Influence of desmethylimipramine on natural killer cell activity

9

Psychiatry Research, 19, 9- I5 Elsevier

Influence Activity Andrew Norin Received 1986.

of Desmethylimipramine

H. Miller,

November

Gregory

M. Asnis,

on Natural

Herman

Killer

M. van Praag,

Cell

and Allen

J.

11, 1985; revised version received February 21, 1986; accepted March 6,

Abstract. Mounting evidence suggests that the central nervous system (CNS) and the immune system are extensively interconnected. One question that arises is whether there is cross-reactivity between psychotropic agents, which are active in the CNS, and immune system function. To explore this notion, we examined the in vitro effect of the tricyclic antidepressant, desmethylimipramine (DMI), on human

natural killer (NK) cell activity in seven separate experiments. At concentrations > 625 ng/ ml, DMI reliably inhibited NK activity. Preincubation of lymphocytes with DMI before assay did not increase the inhibitory effect. Furthermore, removal of the drug from preincubated cells immediately before assay completely eliminated the inhibitory effect. These results demonstrate that DMI reversibly inhibits NK activity at serum concentrations that are not uncommonly found in depressed patients receiving this medication. Key Words. Desmethylimipramine lymphocytes, depression, inhibition.

(DMI),

natural

killer (NK)

cell activity,

Several recent reviews describe multiple points of interdigitation between the central nervous system (CNS) and the immune system (Ader, 1981; Guillemin et al., 1985). Through a complex feedback network of shared hormones, peptides, and neurotransmitters, these two systems extensively communicate, providing the pathways for mutual influence and modulation. CNS effects on immunity have been reported. Dysphoric states and certain psychiatric syndromes have been associated with significant decreases in cell-mediated immune functions (Linn et al., 1982; Kronfol et al., 1983; Locke and Kraus, 1984; Schleifer et al., 1984). Receptors for neurotransmitters and endorphins/enkephalins occur on the lymphocyte surface (Wybran et al., 1980; Besedovsky et al., 1983) and autonomic nervous system innervation of lymphoid tissues has been described (Reilly et al., 1979; Williams and Felton, 1981). As for immunological influences on the CNS, lymphocytes reportedly secrete soluble

Andrew H. Miller, M.D., is Research Fellow; Gregory M. Asnis. M.D., is Associate Professor; and Herman M. van Praag. M.D.. Ph.D.. is Professor and Chairman, Department of Psychiatry, Albert Einstein College of Medicine/ Montefiore Medical Center. Allen J. Norin, Ph.D.. is Associate Professor, Departments of Microbiology & Immunoloky and Surgery, Albert Einstein College of ‘Medicine/ Montefiore Medical Center. Bronx, NY. (Reprint requests to Dr. A.H. Miller, Klau I Psychiatric Outpatient Department, Department of Psychiatry, Montefiore Medical Center, I 1I E. 210 St., Bronx, NY 10467. USA.)

0165-1781/86/$03.50 0 1986 Elsevier Science Publishers

B.V.

