BIOLOGICAL PSYCHOLOGY

ELSEVIER Biological Psychology 38 (1994) 143-155

Conditioned manipulation of natural killer (NK) cells in humans using a discriminative learning protocol

Angelika Buske-Kirschbaum al*, Clemens Kirschbaum a, Helmuth Stierle a, Lea Jabaij b, Dirk Hellhammer a

a Department of Psychology, lJniversi@ of Trier, D-54268 Trier2 Germany b Depart~nt of Clinical Irnrnu~lo~, ~~~ve~s~~ hospital U&e&t, Uirecht, Netherlands

Abstract

There is growing evidence indicating that the immune function can be modified by classical conditioning techniques. This phenomenon, initially explored in animals, is further documented by studies providing evidence that the human immune response can also be influenced by classical conditioning processes. In the present study, we tested the hypothesis that human immune parameters can be modulated by discriminative learning processes. Using a classical discriminative conditioning design, healthy volunteers were provided with a CS + (sherbet sweet/white noise), which was repeatedly paired with an injection of epinephrine (unconditioned stimulus, US). After epinephrine injections (0.2 mg subcuta- neously), a transient increase of natural killer (NK) cell activity (unconditioned response, UR) could be observed. A second stimulus complex (herbal sweet/auditory stimulus, conditioned stimulus, CS - 1 remained without reinforcement. After repeated presentation of the stimuli, re-exposure of the CS + on the test trial 1 resulted in a significantly increased number of NK positive (NK+) cells and in slightly elevated NK cell activity. No alteration of NK cells, however, could be observed after presentation of the CS - . A second re-exposure of the CS + on test trial 3, again resulted in a marked increase of NK’ cell number as well as in significantly elevated NK cell activity. The data presented here extend previous observations of conditioned alteration of immune responses in humans and indicate that the human organism might be able to react differentially to external stimuli, which have been associated with different immunological consequences.

Keywords: Conditioned immunomodulation; Natural killer cells; Discriminative learning; Epinephrine

* ~orres~nding author.

0301-0511/94/$07.00 0 1994 Elsevier Science B.V. AI1 rights reserved SSDI 0301-0511(94)00947-V

144 A. Buske-Kirschbaum et al. / Biological Psychology 38 (1994) 143-155

1. Introduction

Accumulating data indicate an interactive physiological defense network that involves the central nervous system (CNS), the endocrine system, and the immune system (Blalock, 1984; Besedovsky & de1 Rey, 1991). The capabilities of the CNS for active involvement in immunoregulation have been highlighted by demonstra- tions of classicahy ~nditioned aherations of immune function (Ader & Cohen, 1991). Classical conditioning of immune reaction is recognized as the ability of an initially neutral stimulus, termed the conditioned stimulus (CS), to trigger immuno- logic alterations by virtue of a previously learned association with an immunomod- ulating biologic or pharmacologic event, the so-called unconditioned stimulus (US). Using a taste aversion paradigm, Ader and Cohen (1975) were the first to evaluate conditioned immunosuppression. Now, classical ~nditioning protocols are commonly used to show conditioned suppression and conditioned enhance- ment of immune function. The generality of the phenomenon is further docu- mented by a number of studies, suggesting that classical conditioning can influence cell-mediated as we11 as humoral-mediated immune responses. Thus, conditioned depression of natural killer (NK) cell activity (O’Reilly & Exon, 1986), white blood cell counts (~osterhalfen & ~osterhalfen, 1987), l~ph~yte proliferation (Kusnecov, Husband & King, 1988; Neveu, Dantzer & Le Moal, 1988) or antibody responses (Rogers, Reich, Strom & Carpenter, 1976; Wayner, Flannery & Singer, 1978) could be demonstrated. In addition, conditioned increase of various immune parameters including NK cells (Ghanta, Hiramoto, Solvason & Spector, 1985) or cytotoxic T-cell precursors (Gorczynski, Macrae & Kennedy, 1982) have been reported. The potential biological and clinical impact of conditioned immunomod- ulation was demonstrated by several investigators showing that, based on condi- tioning principles, the onset and course of autoimmune disorders as murine lupus erythematodes (Ader & Cohen, 1982) or adjuvant arthritis (Klosterhalfen & Klosterhalfen, 1983) could be modified. Furthermore, conditioned increase of immune response has been used to influence tumor growth and survival rate in murine cancer (Ghanta, Hiramoto, Solvason & Spector, 1987).

