J Pineal Res 1997; 22:75-80 Printed in the United States of America---all rights reserved.

Cupyrighi 0 Munksgoard. 1997 Journal of Pineal Research

ISSN 0742-3098

Effect of melatonin and corticosteroid on in vitro cellular immune function in humans

Rogers N, van den Heuvel C, Dawson D. Effect of melatonin and corticosteroid on in vitro cellular immune function in humans. J. Pineal Res. 1997; 22:75-80.0 Munksgaard, Copenhagen.

Abstract: It is well accepted that the immune system shows circadian rhythmicity and that circadian disruption can significantly alter indices of immune function. Recently, a functional link between the endocrine and immune systems has been proposed to explain circadian rhythms in immune activity. Of particular interest is the finding that hormones such as melatonin and corticosteroid are able to exert modulating effects on lymphocyte proliferation. Previous research examining the effects of melatonin in vitro, however, has produced equivocal results. The aim of this study, therefore, was to examine the effects of melatonin and corticosteroid, both separately and together, on mitogen-stimulated human lymphocyte proliferation. Purified human lymphocytes were stimulated with concanavalin A (Con A, 4 pg/mL). Melatonin and/or corticosteroid were added to the culture medium during incubation. All cultures were done in quadruplicate. As expected, corticosteroid (25-1,000 ng/mL) significantly reduced proliferation by between 30 and 60% in a dose-related manner (P<O.OOOl). Melatonin alone (1 0-1,000 fmol/rnL) did not significantly affect lymphocyte proliferation. However, when lymphocytes were cultured in the presence of melatonin and corticosteroid, a significant decrease in proliferative function of 50435% was observed (P<O.OOOl). Hence, the effect of melatonin and corticosteroid combined was significantly greater than that observed with corticosteroid alone (P<O.OOOl). Therefore, it appears that the in vitro effect of corticosteroid on immune function may be modulated by melatonin in physiological to pharmacological concentrations.

Naomi Rogers,' Cameron van den Heuvel,2 and Drew Dawson' 'The Centre For Sleep Research, The Faculty of Humanities and Social Sciences, The University of South Australia, The Queen Elizabeth Hospital; 'Department of Obstetrics and Gynaecology, University of Adelaide, The Queen Elizabeth Hospital, Woodville SA 501 1, Australia

Key words: Melatonin - corticosteroid - immune modulation

Address reprint requests to Drew Dawson, The Centre For Sleep Research, The Queen Elizabeth Hospital, Woodville Rd, Woodville SA 501 1, Australia.

Received July 26; accepted December 11, 1996

Introduction

Circadian variations have been reported in a wide range of physiological parameters [Moore-Ede et al., 19821, including several functional indices of the immune system [Carandente et al., 19881. It is now well accepted that the immune system shows sig- nificant circadian rhythmicity and that disruption to the timing of the circadian system can alter indices of immune function [Nakano et al., 1982; Ohkoshi et al., 19911. In addition, a functional link between the neuroendocrine and immune systems has been proposed [Blalock and Smith, 1985; Blalock, 1989; Goetzl and Sreedharan, 1992; Homo-Delarche and Dardenne, 19931. Of particular interest is the find- ing that the hormones melatonin and corticosteroid are able to exert modulating effects on lymphocyte

proliferation [Gala, 19911. This finding suggests a possible role for these hormones in mediating cir- cadian rhythms in immune function.

Several studies have reported corticosteroid to possess potent immunosuppressive properties, both in vitro and in vivo. Lymphocyte proliferative re- sponses to stimulation with the mitogens concanava- lin A, phytohaemagglutinin, and pokeweed mitogen are reported to be significantly depressed, follow- ing the administration of either in vivo or in vitro corticosteroid [Heilman, 1972; Cupps and Fauci, 19821. In addition, responses to antigenic stimula- tion have also been found to be suppressed follow- ing in vivo administration of corticosteroid to a variety of species, including humans [Fauci and Dale, 1974; Ilfeld et al., 1977; Katz and Fauci, 1979; MacDermott and Stacey, 19811.

75

Rogers et al.