10

factors that stimulate cortisol secretion both through the hypothalamus (Besedovsky et al., 1985) and directly via the adrenal gland (lymphoid adrenal axis) (Smith et al., 1982). Thymic secretions (thymosin) also appear to activate the hypothalamicpituitary-adrenal axis (Hall et al., 1985). In addition, adrenocorticotropic hormone (ACTH) and fi-endorphin are concomitantly released by lymphocytes, providing evidence that these cells are capable of transcribing the pro-opiomelanocortin gene (Smith, 1985). Finally, stimulated lymphocytes produce presently uncharacterized factors that substantially alter norepinephrine activity in the hypothalamus (Besedovsky et al., 1983). In view of the shared communication between the immune system and the CNS, the question arises as to whether there is cross-reactivity between psychotropic agents, which are active in the CNS, and immune system function. If lymphocytes and nerve cells respond to similar neurotransmitters and share surface receptors for these molecules, then psychotropic agents are likely to have specific immunological effects. One area of particular interest is the potential interaction of tricyclic antidepressants (TCAs) with the immune system. Depression (especially major depressive disorder) may be associated with both lowered cellular immune function (Kronfol et al., 1983; Schleifer et al., 1984) and an increased incidence of morbidity and mortality (Shekelle et al., 1981; Whitlock and Suskind, 1979). Since TCAs are extensively used for the treatment of depression, the effect of these agents on immune functioning is an important consideration in treating depressed patients. Several studies have examined the influence of TCAs on immunity, and an inhibitory effect has generally been observed. To date, all investigations have assessed the action of these drugs on in vitro mitogen-induced lymphocyte proliferation. Inhibition is first detected at TCA concentrations 2 3.0 pg/ ml with proliferation being inhibited by 50% (inhibitory concentration 50 or IC,,,) at concentrations ranging from 4.1 pgi ml to 15.6 pg/ ml in four reports (Baker et al., 1977; Fu et al., 1977; Nahas et al., 1979; Audus and Gordon, 1982). A correlation was noted between length of exposure and amount of inhibition. In no instance, however, regardless of the exposure period, did inhibitory concentrations approach serum levels found in patients taking these drugs. One group suggested that the inhibitory effect may be related to the liposolubility of these agents (Nahas et al., 1979). A more recent hypothesis is that these drugs may be acting on TCA binding sites that have been detected on the lymphocyte membrane (Audus and Gordon, 1982; Zvolsky et al., 1985). A controlled study on depressed patients found that mitogen-induced lymphocyte proliferation was decreased following a successful course of various somatic treatments including TCAs (Albrecht et al., 1985). Since all patients recovered, it is unclear whether recovery per se, somatic treatment, or their interaction contributed to the findings. To avoid this potentially complicating interaction of both psychotropics and depression with immune function, we studied the in vitro effects of a TCA, desmethylimipramine (DMI), on natural killer (NK) cell activity on lymphocyte preparations obtained from normal subjects. As described above, previous studies have used in vitro mitogen stimulation to evaluate the effects of TCAs on immune function. We have examined NK cell activity because unlike the cells in the mitogen assays, which are activated in culture (in vitro), NK cells are an effector cell population that has been activated in vivo. Furthermore,

11 in vitro mitogen responses have no known in vivo counterpart, whereas NK cells are believed to have an important role in providing natural protection against neoplasms and viral

infections

(Herberman

and Ortaldo,

198 1).

Methods Cytotoxicity Assays. Cell-mediated cytotoxicity assays were performed by standard techniques (Ortaldo et al., 1977) using K562 target cells from a human myeloid cell line. Peripheral blood lymphocytes from five healthy volunteers were isolated on Ficoll-Hypaque cushions and cultured in RPM1 1640 medium with 15% fetal calf serum, 2 mm glutamine, 5 x IO-5 A4 2-mercaptoethanol, 100 IUjml penicillin, and 100 pg/ml streptomycin (RPM1 complete). Concentrations of DMI ranging from 156 to 12,500 ng/ml were added to separate lymphocyte mixtures. Target cells ( 1 to 2 x 106) were suspended in 0.3 ml of medium, 0.2 ml of Tris phosphate buffer @H 7.4) and 100-200 &i of sodium 5lCr chromate (Amersham Corp., Arlington Heights, IL). The target cells were then incubated for 30 minutes at 37’C, washed three times and resuspended in medium at 2.5 x 104 viable cells/ ml. One hundred ~1 of viable lymphocytes and DMI were added to 100 ~1 of 51Cr labeled K562 cells in 96 well V-bottom microtiter plates (Nunc, Roskilde, Denmark). The effector-to-target ratio was lOO:l, with effector cell concentrations being 2.5 x IOhcells/ml. After 3 hours at 37”C, 100 ~1 of supernatant from quadruplicate samples was harvested, and the 51Cr radioactivity was determined in a gamma spectrophotometer (Model 1185, Searle Analytic, Inc., Des Plaines, IL). In some experiments lymphocytes were preincubated with the various concentrations of DMI at 37’C for 3 or 24 hours. After preincubation, half of the cells at each DMI concentration were then washed three times to remove drug while the others (unwashed) were assayed for cytotoxicity in the presence of drug. The 70 cytotoxicity was calculated by the formula [(CPM ExperimentalCPM spontaneous) + (CPM Maximum - CPM spontaneous)] x 100. Percent inhibition was calculated by the formula [I - yo cytotoxicity at experimental drug concentration) + (Y_Y cytotoxicity of no-drug control)] x 100. To determine the point at which mean ‘$6inhibition exceeded 0. a two-tailed Student’s f test was used at each drug concentration.

