What do the immune system and the brain know about each other?

342

References 1 Mackay, I. R. and Burnet, F. M. (1963)Autoimmune D~eases: Pathogenesis,

Chemistry and Therapy, Thomas, Springfield, IL 2 Hooper, B., Whittingham, S., Mathews, J . D. eta/. (1972) Clin. Exp.

Immunol. 12, 79-87 3 Hawkins, B. R., Cheah, P. S., Dawkins, R. L. eta/. (1980) Lancet ii,

1057-1059 4 Roberts, I. M., Whittingham, S. and Mackay, I. R. (1973) Lancet ii,

936-940 5 Primi, D., Hammarstr6m, L., Smith, C. I. E. and M611er, G. (1977)J.

Exp. Med. 145, 21-30 6 Weigle, W. O., ChiUer, J . M. and Habicht, G. S. (1972) Transplant. Rev.

Immunology Today, vol. 4, No. 12, 1983

8, 3-25 7 Grabar, P. (1975) Clin. ImmunoL Immunopathol. 4, 453-466 8 Teale, J . and Mackay, I. R. (1979) Lancet ii, 284-287 9 Daar, A. S. and Fabre, J . W. (1981) Clin. Exp. Immunol. 45, 37-47

10 Guilbert, B., Dighiero, G. and Avreamos, S. (1982)J. Immunol. 128, 2779-2787

11 Dighiero, G., Guilbert, B. and Avreamos, S. (1982)~ Immunol. 128, 2788-2792

12 Pedersen, J . S., Toh, B. H., Mackay, I. R. eta/. (1982) Clin. Exp. ImmunoL 48, 527-532

13 Mackay, I. R. (1956)Aust. Ann. Med. 5, 244-253 14 Raft, M. (1982)Nature (London)298, 791-792 15 Nossal, G. J . V. (1983)Annu. Rev. ImmunoL 1, 33-62

1 What do the immune system and the brain know about

each other? Hugo O. Besedovsky, Adriana E. del Rey and Ernst Sorkin

The immune system is generally viewed as being regulated by a variety of mechanisms 'from within'. This internal regulation is conceived to be mediated by different subsets of T cells, by antibodies including their idiotypic determinants, monokines, lymphokines, etc., which appear in a given sequence and quantity. This autoregulation confers a high degree of autonomy on the immune system. However, processes essential for the functioning of immunological cells, such as meta- bolism, transport of substances, allosteric changes in membranes, lymphoid cell proliferation and transformation, and lymphokine synthesis (for bibliography, see Ref 1), are affected by several hormones and neurotransmitters. These facts constitute by themselves good arguments for another kind of regulation, one 'from without' (a term used by Medawar in 1973) which is, as Hugo Besedovsky, Adriana del Rey and Ernst Sorkin have proposed, superimposed upon and inter- woven with autoregulation. This postulate requires the existence of information channels between the immune system and the central nervous system, i. e. that they should know about each other. The apparently lofty title of this presentation raises a number of critical questions about the workings of the immune system in its natural environment. Here the authors discuss these questions." What do immune cells know about the brain? What does the brain know about the immune system? How do the brain and the immune system communicate with each other?Are brain-immune system interactions

linked in regulator.y feedback circuits?

What do immunological ceils know about the brain and mechanisms under brain control?

In-vitro studies of the immune system have led to a con- siderable expansion of our knowledge of its cellular com- position and the function of the multiple subsets of lymphocytes and of accessory cells. Necessary as these in-vitro techniques are for studying autoregulation, they carry with them the disadvantage of removing immunological cells from their natural environment. Under these artifi- cial conditions possible regulatory neuroendocrine signals for immunological cells obviously cannot be detected. There is no doubt, however, that immunological cells are exposed in vivo to brain messengers such as hormones and neurotransmitters, since these powerful agents are ubiquitous in tissues and body fluids. There is also growing experimental evidence that immunological organs are innervated, mainly by sympathetic fibers. For example, perivascular plexuses within the splenic white pulp send single noradrenergic fibers between the sur- rounding lymphocytes. Some of these nerves form an intimate relationship with these cells 2'3. It is of course still possible that immunological cells show an exceptional characteristic, namely they may consider these agents as

Medical Department, Swiss Research Institute,. CH-7270 Davos-Platz, Switzerland.

part of the 'background noise'. Since hormones and neurotransmitters are present in the microenvironment of immunological cells, an essential requirement for the i r participation in immunoregulation is the presence of receptors for these messengers on such cells. In fact receptors for corticosteroids, insulin, prolactin, growth hormone, oestradiol, testosterone, fl-adrenergic agents, acetylcholine and endorphins have been demonstrated in lymphoid or accessory cells 4.