In contrast to the immunosuppressive actions of corticosteroid, melatonin is reported to demonstrate enhancing properties on immune responses in vivo. It has been reported that when the normal melatonin secretion in hamsters was suppressed by administration of a P-adrenergic antagonist, immune activity was also depressed [Maestroni and Pierpaoli, 198 I]. This provided indirect evidence of an immunoenhancing role for melatonin. Further evidence of an immunoenhancing effect of melato- nin was reported with the findings that evening ad- ministration of P-adrenergic antagonists produced a decrease in immune reactivity, which was reversed by evening supplementation of melatonin in ham- sters [Maestroni et al., 1986, 1987al. Subsequent studies, in a variety of species, have reinforced these in vivo immunoenhancing properties of melatonin [Maestroni et al., 1987b, 1988a; Becker et al., 1988; Champney and McMurray, 199 11.

Despite the many studies carried out to examine the effects of melatonin in vivo, there have been few studies examining the in vitro effects of this hor- mone. It has been observed that the immunoenhancing effect of melatonin is aimed primarily at immune re- sponses involving T lymphocytes [Maestroni et al., 1987a,b, 1988a,b, 19891. Hence, a purified lympho- cyte proliferation assay, which produces a general in vitro measure of immune status, has been utilized to examine the in vitro effects of melatonin and corticosteroid on immune function [Eskola et al., 1976; Kaplan et a]., 1976; Cove-Smith et al., 1978; Nakano et al., 1982; Carandente et al., 19881.

The aim of this study was to examine the in vitro immunomodulating effects of melatonin and corti- costeroid, both alone and administered concurrently, using a standard measure of lymphocyte prolifera- tive activity.

Materials and methods

An in vitro correlate of cell-mediated immune func- tion was measured using a lymphocyte proliferation assay based on that described by Nakano et al. [1982].

The lectin mitogen Concanavalin A (Con A, Sigma Chemical Company, St. Louis, USA) was used to stimulate lymphocyte proliferation and was reconstituted and diluted in Rosewell Park Memorial Institute 1640 medium (RPMI-1640, Sigma Chemical Company, St. Louis, USA), and stored at 4°C.

The melatonin used in this study was purchased from Sigma Chemical Company, St. Louis, USA. The melatonin was dissolved in ethanol and diluted in RPMI-1640 medium to a standard stock solution that is stable for 6 months, stored at 4°C in the dark. Further dilutions of this stock solution were pre- pared every 2 weeks.

76

The corticosteroid (hydrocortisone) used in this study was purchased from Sigma Chemical Com- pany, St. Louis, USA and was diluted in RPMI-1640 medium and stored at 4°C.

Human blood samples from four healthy male volunteers, aged between 22 and 23 years old (mean f 0.29), were collected between 1200 and 1300 hr in a syringe containing heparin (1,000 U, Delta West Ltd., Bentley, Australia) and separated on Ficoll-Paque gradient (Phannacia Biotech, Uppsala, Sweden) by centrifugation for 30 min at 1900 rpm. The top se- rum layer was discarded and the “buffy coat” layer, containing peripheral blood lymphocytes, was washed in 10 ml of Medium-199 (M-199, Gibco, Gaitherburg, USA). The cells were centrifuged for 30 min at 1900 rpm, resuspended in 10 mL M-199, then washed twice in 10 mL M-199, and resus- pended in 10 mL M-199. Lymphocytes were counted by a Coulter Counter (Coulter Electronics Ltd, Harpenden Herts, UK), and resuspended in RPMI- 1640, supplemented with 5% fetal bovine se- rum (5% v/v, CSL Pty Ltd, Melbourne, Australia) to a final lymphocyte concentration of 2 x lo5 cells per well, in a 96 well culture plate (Disposable Prod- ucts, Adelaide, South Australia).

Background measurements were obtained by the culture of 2 x lo5 cells in 200 mL RPMI-1640, in the absence of mitogen.

Con A and hormones were added to the plates prior to the addition of cells, to final volume of 200 mL. The plates were incubated for 72 hr in humidi- fied air and 5% COz, at 37°C. Six hours prior to har- vesting, individual wells were pulsed with 20 mL 0.25 mCi/mL [3H]-thymidine (specific activity 5 Ci/ mmol, Amersham International, Buckinghamshire, England). After 72 hr cells were harvested using a Skatron Instruments Cell Harvester (Leirbyen, Nor- way) onto filter paper which was left to dry over- night. Filter discs were transferred to scintillant tubes and 2 mL scintillant cocktail added. p ac- tivity of harvested cells bound to filter paper was measured using a p-scintillation counter (Wallac 1409 Scintillation Counter, Wallac Oy, Turku, Fin- land). Data were expressed as individual mean dis- integrations per minute (dpm) divided by mean background dpm to give stimulation indices. Stimulation indices were then converted to a per- centage response, relative to the response of pu- rified lymphocytes to mitogen alone, which was defined as 100 percent.