Although conditioning of immune responses may be a useful model to under- stand the interplay between the CNS and the immune system, there is only a small body of data suggesting that also the human immune function can be manipulated by classical conditioning processes. Of note are two studies, suggesting modulation of immune function by psychological mediation in humans. Thus, Smith and McDaniels (1983) reported decreased delayed hypersensitive reaction to tuber- culin, when a paradigm was followed, where the subjects anticipated their reac- tions to be negative. This observation was supported by findings demonstrating that chemotherapy treated women showed anticipatory suppression of lymphocyte proliferation, when re-exposed to the clinical stimuli (Bovbjerg et al., 1990).

Following these lines of evidence our group examined conditioned immunomod- ulation in humans using a classical conditioning paradigm. Healthy subjects were conditioned by pairing a neutral sherbet sweet (CS) with a subcutaneous injection of epinephrine (US). Epinephrine injection has been shown to cause a marked

A. Buske-Kirschbaum et al. /Biological Psychology 38 (1994) 143-155 145

increase of the absolute number of NK cells as well as NK cell activity (NKCA; Toennessen, Toennesen & Christensen, 1984). After repeated pairing of the sherbet sweet with epinephrine injection, conditioned subjects showed significant increase of NKCA when re-exposed to the CS. No alteration of immune function, however, could be observed in control groups which have been given the sherbet sweet combined with saline injections (saline control) or with epinephrine, but in a non-paired manner (unpaired control) (Buske-Kirschbaum, Kirschbaum, Stierle, Lehnertz & Hellhammer, 1992; Kirschbaum et al., 1992). These data suggested that also in humans immune function can be manipulated by conditioning tech- niques.

In order to extend our previous results, the present study was designed in an effort to investigate classical conditioning of human immune function in a discrimi- native design. An important consideration to choose a discriminative learning protocol was to evaluate the possibility that (1) the human organism might be able to learn to differentiate between two stimuli of different immunological conse- quences and (2) to react with a learned immune response after re-exposure of the respective cue. In order to test these hypothesis, two conditioned stimuli were introduced. The stimulus paired with the US is referred to as CS + , and the stimulus presented without the US, that is alone, is referred to as CS - . It was expected that with continued training, only re-exposure of the CS + , but not the CS - would result in a conditioned response (Marx, 1971; Bolles, 1979).

2. Method

2.1. Subjects

Healthy volunteers (10 males; 10 females; PZ = 20) were recruited among under- graduate students at the University of Trier. They were invited to participate in a study, assessing the extent to which psychological factors might affect certain physiological components. They were asked to participate in a study involving blood draws and injections of a substance identical to an endogenous substance of the organism. Furthermore, they were told that they would be informed about the research program (e.g. goal of the study, injected agent) after the experiment. Their informed consent was obtained, and they were paid DM 350.- for participa- tion. Because of a suspected virus infection one subject was excluded from the experiment.

2.2. Procedure

A classical discrimination conditioning design was used to train the volunteers with each subject acting as his/ her own control. As in previous studies (Buske- Kirschbaum et al., 1992; Kirschbaum et al., 1992) two subjects per week partici- pated in the experiment. They were provided with ten acquisition trials, five trials

146 A. Buske-Kirschbaum et al. /Biological Psychology 38 (1994) 143-155

with the CS + and five trials with the CS - . The order of the presentation of the stimuli (CS + ; CS) was randomized and was identical for all subjects. According to previous reports of successfully conditioned NK cell function in animals, com- pound stimuli including different sensory modalities were used for this experiment (Dyck, Greenberg & Osachuk, 1986; Dyck, Driedger, Nemeth & Osachuk, 1987). In the CS + training, subjects were treated as follows. After entering the experi- mental room, subjects were informed about the experimental protocol. In order to allow blood draws, a small catheter was inserted into an anticubital vein and kept open by physiological saline. 30 min later, a first blood sample was drawn (baseline). After an additional 10 min, subjects were provided with a neutral sherbet sweet (‘Dynamit”, Frigeo, Germany) combined with white noise (CS + > for 10 s. Immediately after presentation of the CS + , the subjects were injected subcutaneously with 0.2 mg epinephrine (SuprareninR, Hoechst AG, Germany) which constituted the US. 20 min after stimulus presentation, a second blood sample was obtained and the experimental session was concluded.

In the CS - training subjects were treated identically, apart from the fact that subjects were exposed to a herbal sweet combined with a specific tone (CS - ) which remained without any subsequent treatment.

