molecule expression. M+ from control (panel A) or MEL-treated (panel B) mice were labeled with anti-I-A and anti-I-E biotin-conjugated mAbs which were revealed by phycoerythrin-conjugated streptavidin. Background staining was measured on Md~ incubated with phycoerythrin-conjugated streptavidin alone (dotted line). The fractions of Ia + cells were 7°70in M+ from control mice and 21% in M+ from treated mice. phycoerythrin-conjugated streptavidin alone represent the blank control. The fraction of Ia + cells is 7o7o in Md? from control mice (panel A) whereas it is 21o7o (panel B) in Mdp from MEL-treated mice. Similar results have been obtained in 2 other experiments.
DISCUSSION Our results show that MEL treatment is able to increase antigen presentation by splenic M~. There are two critical events in antigen presentation, namely the expression of MHC class II molecules, which are essential for protein antigen recognition by T-cells, and the synthesis of IL-1 and other ancillary molecules which serve as a second signal for T-cell proliferation (Unanue & Allen, 1987). To clarify which mechanism of antigen presentation is influenced by MEL, we have studied cytokine production and MHC class I! molecule expression. We found that M~ from MEL-treated mice are able to produce more IL-I (2.2-fold) and TNF-a (1.2 -4.8-fold) than do M~ from control mice. Moreover, M~ from
MEL-treated mice express more MHC class II molecules as compared to Mdp from control mice. It is clear from several studies that IL-I is a necessary co-stimulator for T-cell growth (Gery, Gershon & Waksman, 1972; Durum, Schmidt & Oppenheim, 1986; Weaver & Unanue, 1990) and may act by increasing the expression of IL-2 receptor on T-cells (Mannel, Mizel, Diamanstein & Falk, 1985). Also TNF-a stimulates T-cells, by increasing the proliferative response to antigen and the expression of IL-2 receptor and by inducing IFN-y production (Scheurich, Thoma, Ucer & Pfizenmaier, 1987; Yokota, Geppert & Lipsky, 1988). rTNF-a treatment, likewise the endogenously produced T N F - a , is able to increase antigen presentation in human monocytes (Zembala et al., 1990). The effects of IL-1 and T N F - a on antigen presentation could also be mediated, as suggested by others (Dustin, Rothlein, Bhan, Dinarello & Springer, 1986; Pober, Gimbone, Lapiere & Mendrick, 1986) through induction of adhesion molecules, such as ICAM-1, or other cytokines, such as IL-6. Mdp harvested from lymphoid tissues or from the peritoneal cavity do not constitutively express IL-1 protein or messenger RNA (Kurt-Jones, Belier,
Melatonin Increases Antigen Presentation Mizel & Unanue, 1985; Fuhlbrigge, Chaplin, Kiely & Unanue, 1987) and, likewise, resting B-cells are IL-1 negative (Hawrylowicz, Duncan, Fuhlbrigge & Unanue, 1989). Also the expression of MHC class II molecules on the cell surface is influenced by the immune status of the individual (Durum & Oppenheim, 1989). Thus, endogenous or exogenous stimuli are necessary to induce both IL-1 and TNF-a production as well as MHC class II expression. In our experiments we have found that MEL treatment can increase the expression of MHC class II molecules on the M~ surface. Conversely, MEL treatment alone is not sufficient to induce IL-1 production but it renders M~, more susceptible to LPS stimulation. Thus, it appears that MEL treatment is able to induce in M+ an increase in specific (antigen-MHC class II complex) and nonspecific (IL-1 and TNF-a) signals that trigger T-cell proliferation. These findings provide further evidence for the role of the pineal gland in the control of the immune response. The improvement of the accessory function of M+, as induced by MEL, may indeed modulate the antibody response. It has been shown that MEL treatment increases the antibody production in mice immunodepressed by corticosterone treatment (Maestroni, Conti & Pierpaoli, 1988). Our preceding results have shown that MEL increases the antibody production in normal young mice or in mice immunodepressed by CY treatment or aging (Caroleo et al., 1992). The increase in antibody response could result from Md~ activation and IL-I production leading to stimulation not only of T-cells but also of B-cells. Under IL-I stimulation pre-B-cells express complete Ig on the plasma membrane (Giri, Kincade & Mizel, 1984) and, subsequently, during antigen activation mature
467
B-cells increase proliferation and Ig production (Hoffman, Gilbert, Hirst & Scheid, 1987). Moreover, indirect effects of IL-I on B-cells may also occur through the mediation of secondary cytokines: IL-1 induces several cell types to produce IL-6, which promotes Ig secretion by activated B-cells. IL-1 also augments the production of other cytokines by T-helper cells, including IL-2, IL-4, and IL-5 which control various stages of B-cell activation and function. Moreover, our preliminary results show that B-cell mitotic response to LPS in vitro is increased by MEL treatment. In preliminary experiments we have studied the in vitro effect of MEL on antigen presentation and IL-I production by splenic M+ but under our experimental conditions we did not find any effect (data not shown). Although a direct effect of MEL on M+ cannot be excluded, it is conceivable that MEL may act indirectly by inducing the secretion of soluble factors activating M+. One of these factors may be IFN-y which induces the expression of MHC class II molecules on the M+ surface (Kelso, Glasebrook, Kanagawa & Brummer, 1982). Moreover, in vitro and in vivo M~ priming by IFN-y increases the capacity of LPS to induce TNF-a production in vitro (Gifford et al., 1987). Experiments are in progress to test this possibility. In conclusion, our data show that MEL affects antigen presentation by increasing the expression of MHC class II molecules and cytokine (IL-1 and TNF-a) production, suggesting relevant mechanisms that may account for the role of the pineal gland in immunoregulation.