Differences in the relative response with the vari- ous concentrations of melatonin and corticosteroid were compared by way of a two-factor ANOVA and Fisher’s PLSD post-hoc comparison. Differences in the response of purified lymphocyte culture with low and high concentrations of corticosteroid and

Melatonin, corticosteroid and immune function

120 -

100 -

80- 8

2 a 60- Q

c C al

al a 40-

20 -

0 -

varying concentrations of melatonin were compared by way of ANOVA and planned comparisons.

120 -

100 -

I 0 0 - C 0 Q $ 6 0 - U

i i 40- 2

c

Q,

a 2 0 -

0 -

Results

As can be seen in Figure 1, addition of corticoster- oid to the culture produced a reduction in the pro- liferative response of lymphocytes to stimulation with Con A. The addition of 25 ng/mL of corticos- teroid produced a decrease of 27 f 2%, compared to when no corticosteroid was present (P<O.OOOl). The addition of 250 ng/mL produced a suppression of 62 & 1% in proliferative response compared to the response in the absence of corticosteroid (RO.01). When 1000 ng/mL of corticosteroid was added to the culture, no significant difference in sup- pression of response was found compared to that found with 250 ng/mL.

Addition of melatonin to the culture at concen- trations of 10, 250, and 1,000 fmol/mL, produced no significant difference in the proliferative response of the lymphocytes compared to cultures performed in the absence of melatonin, as seen in Figure 2.

As can be seen in Figure 3, the addition of both melatonin and corticosteroid to the culture produced a significant reduction in response, at both low and high concentrations of corticosteroid (P<0.0001). Not only was the response decreased compared to culture without the presence of either hormone, but the addition of both hormones produced a reduction in response greater than that seen with corticoster- oid alone (P<0.05). At a corticosteroid concentra-

I

0 25 250 1000 2000

Corticosteroid Concentration (ng/mL)

Fig. I . Effects of administration of corticosteroid on mito- gen (Con A)-induced lymphocyte proliferation. Corticosteroid concentrations are in ng/mL. Each column represents n = 4 f SEM. Lymphocyte proliferation was measured in dpm, con- verted to a stimulation index, then to a percent response, where response to Con A alone equals 100% response.

-

I

10 L 250

1000 2000

Melatonin Concentration (fmollmL)

Fig. 2 . Effects of administration of melatonin on mitogen (Con A)-induced lymphocyte proliferation. Melatonin concen- trations are in fmol/mL. Each column represents n = 4 f SEM. Lymphocyte proliferation was measured in dpm, converted to a stimulation index, then to a percent response, where response to Con A alone equals 100% response.

tion of 25ng/mL the proliferative response was sup- pressed by 27 & 2% of that without corticosteroid (P<O.OOOl). The addition of 10 fmol/mL melatonin further suppressed this response to 51 f 4% of the mitogen only culture, with 250 fmol/mL and 1,000 fmol/mL suppressing the response further to 62 +_

Corticosteroid Concentration IZ0 7

250nglmL

25ng/m~

0 10 250 1 wo

Melatonin Concentration (fmol/mL)

Fig . 3 . Effects of administration of melatonin and corticos- teroid on mitogen (Con A)-induced lymphocyte proliferation. Melatonin concentrations are in fmol/ml. Corticosteroid con- centrations are in ng/mL. Each column represents n = 4 f SEM. Two concentrations of corticosteroid were administered: 25 ng/ mL (low) and 250 ng/mL (high). Three concentrations of me- latonin were administered: 10 fmol/mL (low); 250 fmol/mL (medium) and 1000 fmol/mL (high). Lymphocyte proliferation was measured in dpm, converted to a stimulation index, then to a percent response, where response to Con A alone equals 100% response. (Error bars have not been omitted from this figure, but are of small magnitude.)

77

Rogers et al.

0.4% and 62 f 1% respectively (P<O.OOOl). When 250 ng/mL of corticosteroid was added, the prolif- erative response was suppressed by 62 k 1% of the response without corticosteroid. This response was reduced to 82 f 1%; 85 f 196, and 77 f 1% of the mitogen only response with the addition of the high, medium and low concentrations of melatonin, re- spectively (RO.0001).