Conditioning test trials began after ten acquisition days. Test trials were identical to acquisition trials, except for the absence of the US. In the first and third test trial subjects were re-exposed to the CS + , in the second test trial they were reintroduced to the CS - . The first two conditioning test trials were designed to demonstrate conditioned alteration of NK cell function based on discrimination learning. The third test trial, however, was performed to assess possible extinction phenomena. The basic design of the treatment schedule is summarized in Table 1.

Table 1

Experimental protocol

Treatment

Acquisition trials

2

6

8 9

10 Test trials

11

12

13

cs- a cs- a+ b cs- cs- cs+ cs+ cs- cs+ cs+

cs+ cs- cs+

a CS - = herbal sweet/specific tone.

b CS + = sherbet sweet/white noise + 0.2 mg epinephrine (s.c.).

A. Burke-Kirschbaum et al. /Biological Psychology 38 (1994) 143-155 147

The experiment started at 9.00 a.m. and ended at 10.30 a.m. It should be noted that time of conditioning treatment as well as the testing conditions were kept identical for all experimental subjects.

2.3. Immunological assays

Two immunological measurements were performed: a functional assay (NKCA) and a quantitative assay (absolute number of NK+ cells). Both assays were of primary interest because NKCA as well as the absolute number of NK+ cells are known to be sensitive to behavioral influences (Irwin, Daniels, Bloom & Weiner, 1986; Kiecolt-Glaser et al., 1985; Ottaway & Husband, 1992). Additionally, recent studies have suggested that one mechanism to induce enhancement of NKCA appears to be mediated by elevated number of NK+ cells (Trinchieri, 1985). Thus, quantitative determination of NK cells was also performed to explore its possible contribution to alterations in the functional assay.

In order to assess the number of NK+ cells, fluorescence-activated cell sorting (FACS) analysis was performed on acquisition days 3 and 8 as well as on the conditioning test days 11, 12 and 13. For lymphocyte subset analysis, cells were stained using “Simultest” sets (Becton-Dickinson, Mountain View, USA), contain- ing directly FITC and PE conjugated monoclonal antibodies against CD16/CD56 (both NK cell markers). Cells (0.5 X 10h) were pelleted and incubated with the different monoclonal antibodies for 20 min on ice. Cells were washed twice using MEM/ps, supplemented with 0.5% bovine serum albumin and 0.1% sodium azide, after which they were analyzed using a flow cytometer (FACScan, Becton- Dickinson, USA).

Blood samples were assayed for NKCA on acquisition days 1, 3, 6 and 8 as well as on conditioning test days 11, 12 and 13. NKCA was measured using a standard “chromium release assay. This method has been recently described in detail (Kiecolt-Glaser et al., 1984). Briefly, 15 ml of heparinized venous blood was diluted with an equal volume of balanced salt solution (BSS) and layered onto 50 ml Ficoll-Hypaque (Pharmacia, Upsala, Sweden). Samples were centrifuged for 35 min at 400 g. Mononuclear cells were harvested from the interface, washed three times in BSS and subsequently counted under typran blue exclusion. 2 x lo6 target K-562 cells were suspended in 100 ~1 BSS and incubated for 1 h with 100 PCi Na$ZrO, at 37 “C and 5% CO,. After incubation, the labeled cells were washed three times in BSS and resuspended at a concentration of lo5 cells per ml in Dulbecco’s modified Eageles medium (DME). Mononuclear cells were co-cultured with labeled target cells at effector-target cell ratios (E: T) of 50: 1 and 25 : 1, respectively, for 6 h at 37 “C and 5% CO,. Thereafter, 100 ~1 of cell-free supernatant was harvested form each well and radioactivity was measured from the supernatant of K-562 cells incubated without lymphoid cells, and the maximum release was obtained after adding 1 M NaOH to lo4 targets. Cell lysis was calculated according to the following formula: % Cytotoxicity = (Experimental release - Spontaneous release/ Maximal release - Spontaneous release) x 100%.

148 A. Buske-Kirschbaum et al. /Biological Psychology 38 (1994) 143-155

3.3. Co&sol assay

There have been numerous reports indicating that a variety of stressors, including experience of venipuncture or donating blood, are associated with concomitant increases in adrenocortical activity and suppressed lymphocyte func- tion, e.g. NKCA (Girgis, Shea & Husband, 1988; Schlesinger & Yodfat, 1988). In order to control modulation of NKCA in response to altered concentration of cortisol, salivary cortisol was assessed on the conditioning test days 11, 12 and 13. Saliva samples were obtained 20 min and 10 min before, 10 min, and 20 min after stimulus presentation using the “alivette” sampling device (Sarstedt, Rommels- dorf, Germany). This device consists mainly of a small cotton swab on which the subject gently chews for 0.5-l min. Thereafter, the swab is transferred into a small plastic tube. The samples were stored at -20 “C before analysis. After thawing the devices were centrifuged at 2000 rev mini resulting in a clear, watery supernatant. For cortisol determination 100 ~1 of saliva were used for duplicate analysis with a recently developed time-resolved fluorescence immunoassay which is described elsewhere (Dressendorfer, Kirschbaum, Rohde, Stahl & Strasburger, 1992). The intra- and interassay coefficients of variance for this assay were less than 8% at a cortisol concentration of 5 nmol 1-l.