Acknowledgement - - C. Pioli is a recipient of an Italian Association for Cancer Research (AIRC) fellowship.
REFERENCES
AXELROD, J., FRASCHINI,F. & VELO, G. P. (eds) (1982). The pineal gland and its neuroendocrine role. Proc. Nato Adv. Study, Erice, Italy. Plenum, New York. CAROLEO, M. C., FRASCA,D., NISTICO',G. & DORIA, G. (1992). Melatonin as immunomodulator in immunodeficient mice. Immunopharmacology, 23, 8 1 - 89. CORRADIN,G., ETLINGER,H. M. & CHILLER,J. M. (1977). Lymphocyte specificity to protein antigens. I. Characterization of the antigen-induced in vitro cell-dependent proliferative response with lymph node cells from primed mice. J. Immun., 119, 1048-1053. DURUM, S. K. & OPPENHE1M) J. J. (1989). Macrophage-derived mediators. In Fundamental Immunology (ed. Paul, W. E.), pp. 639-661. Raven Press, New York. DURUM, S. K., SCHMIDT, J. A. & OPPENHEIM, J. J. (1986). Interleukin 1: an immunological perspective. Ann. Rev. Immun., 3, 263-287.
468
C. PIOLI et al.
DUSTIN, M. L., ROTHLEIN, R., BHAN, A. K., DINARELLO, C. A. & SPRINGER, T. A. (1986). Induction by IL-l and interferon-gamma, tissue distribution, biochemistry, and function of a natural adherence molecule (ICAM-I). J. Immun., 137, 245 - 250. FUHLBRIGGE, R. C., CHAPLIN, D. D., KIELY, J. M. & UNANUE, E. R. (1987). Regulation of interleukin 1 gene expression by adherence and lipopolysaccharide. J. Immun., 138, 3799-3802. GERY, I., GERSHON, R. K. & WAKSMAN,B. H. (1972). Potentiation of the T lymphocyte response to mitogens. I. The responding cell. J. exp. Med., 136, 128- 138. GIFFORD, G. E. & LOHMAN-MATTHES, M. L. (1987). Gamma interferon priming of mouse and human macrophages for induction of tumor necrosis factor production by bacterial lipopolysaccharide. JNCI, 78, 121 - 124. GIRl, J. G., KINCADE, P. W. & MIZEL, S. B. (1984). Interleukin 1-mediated induction of kappa-light chain synthesis and surface immunoglobulin expression on pre-B cells. J. Immun., 132, 2 2 3 - 228. HAWRYLOWlCZ, C. M., DUNCAN, L. M., FUHLBRIGGE,R. C. & UNANUE, E. R. (1989). Regulation of antigen presentation. II. Anti-Ig and IL-2 induces IL-1 production by murine splenic B cells. J. Irnrnun., 142, 3361 -3368. HOFFMAN, M. K., GILBERT, K. M., HIRST, J. A. & SCHE1D, M. (1987). An essential role for interleukin 1 and a dual function for interleukin 2 in the immune response of murine B lymphocytes to sheep erythrocytes. J. Molec. Cell. Immun., 3, 2 9 - 33. KELSO, A., GLASEBROOK, A. L., KANAGAWA,O. & BRUMMER, K. T. (1982). Production of macrophage-activating factor by T lymphocyte clones and correlation with other lyrnphokine activities. J. Immun., 129, 5 5 0 - 556. KURT-JONES, E. A., BELLER, D. I., MIZEL, S. B. & UNANUE, E. R. (1985). Identification of a membrane-associated interleukin 1 in macrophages. Proc. natn. Acad. Sci. U.S.A., 82, 1204- 1208. LEWy, A. J. (1983). Biochemistry and regulation of mammalian melatonin production. In The Pineal Gland, pp. 77 - 121. Elsevier, New York. LISSONI, P., MORELLI, O., MAURI, R., RESENTIN1, M., FRANCO, P., ESPOSTI, D., ESPOSTI, G., FRASCHINI, F., HELBERG, F., SOTHERN, R. B. & CORNELISSEN,G. (1986). Ultradianal