Discussion

In the present study the effects of corticosteroid and melatonin on human in vitro cellular immune func- tion were examined. The addition of corticosteroid to a purified human lymphocyte assay produced a reduction in proliferative response to mitogenic stimulation. Although the addition of melatonin to this culture produced no observable change in the proliferative response of lymphocytes, when the two hormones were added together, a suppression in re- sponse was evident. Of particular interest is the ob- servation that the suppression observed with melatonin and corticosteroid together is greater than that observed with corticosteroid alone.

The immunosuppressive effect of corticosteroid on mitogen-stimulated lymphocytes reported in the present study is consistent with previous findings examining the effect of corticosteroid on immune response. Corticosteroid’s immunosuppressive prop- erties have been reported in both in vitro [Heilman, 1972; Gordon and Nouri, 19811 and in vivo [Fauci and Dale, 1974, 1975; Cooper et al., 19771 studies. In the present study we administered both physiological and pharmacological doses of corticosteroid to the in vitro culture, prior to incubation, and observed that, for each concentration of corticosteroid, the immune response was suppressed by 30 to 60%.

In the present study, despite the fact that effects of corticosteroid were consistent with previous re- search, the effects of melatonin administration con- trast with some previous findings. In vivo studies examining melatonin administration have reported immunoenhancing effects [Maestroni et al., 1987b, 1988; Becker et al., 1988; Champney and McMurray, 19911, whereas in vitro studies have reported en- hancing [Fraschini et al., 1990; Guerrero and Re- iter, 19921, inhibitory [Persengiev and Kyurkchiev, 1993; Di and Paulesu, 19941, and no effect [Maestroni et al., 1987bl. One shortcoming in some of these previous studies, however, is that the time of blood collection is not reported. Hence, it is pos- sible that the differing responses of melatonin on in vitro immune reactivity may result from circadian variation in lymphocyte sensitivity to the reagents used.

The equivocal in vitro effects of melatonin re- ported here and in previous studies raises the ques-

78

tion of what pathways are involved in the enhanc- ing effects of melatonin on lymphocytes in vivo. It seems likely that the immunoenhancing effect of melatonin is not a direct action of the hormone on lymphocytes. Rather, it could be that melatonin acts on other immunoreactive cells, which in turn act on the lymphocytes, or assist in the immune response. Alternatively, it is possible that other factors must be present at the time of administration to assist melatonin in its actions on lymphocytes. These other factors may be various cytokines released prior to or during the immune response, other blood-borne factors, or other factors released from other cells in the blood, which are removed during the purifica- tion of lymphocytes for this culture.

It has previously been reported that the in vivo immunoenhancing properties of melatonin are stron- gest when the immune system is suppressed by pharmacological treatment with corticosteroids or by acute stress, which is known to increase corticos- teroid levels [Maestroni et al., 1986,1987b,1988b, 19891. Hence, it is of interest to investigate whether this effect of melatonin can be replicated using an in vitro model.

The addition of both hormones together produced a significant reduction in lymphocyte proliferative activity in response to Con A stimulation. It is of interest to note that the suppression of response of lymphocyte proliferation due to melatonin and cor- ticosteroid is significantly greater than when corti- costeroid is added alone. This finding suggests a synergistic effect of the two hormones in the sup- pression of in vitro lymphocyte immune response. Thus, one can speculate that the effect of melato- nin on immune function is dependent on various fac- tors present at the time of reaction. It would seem reasonable to suggest, however, that the effect of melatonin in the present study is due more to an ef- fect of melatonin on corticosteroid’s immunoreac- tivity, rather than being due to a direct action of melatonin on lymphocytes.

This in vitro reduction in response due to mela- tonin and corticosteroid may have implications for immune activity in vivo, when the two hormones are present in the blood at the same time. The ef- fect of melatonin in vivo has been reported to have time-dependent effects. When melatonin is admin- istered late in the afternoon, there is a greater immunoenhancing effect than when it is adminis- tered in the morning [Maestroni et al., 1987bl. Hence, it is possible that at times of the day when melatonin and corticosteroid are released concur- rently, a decrease in immune function may be evident.