Results To evaluate the effect of DMI on NK activity, peripheral blood lymphocytes were incubated with DMI at concentrations ranging from 156 to 10,000 ng/ ml. Fig. 1 shows a dose-response curve generated from seven separate experiments. Fifty percent inhibition (IC 50) occurred at the DMI concentration of 4,250 ng/ml(l.4 x 10m5M), and beginning at 625 ng/ ml the mean Yo inhibition was significantly greater than 0 (t = 3.75, df= 6, p < 0.01). To determine whether the inhibitory effect of DMI on NK activity is reversible, cells were incubated with DMI for 3 or 24 hours, washed extensively to remove the drug, and then assayed for cytotoxicity. Control cultures were incubated with the drug for identical time periods and not washed. Fig. 2 demonstrates that removal of the drug before assay completely eliminates the inhibitory effect whether cells are incubated for 3 or 24 hours. These experiments demonstrate that DMI inhibition of NK activity is reversible. Whether prolonged exposure to DMI enhances the inhibitory effect on NK activity was assessed by preincubating cells with DMI for 3 or 24 hours. These cells (unwashed) were then assayed for NK activity in the presence of drug. Preincubation of up to 24 hours has no significant effect on inhibition when compared to the standard inhibition dose-response curve on cells that were not preincubated (Table 1). We

12 Fig. 1. Effect of desmethylimipramine (DMI) on natural killer cell activity

Fig. 2. Reversible inhibition of natural killer (NK) cell activity in desmethylimipramine (DMI)

A unwashed, 3 hrs A washed, 3 hrs l

unwashed, 24 hrs

o washed, 24 hrs

2000

4000

6000

6000

10000

1000

DMl tqlmlf Lymphocytes and target cells were incubated in the presence of various concentrations of DMI. The mean % inhibition I? SD1 in = 7) is shown. Statistically significant inhibition +I < 0.01 1was observed at DMI doses 2 625 ng/ml

Table 1. Effect of preincubation natural killer (NK) cell activity1

3000 DMI hg/ml)

5000’

NKcelIswerepreincubated with DMIfor3or24hours and then washed to remove thedrug. Control cultures were preincubated with DMI for3 or 24 hoursand then assayed (unwashed 1for NKactivity in the presence of drug. Representative data are shown.

with desmethylimipramine

(DMI)

on

% Inhibition

DMI concentration

No preincubation

_

Preincubation

with DMI

(ng/mU

with DMI

3 hours

24 hours

10,000

90%

ND2

97%

5,000

55%

59%

50%

2,500

26%

32%

19%

1,250

13%

16%

13%

625

9%

2%

3%

312

4%

0

4%

156

3%

0

4%

0

0

0

0

1. Cells were prerncubated for 3 or 24 hours in the presence of DMI at 37” C. Control cells were similarly incubated for 3 or 24 hours without DMI All cells were then assayed for NK activity in the presence of DMI.ThecontrolculturesgavesimilarinhibitioncurvesascomparedtotheaverageDMIinhibitioncurve obtained in 7 experiments (Fig 1 I. For comparative purposes, these average values are shown in this table. 2. Not done.