chronomodulation by melatonin of a placebo effect upon human killer cell activity. Chronobiologia, 13, 3 3 9 - 343. MAESTRONI, G. J. M., CONTI, A. & PIERPAOLI, W. (1986). Role of the pineal gland in immunity. Circadian synthesis and release of melatonin modulates the antibody response and antagonizes the immunosuppressive effect of corticosterone. J. Neuroimmun., 13, 19-30. MAESTRONI, G. J. M., CONTI, A. & PIERPAOLI, W. (1987). Role of the pineal gland in immunity. II. Melatonin enhances the antibody response via an opiatergic mechanism. Clin. exp. Immun., 68, 384-391. MAESTRONI, G. J. M., CONTI, A. & PIERPAOLI, W. (1988). Pineal melatonin and its fundamental imrnunoregulatory role in aging and cancer. Ann. N.Y. Acad. Sci., 521, 140- 148. MANNEL, D. N., MIZEL, S. B., DIAMANSTEIN, T. & FALK, W. (1985). Induction of interleukin 2 responsiveness in thymocytes by synergistic action of interleukin 1 and interleukin 2. J. Immun., 134, 3108-3110. MOLLER, G. (1987). Antigenic requirements for activation of MHC restricted responses. Immun. Rev., 98, 1 - 187. POBER, J. S., GIMBONE, M. A~ JR, LAPIERE, L. A. & MENDRICK, M. (1986). Overlapping patterns of activation of human endothelial cells by interleukin 1, tumor necrosis factor and immune interferon. J. Immun., 137, 1893- 1898. PRESLOCK, J. P. (1984). The pineal gland: basic and clinical correlations. Endocrin. Rev., 5, 282-297. RADOSEVIC-STASIC, B., JONIC, S. & RUKAVINA, D. (1983). Immune response of rats after pharmacologic pinealectomy. Period. Biol., 85, 119- 121. RANGES, G. E., ZLOTNIK, A., ESPEV1K, T., DINARELLO, C. A., CERAM1, A. & PALLADINO, M. A. (1988). Tumor necrosis factor-alpha/cachectin is a growth factor for thymocytes. J. exp. Med., 167, 1472- 1478, SCHEURICH, P., THOMA, B., UCER, U. & PFIZENMAIER, K. (1987). Immunoregulatory activity of recombinant human tumor necrosis factor (TNF-alpha): induction of TNF receptors on human T cells and TNF-alpha-mediated enhancement of T cell responses. J. Irnmun., 138, 1786- 1790. SCHWARTZ, R . H . (1985). T-lymphocyte recognition of antigen in association with gene products of the major histocompatibility complex. Ann. Rev. Irnmun., 3, 2 3 7 - 261. UNANUE, E. R. & ALLEN,P. M. (1987). The basis for immunoregulatory role of macrophages and other accessory cells. Science, 236, 551 - 557. WEAVER, C. Z. & UNANUE,E. R. (1990). The costimulatory function of antigen-presenting cells. Immun. Today, 11, 49 - 55. WURTMAN, R. J., AXELROD, J. & PHILLIPS, L. S. (1963). Melatonin synthesis in the pineal gland: control by light. Science, 148, 1609- 1613. YOKOTA, S., GEPPERT, T. D. & LIPSKY, P. E. (1988). Enhancement of antigen- and mitogen-induced human T lymphocyte proliferation by tumor necrosis factor alpha. J. Immun., 140, 5 3 1 - 536. ZEMBALA, M., KOWALCZYK, D. E., PRYJMA, J., RUGGIERO, I., MYTOR, B., KLVSIK, J. & STEC, W. J. (1990). The role of tumour necrosis factor in the regulation of antigen presentation by human monocytes. Int. Immun., 2, 3 3 7 - 342.