Since the secretory patterns of corticosteroid and melatonin do not typically reach their peak at the same time of day in humans, significant amounts of

Melatonin, corticosteroid and immune function

the two hormones will not generally be released concurrently. However, in addition to the circadian release of corticosteroid, the hormone is typically released as part of the stress response. Therefore, it is possible that if someone is awake during the night, when melatonin reaches peak levels, and is faced with a stressful stimulus, corticosteroid may be se- creted in significant quantities.

The findings of the present study suggest a cau- tious approach in looking at previous studies exam- ining the in vitro effects of melatonin on immune function. It is possible that the equivocal results of in vitro melatonin administration may result from unknown corticosteroid levels in the blood at time of sampling. Alternatively, it may be caused by the absence of other blood-borne factors that are re- moved during purification. Hence, an alternative method of examining in vitro effects of hormone administration on immune function is required.

In conclusion, the findings of the present study demonstrate that the immunosuppressive effects of corticosteroid reach a maximum level of effect at a dose as low as 250bng/mL. In addition, it appears that melatonin has no immunoreactive properties in vitro when administered alone. However, when me- latonin is administered in the presence of corticos- teroid, the two hormones together clearly demonstrate immunoinhibitory effects, greater than seen with cor- ticosteroid alone.

Acknowledgments

This research was supported in part by National Health & Medical Research Council, WorkSafe Australia, and the Aus- tralian Government Employees’ Medical Research Fund.

Literature cited

BECKER, J., G. VEIT, R. HANDGRETlNGeT, A. ATTANASIA, G. BRUCHETT, I. TRENNER, D. NIETHAMMER, D. GUPTA (1988) Cir- cadian variations in the immunomodulatory role of the pineal gland. Neuroendocrinol. Lett. 10:65-80.

BLALOCK, J.E. (1989) A molecular basis for bidirectional com- munication between the immune and neuroendocrine systems. Physiol. Rev. 69:l-32.

BLALOCK, J. E., E. M. SMITH (1985) A complete regulatory loop between the immune and neuroendocrine systems. Fed Proc 44 (NO. 1 Pt. 1):108-111.

CARANDENTE F., A. ANGELI, A. DE VECCHI, F. DAMMACCO, F. HALBERG (1988) Multifrequency rhythms of immunological functions. Chronobiologica 15:7-23.

CHAMPNEY, T.H., D.N. MCMURRAY, (1991) Spleen morphology and lymphoproliferative activity in short photoperiod exposed hamsters, In: Role of Melatonin and Pineal Peptides in Neuroimmunomodulation. F. Fraschini and R.J. Reiter, eds. Plenum, New York, pp. 219-225.

COOPER, D.A., V. PETTS, E. LUCKHURST, R. PENNY (1977) The effect of acute and prolonged administration of prednisolone and ACTH on lymphocyte subpopulations. Clin. Exp. Immu- nol. 28:467-73.

COVE-Smith J.R., P. KABLER,R. POWNALL, M.S. KNAPP (1978) Circadian variation in an immune response in man. Brit. Med. J. 2~253-254.

CUPPS, T.R., AS. FAuCl (1982) Corticosteroid-mediated immu- noregulation in man. Immunol. Rev. 65: 133-155.

DI StEFANo A,, L. PAULESU (1994) Inhibitory effect of melato- nin on production of IFN gamma or TNF alpha in peripheral blood mononuclear cells of some blood donors. J. Pineal Res. 17:164-169.

ESKOLA J., H. FREY, G. MOLNAR, E. SOPPI (1976) Biological rhythm of cell mediated immunity in man. Clin. Exp. lmmu- nol. 26:253-257.

FAUCI, AS. , D.C. DALE (1974) The effect of in vivo hydrocor- tisone on subpopulations of human lymphocytes. J. Clin. In- vest. 53:240-246.

FAUCI, A S . , D.C. DALE (1975) Alternate-day prednisone therapy and human lymphocyte subpopulations. J. Clin. In- vest. 55:22-32.

SACERDOTE (1990) Melatonin action on immune responses. In: Advances in Pineal Research 4. R.J. Reiter, A. Lukaszyk, eds. John Libbey, London, pp 225-233.

GALA, R.R. (1991) Prolactin and growth hormone in the regu- lation of the immune system. Proc. SOC. Exp. Biol. Med. 198:5 13-527.

GOETZL, E.J. S.P. SREEDHARAN (1992) Mediators of communi- cation and adaptation in the neuroendocrine and immune sys- tems, FASEB J 6:2646-2652.