3.4. Statistical analysis

Changes in NK+ cell number and NKCA were tested for statistical significance with Student’s t-test for dependent measures. Differences between measures were considered statistically significant at (Y =Z 0.05. Furthermore, Pearson’s correlations were computed between NK+ cell number and NKCA.

4. Results

Fig. 1 shows alteration of the number of NKf cells (FACS analysis) on the acquisition days (Fig. l(A)) as well as on the conditioning test days (Fig. l(B)).

On acquisition day 3, analysis of NK+ cells revealed a significant increase of cell number after exposure of the CS + combined with epinephrine injection (day 3: t = 7.2; df = 15; p < 0.001). No alteration of NKf cell number, however, could be determined after exposure of the CS - on the acquisition day 8 (t < 1).

On the critical test days re-exposure of the CS + on day 11 as well as on day 13 resulted in a significantly increased number of NK+ cells (day 11: t = 2.6; df = 18; p = 0.02; day 13: t = 4.1; df = 15; p < 0.001). Again, no alteration could be deter- mined after presentation of the CS - .

As can be seen in Fig. 2 analysis of NKCA revealed a less pronounced effect than did analysis of NK+ cell number. After presentation of the CS + on day 3 and 6 subjects showed significantly elevated NK cell activity (day 3: t = 6.5; df = 15; p < 0.001; day 6: t = 7.2; df = 18; p < 0.001). Furthermore, comparison between both CS + acquisition trials indicated that alterations of NK cell activity on day 6

A. Buske-Kirschbaum et al. /Biological Psychology 38 (1994) 143-155 149

22 -

- B 20

3 8

cs+ cs-

DAYS / TREATMENT

0 Baseline *

m +20 min **

** *

L = L

11 12 13

cs+ cs- cs+

DAYS / TREATMENT

P<O.O5 P<O.OOl

Fig. 1. Natural killer (NK) cell number (CD16, CD 56) obtained 10 min before (baseline) and 20 min

after stimulus presentation. The number of NK positive cells was determined (A) on acquisition days 1,

3, 6 and 8 and (B) on test days 11, 12 and 13 by FACS analysis c” p < 0.05; ** p < 0.001).

I30 -

.A 50 -

: 40 -

k il

K 30-

s

E 20-

10

60

** .B **

T 50

40 - ** *

===lT 30

_L I =

20

10

Ai-__-i O- O-

3 6 1 8 11 12 13

cs+ cs- csc cs- cs+

DAYS / TREATMENT DAYS / TREATMENT

0 Baseline * : P=O.O55 m +20 mln ** : P<O.OOl

Fig. 2. Natural killer cell (NK) activity (in % lysis) obtained 10 min before (baseline) and 20 min after

stimulus presentation. NK cell activity was determined (A) on acquisition days 1, 3, 6 and 8 and (B) on test days 11, 12 and 13 by a standard 53Cr-release assay (4 h) c” p < 0.055; ** p < 0.001).

150 A. Buske-Kirschbaum et al. /Biological Psychology 38 (1994) 143-155

A 35

r = 0.57 ; 30 - x,Y

P i 0.001

0 10 20 30 40 50 60 70

NKCA (X Lysis)

.

.

0 10 20 30 40 50 60 70

NKCA (% Lysis)

Fig. 3. Linear regression of natural killer (NK) cell activity and total number of NK positive cells

computed (A) 10 min before and (B) 20 min after stimulus presentation.

was significantly higher than on day 3 (t = 2.4; p = 0.032) suggesting a stable effect of epinephrine on NK cell function without habituation effects.

On the critical test day 11, re-exposure of the CS + resulted in a slight elevation of NK cell activity which, however, did not reach statistical significance (day 11: t = 2.1; df = 18; p = 0.055). Again, no alteration of NK cell activity could be determined after presentation of the CS - (day 12: t = 1.3; &= 18; p = 0.22). However, after the second re-exposure of the CS + on the test day 13, significant elevation of NK cell activity was found (day 13: t = 3.6; df = 17; p = 0.002).