13 conclude, therefore, that the inhibitory influence on NK activity by concentrations of DMI < 10,000 ng/ ml occurs only during the cytotoxicity assay (target cell lysis). No additional effect is observed if the NK cells are exposed to DMI before assay. Discussion The inhibitory effect of DMI on NK activity in vitro does not appear to result from permanent damage to cellular components since exposed cells resume normal function when DMI is removed. In addition, the effect is not enhanced by prolonged exposure (3-24 hours preincubation) to the drug. This observation further suggests that toxicity is not involved. Apparently, the inhibitory effect only occurs during target cell killing, suggesting that DMI has a specific effect on the lytic process. There are several steps in this process where DMI might be exerting an influence: (1) inhibition of effector and target cell binding, (2) activation of the effector cell, (3) exocytosis of effector cell lytic granules, or (4) interaction of lytic granules with the target cell. Investigations on the effect of DMI on these processes are underway. Note that the inhibitory effect of DMI occurred at concentrations not uncommonly seen in depressed patients treated with this TCA. For example, in one treatment study, steady-state plasma concentrations in 23 depressed patients treated with DMI (300 mg/day) ranged from 48 to 877 ng/ml (Stewart et al., 1980). Three of these patients had levels of 557, 713, and 877 ng/ml. In vitro DMI inhibition is detected at 3 625 ng/ ml. Our data suggest that DMI treatment in most patients would have little effect on NK activity, but based on the data of Stewart et al. (1980), approximately 10% of patients receiving 300 mg of DMI per day may have blood concentrations sufficient to result in significant inhibition of NK activity. The full extent of this potential inhibition is difficult to estimate because peak plasma drug levels are not reported and could be l-2 times greater than steady-state levels (Gram et al., 1977). The active metabolite of DMI, 2-OH desipramine, is also not measured and may exert additional immunological effects (Javaid et al., 1979). Furthermore, depressed patients using TCAs such as DMI are usually treated for 6 months to a year or more. Although preincubation of lymphocytes with DMI from 3 to 24 hours does not enhance the inhibitory effect on NK cells, more prolonged exposure could have a greater effect. Tolerance to the immunological effects of the drug might develop in chronically treated patients. Finally, the healthy individuals we studied may differ from depressed patients in their immunological sensitivity to DMI. In fact, preliminary investigations comparing DMI binding sites on lymphocytes in depressives and controls suggest that such differences exist (Zvolsky et al., 1985). The in vitro model may not accurately reflect the inhibitory effect of DMI on in vivo NK activity. However, assessment of the in vivo effect of psychotropic agents (e.g., DMI) is problematic; in addition to direct interactions with the immune system, these psychotropic agents may alter hormonal and/ or neurotransmitter balances as well as mental state (Asnis et al., 1985). Such changes may in turn influence the immune response. Since the effect of DM I is reversible and cells are washed before assay, it should be possible to assess the influence of depression on NK activity in patients who are receiving DMI. Preliminary reports that depressed patients exhibit evidence of

14 lowered cell-mediated immune function underline the need to explore the potential immunological effects of DMI and other psychotropic agents. To date, the autonomic and cardiovascular side effects of TCA treatment have been the major considerations. Because the functional reserve of the immune system is poorly understood, the clinical repercussions of low level inhibition of N K cell activity by DM 1 remain unclear. In the future, it may be necessary to include the immune system in such considerations. Acknowledgments. We thank Christine Lackner, Rattan Cherwoo, and Rimma Mushnitsky for laboratory assistance, and Lorna Elphic and Joyce Sloman for manuscript preparation. This study was supported in part by a grant (MH-40903) from the National Institute of Mental Health.

References Ader, R., ed. Ps.whoimmunology. Academic Press, Orlando, FL (1981). Albrecht, J., Helderman, J. H., Schlesser. M.A., and Rush, A.J. A controlled study of cellular immune function in affective disorders before and during somatic therapy. Ps~~~Aiarr~~ Research,

15, 195 (1985).

Asnis, G.M., Halbreich, U., Rabinovich,

H., Ryan, N.D.. Sachar, E.J., Nelson, B., PuigAntich, J., and Novacenko, H. The cortisol response to desipramine in endogenous depressives and normal controls: Preliminary findings. P.yJ,chiutry Research, 14, 224 (1985). Audus. K.L.. and Gordon, M.A. Tricyclic antidepressant effects on the immune lymphocyte mitogen response. Journal qf’Immune Pharmacology, 4, 13 (1982). Baker, G.A., Santalo. R., and Blumenstein. J. Effect of psychotropic agents upon the blastogenic response of human T-lymphocytes. Biological Psychiatry,, 12, 159 (1977). Besedovsky. H.O., de1 Ray, A., and Sorkin, E. Neuroendocrine immune regulation. In: Fabris, N., Garaci. E., Hadden, J.. and Mitchison. N.A., eds. Immunoregulation. Plenum Press, London (1983). Besedovsky, H.. del Ray, A., Sorkin, E., Da Prada, M., Burri, R.. and Honegger, C. The immune response evokes changes in brain noradrenergic neurons. Science, 221, 564 (1983). Besedovsky. H.O., Sorkin, E., Keller, M., and Muller, J. Changes in blood hormone levels during the immune response. Proceedings of’ the Society’ of’ E.rperimenral Biology and Medicine, 150, 466 ( 1985). Fu, T.-K., Jarvik, L.F. Yen, F.-S.. and Matsuyama, S.S. In vitro effects of imipramine on proliferation of human leukocytes. Archives of Genera/ Ps.vchiarr.v, 34, 728 (1977). Gram, L.F., Sondergaard, I.B., Christiansen, J., Peterson, G.O.. Beth, P., Reisby, N., Ibsen, I., Ortmann, J., Nagy. A., Dencker. S., Jacobson, O., and Kautwald. 0. Steady-state kinetics of imipramine in patients. Psychopharmacology, 54, 255 (1977). Guillemin, R.. Cohn. M.. and Melnechuk. T., eds. Neural Modulation of Immunit~~. Raven Press, New York (1985). Hall, N.R.. McGillis, J.P., Spangelo, B.L., Healy. D.L., Chrousos. G.P., Schultz, H.M.. and Goldstein. A.L. Thymic hormone effects on the brain and neuroendocrine circuits. In: Guillemin, R., Cohn, M., and Melnechuk, T., eds. Neural Modularion qf’Immunit~~. Raven Press, New York (1985). Herberman. R.B.. and Ortaldo. J. R. Natural killer cells: Their role in defense against disease. Science, 214, 24 (1981). Javaid, J.R.. Perel, J.M., and Davis, J.M. Inhibition of biogenic amines uptake by imipramine, desipramine. 2-OH-imipramine. and 2-OH-desipramine in rat brain. L$e Sciences, 24, 21 (1979).