GORDON, D., A.M.E. NOURI (1981) Comparison of the inhibition by glucocorticosteroids and cyclosporin A of mitogen-stimulated human lymphocyte proliferation . Clin. Exp. Immunol.

GUERRERO, J. M., R.J. REITER (1992) A brief survey of pi- neal gland-immune system interrelationships. Endocr. Res. 18:91-113.

HEILMAN, D.H. (1972) Failure of hydrocortisone to inhibit blas- togenesis by pokeweed mitogen in human leukocyte cultures. Clin. Exp. Immunol. 11:393403.

HOMO-DELARCHE, F., M. DARDENNE ( 1993) The neuroendocrine- immune axis. Springer Seminars Immunopathol. 14: 221-238.

ILFELD, D.N., R. S. KRAKAUER, R.M. BLAESE (1977) Suppres- sion of the human autologous mixed lymphocyte reaction by physiologic concentrations of hydrocortisone. J. Immunol. 1 19:428-34.

FUNDENBERG, L.N. LECAM (1975) Circadian rhythm of lym- phocyte blastogenesis. J. Allergy Clin. Immunol. 58:180-189.

KATZ, P., A. S. FAUCI (1979) Autologous and allogeneic inter- cellular interactions: modulation by adherent cells, irradiation, and in vitro and in vivo corticosteroids. J. Immunol. 123:2270-2277.

MACDERMOTT, R. P., M. C. STACEY (1981) Further character- ization of the human autologous mixed leukocyte reaction (MLR). J. Immunol. 126:729-34.

trol of the hormonally mediated immune-response. In: Psy- choneuroimmunology. R. Ader, ed. Academic Press, New York, pp. 405425.

MAESTRONI, G.J.M., A. CONTI, W. PIERPAOLI (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. Neuroimmunol.

MAESTRO”, G.J.M., A. CONTI, W. PIERPAOLI (1987a) The pineal gland and the circadian opiatergic immunoregulatory role of melatonin. Ann. N.Y. Acad. Sci. 496:67-77.

MAESTRONI, G.J.M., A. CONTI, W. PIERPAOLI (1 987b) Role of the

FRASCHINI, F., F. SCAGLIONE, G. DEMARTINI, v. LUCINI, P.

44: 287-294.

KAPLAN, M.S., V.S. BYERS, A.S. LEVIN, D.F, GERMAN, H.H.

MAESTRO”, G.J.M., w. PIERPAOLI (1981) Pharmacological con-

13: 19-30.

79

Rogers et al.

pineal gland in immunity. 11. Melatonin enhances the antibody response via an opiatergic mechanism. Clin. Exp. Immunol. 68:384-391.

MAESTRO", G.J.M., A. CONTI, W. PIERPAOLI (1988a) Pineal me- latonin: Its fundamental immunoregulatory role in aging and cancer. Ann. N.Y. Acad. Sci. 521:140-148.

MAESTRO", G.J.M., A. CONTI, W. PIERPAOLI (1988b) Role of the pineal gland in immunity. 111. Melatonin antagonizes the im- munosuppressive effect of acute stress via an opiatergic mechanism. Immunology 63:465-469.

MAESTRONI, G.J.M., A. CONTI, W. PIERPAOLI (1989) Melato- nin, stress and the immune system. In: Pineal Research Reviews, Vol. 7. R.J. Reiter, ed. Alan R. Liss, New York, pp. 203-233.

MOORE-EDE M.C., F.M. SULZMAN, C.A. FULLER (1982) In: The Clocks That Time Us. Moore-Ede M.C., Sulzman F.M. and Fuller C.A., eds. Harvard University Press, Cambridge, Mas- sachusetts.

NAKANO Y., I. HARA, T. MIURA, H. AONO, N. MIYANO, K. MIYAJIMA, T. TABUCHI, H. KOSAKA (1982) The effect of shift work on cellular immune function. J. Hum. Ergol. Suppl:131-137.

OHKOSHI H., I. Asukata, N. TAJIMA, K. YAMAMOTO, M. SASAKI, M. HOKARI, T. SAKAI (1991) The influence of transmeridian flight on human circulating lymphocytes. Aviation, Space Environ. Med. 62 :14-18.

PERSENGIEV, S. P., S. KYURKCHIEV (1993) Selective effect of melatonin on the proliferation of lymphoid cells. Int. J. Bio- chem. 25 :44 1-444.

80