As shown in Fig. 3, modulation of NKCA was significantly and positively correlated with alteration of NKf cell number suggesting that increased number of NK+ cells appears to be one relevant mechanism to induce elevation of NK cell activity.

Measurement of salivary cortisol indicated no significant increase of cortisol concentration (data not shown). On all conditioning test days, cortisol concentra- tion showed a steady decline from the first to the last sample reflecting circadian rhythm fluctuations during morning hours. Thus, no elevated cortisol concentra- tion due to possible stressful stimulation could be determined.

5. Discussion

20 min after epinephrine injection healthy subjects showed a significantly elevated number of NK+ cells as well as a significantly increased NKCA. These data confirm recent observations demonstrating a stimulating effect of epinephrine on NK cell function (Toennesen et al., 1984; Toennesen, Christensen & Brinklov, 1987). Consistent with our previous conditioning experiments (Buske-Kirschbaum et al., 1992; Kirschbaum et al., 1992), this experiment confirmed the efficacy of epinephrine as a US which again elicited a significant and reliable unconditioned

A. Buske-Kirschbaum et al. /Biological Psychology 38 (1994) X43-155 151

response. No alteration of NK+ cell number or NKCA was found after CS - presentation, indicating that the experimental procedure per se did not effect NK cell function. This latter observation is especially noteworthy, because it is re- garded as vital for discriminative conditioning that the two different stimuli used as the alternative conditioned stimuli CS + and CS - , are correlated with the pres- ence or absence of the unconditioned response (Mackintosh, 1983).

Data analysis of the critical test days revealed a significantly enhanced number of NK+ cells after re-exposure of the CS + on both test days with the greater response on the second test day 13. As predicted, no alteration of NK+ cells could be determined after re-exposure of the CS - . Analysis of NKCA showed a similar pattern of results. However, while the total number of NK+ cells was significantly effected by the first and the second re-exposure of the CS + , significant elevation of NKCA was only observed after the second CS + re-exposure on day 13. Presentation of the CS + on day 11 resulted in a transient, but not statistically significant elevation of NKCA. Taken together, these results are generally consis- tent with a discriminative conditioning analysis suggesting that only re-exposure of the CS previously paired with epinephrine injection resulted in conditioned immu- noenhancement while a comparable CS without previous association with an immunomodulating treatment did not cause this effect. This consideration is supported by previous studies suggesting that immune function can be modulated by discriminative learning processes. Using a comparable design, Russel et al. (1984) reported that, after repeated pairing of a sulfur odor (CS + ) with injections of bovine serum albumin (US), guinea pigs showed conditioned increase of histamine release when reintroduced to the odor alone. No altered histamine release was found in response to a fishy odor (CS - 1, which previously has been presented without antigenic challenge. Comparable findings have been reported by others (Dark, Peeke, Ellman & Salfi, 1987). Furthermore, there is evidence indicating that the total number of white blood cells can be manipulated by classical conditioning processes in rats Wosterhalfen & Klosterhalfen, 1987; Gauci, Bull, Schedlowski, Husband & King, 1991). These findings have been extended by others showing, that besides classically conditioned modulation of white blood cell counts, the total number of different lymphocyte subsets can be influenced by classical conditioning processes. Husband, King and Brown (1987) recently reported that presentation of saccharin resulted in conditioned increase of the helper : suppressor (CD4 : CDB) T cell subset ratio when previously paired with the immunostimulant lavamisole.

The present data extend previous findings and suggest that also in humans the total number of a specific leukocyte subset, i.e. NK+ cells can be influenced by discriminative conditioning techniques. The reason why significant alteration of NKCA could be observed only after the second CS + re-exposure remains to be determined. It can be argued that differences in the reliability of the immunologi- cal assays used in the experiment might have contributed to the differences in the conditioned responses. It is known that biological assays are less reliable than quantitative immunological methods such as cell sorting. Alternatively, there are some reports indicating that, under certain circumstances, two re-exposures of the

152 A. Burke-Kirschbaum et al. /Biological Psychology 38 (1994) 143-155

CS are necessary to induce conditioned immunomodulation (Rogers et al., 1976; Wayner et al., 1978). Further research, however, is needed to clarify whether both response systems used in the present conditioning protocol are dissociated with respect to their sensitivity to the experimental manipulations.