15

Kronfol, Z.. Silva, J., Greden. J.. Demibrimski, S., Gardner, R., and Carroll, B. Impaired lymphocyte function in depressive illness. L
Locke. activity. Nahas, cultured 344

Medicine,

44, I28 ( 1982).

S.E., and Kraus. L. Life change stress, psychiatric Psychosomnric

Medicine,

symptoms

and natural

killer cell

46, 441 (1984).

G.G.. Desoize, B.. and Leger. C. Effects of psychotropic lymphocytes. Proceedings of’the Society qf’Experimental

drugs in DNA synthesis Bio1og.v and Medicine,

in 160,

( 1979).

Ortaldo. J.R.. Oldham, R.K., Cannon, G.C.. and Herberman, R.B. Specificity of natural cytotoxic reactivity of normal human lymphocytes against a myeloid leukemia cell line. Journal qfthe

National

Cancer

Institute,

59, 77 (1977).

Reilly, F.D.. McCuskey, P.A.. Miller, M.L.. McCuskey, R.S.. and Meineke. H.A. Innervation of the periarterioiar lymphatic sheath ofthe spleen. Tissueand Cell, 11, 121(1979). Schleifer. S.J., Keller. S.E.. Meyerson. A., Raskin. M.. Davis, K., and Stein, M. Lymphocyte function in major depressive disorder. Archives of’ General Ps.pchiatry, 41,484 (1984). Shekelle, R.B.. Raynor, W.J.. Ostfeld, A.M., Garron, D.C., Bieliavskas, L.A., Lui, S.C., Maliza, C.. and Oglesby, P. Psychological depression and l7-year risk from cancer: P.~ychosomaric Medicine, 43, I I7 ( 198 I ). Smith, E.M., Harbour-McMenamin, D., and Blalock, J.E. Lymphocyte production of endorphins and endorphin-mediated immunoregulatory activity. Journal of Immunology, 135, 779s (1985).

Smith, E.M., Meyer, W.J.. and Blalock, J.E. Virus-induced increases in corticosterone in hypophysectomized mice: A possible lymphoid-adrenal axis. Science, 218, I31 I (1982). Stewart, J.W., Quitkin, F.. Fyer. A., Rifkin. A., McGrath. P., Liebowitz, M., Rosnick. L., and Klein, D.F. Efficacy of desipramine in endogenomorphically depressed patients. Journalqf A,frective Disorders, 2, I65 ( 1980). Whitlock. F.A., and Siskind. M. Depression and cancer: A follow-up study. Ps,,chologica/ Medicine,

9, 747 ( 1979).

Williams. J.M.. and Felten, D.L. Sympathetic innervation of murine thymus and spleen: A comparative histofluorescence study. Anatomical Record, 199, 53 I ( I98 I). Wybran, J., Appelboom, T., Famaey, J., and Govaerts. A. Receptors for morphine and methionine-enkephalin on human T lymphocytes: The two hits opioid lymphocyte receptor hypothesis. In: Serrou, B., and Rosenfeld, C., eds. New Trends in Human Immunology and Cancer Immunotherap~~. Doin Publishers, Paris (1980). Zvolsky. P.. Krulik, R., Silva, D., Sikora, J.. Farska. I.. and Fuxoua, L. The binding of imipramine and desmethyl-imipramine in the human lymphocyte. Abstracts qfrhe IVth World Congress qf Biological Psvchiatrv, Philadelphia, PA. p, 204 (1985).