The identification of the physiological basis for modulated NK+ cell number or NKCA in a discriminative conditioning protocol is as yet unknown. There are numerous pathways that might be involved. Actually, the correlated changes of NKCA and NK+ cell number after epinephrine administration support the view that a release of NKf cells from lymphoid organs into the peripheral blood may be one mediating mechanism of the unconditioned and the conditioned response. These considerations are in agreement with previous observations demonstrating increased number of NK+ cehs in the peripheral blood after epinephrine injection (Crary et al., 1983; Kappel et al., 1991) or epinephrine infusion (Brohee, Vanhae- verbeek, Kennes & Neve, 1990) in humans. Furthermore, exercise-induced eleva- tion of epinephrine is often described to be accompanied by an elevated number and activity of NK cells (Kendall, Hoffman-Goetz, Houston, MacNeil & Aru- mugam, 1990; Pedersen et al., 1988), suggesting a close relationship between sympathetic nervous system activity and NK cell function. fn addition, regulation of NKCA via beta-adrenoceptors was reported by Hellstrand, Hermodsson and Strannegard (1985), demonstrating that after incubation of lymphocytes with epinephrine, a significant increase of NKCA could be observed. These data are supported by findings indicating that NK cells, compared with other mononuclear cells (monocytes, T-lymphocytes, B-lymphocytes) are characterized by a high density of beta-adrenergic receptors (Maisel, Harris, Rearden & Michel, 1990). Additionally, acute in vivo exposure to catecholamines or beta-adrenergic agonists increases the number of beta-adrenergic receptors (Tohmeh & Cryer, 1980; Butler, Kelly, O’Malley & Pidgeon, 1983). Finally, reports of decreased NKCA after lesions with 6-OHDA suggest that regulation of NK cells by direct innervation of lymphoid tissues may be evoked to explain elevated NKCA in our conditioning protocol (Felten & Felten, 1991; Cross & Roszman, 1988).

The present data provide additional substantiation of classically conditioned modulation of NK cell function in humans. The present findings further suggest that also in humans the organism might be able to differentiate between two cues of different immunological relevance and to react anticipatorily with a learned immune response. The possible role of discriminative learning in host defense mechanisms remains to be determined in future research.

Acknowledgments

We thank Prof. Dr. R. Ballieux and Prof. Dr. H. Kimmel for their most helpful criticisms and comments. Furthermore, we thank Mrs. I. Rummel-Friihauf for her excellent technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft, Grant He 1013/4-l.

A. Buske-Kirschbaum et al. /Biological Psychology 38 (1994) 143-155 153

References

Ader, R., & Cohen, N. (1975). Behaviorally conditioned immunosuppression. Psychosomatic Medicine,

37, 333-340. Ader, R., & Cohen, N. (1982). Behaviorally conditioned immunosuppression and murine systemic lupus

erythematosus. Science, 214, 1534-1536. Ader, R., & Cohen, N. (1991). The influence of conditioning on immune response. In R. Ader, D.L.

Felten & N. Cohen (Eds), Psychoneuroimmunology (pp. 609-640). San Diego: Academic Press. Besedovsky, H.O., & de1 Rey, A. (1991). Physiological implications of the immune-neuro-endocrine

network. In R. Ader, D.L. Felten & N. Cohen (Eds), Psychoneuroimmunology (pp. 589-604). San

Diego: Academic Press. Blalock, J.E. (1984). The immune system as a sensory organ. Journal of Immunology, 132, 1067-1070. Belles, R.C. (1979). Learning theory. New York: Holt, Rinehart & Winston, 1979

Bovbjerg, D.H., Redd, W.H., Maier, L.A., Holland, J.C., Lesko, L, M., Niedzwiecki, D., Rubin, SC., &

Hakes, T.B. (1990). Anticipatory immune suppression and nausea in women receiving cyclic

chemotherapy for ovarian cancer. Journal of Consulting and Clinical Psychology, 38, 153-157. Brohee, D., Vanhaeverbeek, M, Kennes, B., & Neve, P. (1990). Leukocyte and lymphocyte subsets after

a short pharmacological stress by intravenous epinephrine and hydrocortisone infusion in healthy

humans. International Journal of Neuroscience, 53, 53-62. Buske-Kirschbaum, A., Kirschbaum, C., Stierle, H., Lehnert, H., & Hellhammer, D.H.tl992). Condi-

tioned increase of natural killer (NK) cells in humans. Psychosomatic Medicine, 54, 123-132.

Butler, J., Kelly, J.G., O’Malley, K., & Pidgeon, F. (1983). Beta-adrenoceptor adaptation to acute

exercise. Journal of Physiology, 344, 113-117. Crary, B., Hauser, S.L., Borysenko, M., Keutz, I., Hobem, C., Ault, K.A., Weiner, H.L. & Benson, H.

(1983). Epinephrine-induced changes in the distribution of lymphocyte subsets in peripheral blood.

Journal of Immunology, 131, 1178-1181. Cross, R.J., & Roszman, T.L. (1988). Central catecholamine depletion impairs in vivo immunity but not

in vitro lymphocyte activation. Journal of Neuroimmunology, 19, 33-45. Dark, K., Peeke, H.V.S., Ellman, G., & Salfi, M. (1987). Behaviorally conditioned histamine release.

Prior stress and conditionability and extinction of the response. Annals of the New York Academy of Science, 496, 578-582.

Dressendorfer, R.A., Kirschbaum, C., Rohde, W., Stahl, F., & Strasburger, C.J. (1992).Synthesis of a

cortisol-biotin conjugate and evaluation as a tracer in an immunoassay for salivary cortisol

measurement. Journal of Steroid Biochemistry and Molecular Biology, 43, 683-692 Dyck, D.G., Greenberg, A.H., & Osachuk, T.A.G. (1986). Tolerance to drug-induced(Poly I : C) natural

killer cell activation: Congruence with a Pavlovian conditioning model. Journal of Experimental Psychology, 12, 25-31.

Dyck, D.G., Driedger, S.M., Nemeth, R., & Osachuk, T.A.G. (1987). Conditioned tolerance to

drug-induced (Poly I: C) natural killer cell activation: Effects of drug-dosage and context-specificity parameters. Brain, Behavior, and Immunity, I, 251-266.

Felten, D.L., & Felten, S. (1991). Innervation of lymphoid tissue. In R. Ader, D.L. Felten & N. Cohen

(Eds.), Psychoneuroimmunology (pp. 27-53). San Diego: Academic Press.

Gauci, M., Bull, D.F., Schedlowski, M., Husband, A.J., King, M.G. (1991). Lithium chloride and immunomodulation in taste aversion conditioning. Physiology and Behavior, 51, 207-210.

Ghanta, V., Hiramoto, R.N., Solvason, B., & Spector, N. (1985). Neural and environmental influences

on neoplasia and conditioning of NK cell activity. Journal of Immunology, 135, 848s-852s. Ghanta, V., Hiramoto, R.N., Solvason, B., & Spector, N. (1987). Influence of conditioned natural

immunity on tumor growth. Annals of the New York Academy of Science, 496, 637-646. Girgis, A., Shea, J., & Husband, A (1988). Immune and psychological response to acute venipuncture

stress. Medical Science Research, 16, 351-352. Gorczynski, R.M., Macrae, S., & Kennedy, M. (1982). Conditioned immune response associated with

allogeneic skin grafts in mice. Journal of Immunology, 129, 704-709. Hellstrand, K., Hermodsson, S., & Strannegard, D. (1985). Evidence of a beta-adrenoceptor mediated

regulation of human natural killer cells. Journal of Immunology, 134, 4059-4095-4099

154 A. Buske-Kirschbaum et al. /Biological Psychology 38 (1994) 143-155

Husband, A.J., King, M.G., & Brown, R. (1987). Behaviorally conditioned modification of T cell subset ratios in rats. Immunology Letters, 14, 91-94.

Irwin, M., Daniels, M., Bloom, W., & Weiner, H. (1986). Life events, depression and natural killer cell activity. Psychophamacological Bulletin, 22, 1093-1096.

Kappel, M., Tvede, N., Galbo, H., Haahr, P.M., Kjaer, M., Linstow, M., Klarlund, K. & Pedersen, B.K. (1991). Evidence that the effect of physical exercise is mediated by epinephrine. Journal of Applied

Physiology, 70, 2530-2534.

Kendall, A., Hoffmann-Goetz, L., Houston, M., McNeil, B., & Arumugam, Y. (1990). Exercise and

lymphocyte subset responses: Intensity, duration, and subject fitness effects. Journal of Applied

Physiology, 69, 251-260.

Kiecolt-Glaser, J.K., Garner, W., Speicher, C., Penn, G.M., Holliday, J., & Glaser, R.(1984). Psychoso-

cial modifiers of immunocompetence in medical students. Psychosomatic Medicine, 46, 7-14.

Kiecolt-Glaser, R.K., Glaser, R., Williger, D., Stout, J., Mesich, G., Sheppard, S., Richer, D.,

Romisher, S.C., Briner, W., & Bonell, G. (1985). Psychosocial enhancement of immunocompetence in a geriatric population. Health Psychology, 4, 25-41.

Kirschbaum, C., Jabaaij, L., Buske-Kirschbaum, A., Hennig, J., Blom, M., Dorst, K., Bauch, J., DiPauli, R., Schmitz, G., Ballieux, R., & Hellhammer, D.H. (1992). Conditioning of drug-induced im-

munomodulation in human volunteers: A European collaborative study. British Journal of Clinical

Psychology, 31, 459-472.

Klosterhalfen, W., & Klosterhalfen, S. (1983). Pavlovian conditioning of immunosuppression modifies

adjuvans arthritis in rats. Behauioral Neuroscience, 4, 663-666.

Klosterhalfen, S., & Klosterhalfen, W. (1987). Classically conditioned effects of cyclophosphamide on

white blood cells counts in rats. Annals of the New York Academy of Science, 496, 569-577.

Kusnecov, A.V., Husband, A.J., & King, M.G. (1988). Behaviorally conditioned suppression of

mitogen-induced proliferation and immunoglobulin production: Effect of time span between condi-

tioning and re-exposure to the conditioned stimulus. Brain, Behavior and Immunity, 2, 198-211.

Mackintosh, N.J. (1983). Conditioning and associative learning. Oxford: Clarendon Press.

Maisel, A.S., Harris, T., Rearden, C.A., & Michel, M.C. (1990). Beta-adrenergic receptors in lympho-

cyte subsets after exercise. Alteration in normal individuals and patients with congestive heart

failure. Circulation, 82, 2003-2010.

Marx, M.H. (1971). Learning: Processes. New York: Macmillan.

Neveu, P.J., Dantzer, R., & Le Moal, M. (1988). Behaviorally conditioned suppression of mitogen-in-

duced lymphoproliferation and antibody production in mice. Neuroscience Letter, 65, 293-298.

O’Reilly, C.A., & Exon, J.H. (1986). Cyclophosphamide-conditioned suppression of the natural killer

cell response in rats. Physiology and Behauior, 37, 759-764.

Ottaway, C.A., & Husband, A.J. (1992). Central nervous system influences on lymphocyte migration.

Brain, Behavior and Immunity, 6, 97-116.

Pedersen, B.K., Tvede, N., Hansen, F.R., Andersen, V., Bendix, T., Bendixen, G., Bendtzen, U., Galbo,

H., Haahr, P.M., & Klarlund, K. (1988). Modulation of natural killer cell activity in peripheral blood by physical exercise. Scandinavian Journal of Immunology, 27, 673-678.

Rogers, M.P., Reich, P., Strom, T.B., & Carpenter, C.B. (1976). Behaviorally conditioned immunosup-

pression: Replication of a recent study. Psychosomatic Medicine, 38, 447-452.

Russel, M., Dark K.A., Cummins, R.W., Ellmann, G., Callaway, E., & Peeke, H.V. (1984). Learned

histamine release. Science, 225, 733-734.

Schlesinger, M., & Yodfat, Y. (1988). Effect of psychosocial stress on natural killer cell activity. Cancer

Detection and Prevention, 12, 9-14.

Smith, G.R., & McDaniels, S.M. (1983). Psychologically mediated effect on the delayed hypersensitivity

reaction to tuberculin in humans. Psychosomatic Medicine, 45, 65-70.

Toennesen, E., Toennesen, J., & Christensen, N.J. (1984). Augmentation of cytotoxicity by natural killer (NK) cells after adrenaline administration in man. Acta Pathologica, Microbiologica et Immunologica

Scandinavia, 92, 81-83.

Toennesen, E., Christensen, N.J., & Brinklov, M.M. (1987). Natural killer cell activity during cortisol and adrenaline infusion in healthy volunteers. European Journal of Clinical Investigation, 10,

497-503.

A. Buske-Kirschbaum et al./Biological Psychology 38 (1994) 143-155 155

Tomeh, J.F., & Cryer, P.E. (1980). Biphasic adrenergic modulation of /3-adrenergic receptors in man.

Journal of Clinical Inuestigation, 65, 836-840. Trinchieri, G. (1985). Biology of natural killer cells. Advances in Immunology, 47, 187-376. Wayner, E.A., Flannery, G.R., & Singer, G. (1978) The effect of taste aversion conditioning on the

primary antibody response to sheep red blood cells and Bruccella abortus in the albino rat.

Physiology and Behavior, 21, 995-1000.