These facts show that immunological cells can 'see' and interact with brain or brain-linked messengers present in their natural environment. Furthermore, some hormones and neurotransmitters can modulate intracellular nucleo- tide levels (for example cAMP, cGMP) in lymphocytes and accessory cells. These messengers are supposed to participate in the process of lymphocyte activation 5. This suggests the existence of intracellular pathways common to hormones and neurotransmitters on the one hand, and antigens and signals intrinsic to the immune system on the other.

Investigations into the participation of hormones in the immune response have generally involved either the parenteral administration of hormones or the ablation or blockade of endocrine glands. Numerous reports agree that hormone administration can lead to depressed or stimulated immune responses, depending on the kind

© 1983, Elsevier Scien~ Publishers B.V., Amsterdam 0167 - 4919/83/S01 .GO


and dose of hormones and the timing of their administration. In general, glucocorticoids, androgens, oestro- gens and progesterone depress the immune response in vivo, whereas growth hormone, thyroxine and insulin increase the response. Sex differences in the immune response are also well documented and such differences are important in autoimmune disease 6. It is beyond the scope of this presentation to review the vast literature on this subject (see Refs 4, 7-10).

Data on the effects of mediators of the autonomic nervous system on the immune system are contradictory but in the main they indicate that neurotransmitters can influence the immune response, both in vitro and in vivo. Some of the previously obtained data on catecholamines have been discussed elsewhere"'~2.

Parasympathetic agents have been reported to increase antibody formation 13 and cytotoxicity ~. Direct manipulation of the brain can affect the immune response, for example electrolytical lesions or stimulation of different parts of the brain influence the immune response either directly or by altering hormonal controls ~5-~7. Also, studies on effects of stress and conditioning showed that environmental stimuli can interfere with the immune response (see Refs 18, 19).

In summary, the facts discussed above permit us to answer in the affirmative the question formulated at the outset: potential messages from the brain, through neuroendocrine and autonomic mechanisms, can be perceived by immunological cells and affect their function. The question remains to be answered, however, whether these possible messages are generated only under extreme conditions or whether they constitute part of feedback mechanisms operating permanently between brain and immunological cells.

What does the brain know about the immune system? For years our research has been oriented to answer this

question. We believe that in analogy to all types of regulatory mechanisms, a reciprocal flow of information should exist between the immune system and the brain. This is a basic requirement in postulating that central mcA~nisrns are involved in immunoregulation. Since the immune response is a phasic phenomenon, the brain should be able to detect immune signals derived from activated cells at different given steps of the immune response. These signals should give rise to brain, autonomic and endocrine responses.

Our methodological approach in searching for such centrally mediated responses was to elicit and evaluate an immune response, and to try to identify brain or brain- mediated reactions in the same animals. Manipulat ion of the host was reduced to a minimum, just one intraperi- toneal injection of a non-toxic, non-infective, non-neo- plastic and non-pyrogenic antigen. Some of the results were derived from direct studies in the brain while others were obtained from biochemical approaches by meas- uring autonomic and neuroendocrine responses.

Electrophysiological studies were performed in collaboration with D. Felix and H. Haas at the University of Ziirich. In the same animal (rat) we studied both the immune response and the rate of firing of individual hypothalamic neurons at various intervals after injection

343

of sheep red blood cells (SRBC) or TNP-haemocyanin ~°. Animals stimulated with SRBC showed on day 1 no plaque-forming cells (PFC) and no changes in firing fre- quency. On day 5, PFC in spleen were maximal and there wa~ a more than twofold increase in the firing rate ofven- tromedial neurons. In several rats which were immunological non-responders, no increase in firing rates occurred. Furthermore, no changes in firing rates were observed in simultaneous recordings in the anterior hypothalamic nucleus of immunologically responding rats. Increased frequencies in neuronal firing (ventromedial hypothalamic nucleus) were also observed during the immune response to TNP-haemocyanin. The highest activation occurred on day 2, a time close to the peak number of splenic direct PFC. These data clearly show that the hypothalamus has received direct or indirect immunological signals.

Since brain aminergic neurons projecting into the hypothalamus are major controllers of hypothalamic neuronal activity, turnover rates were studied in several parts of the brain. Catecholamine (CA) turnover rates reflect the degree of catecholamine synthesis which is a function of aminergic neuronal activity. These experiments were done in collaboration with M. Da Prada, Hoffmann-La Roche, Basel. SRBC were used for immunization. In those animals which showed a high immune response to SRBC, a marked decrease in hypothalamic noradrenaline (NA) turnover was detected at the peak of the immune response when compared with low responders and non-immunized controls 21. No changes in NA turnover were detected early during the immune response. In the brain cortex no NA changes were noted. Also the dopamine turnover in the hypothalamus remained unchanged.

Our studies were also oriented towards detecting possible peripheral autonomic and endocrine responses in the course of the immune response. The possibility of sympathetic neurotransmitters being involved in immunoregulation was based on the knowledge that lymphoid organs are sympathetically innervated, and on data from this and other laboratories showing that surgical denervation of the spleen or chemical sympathectomy leads to an enhanced immune response 3'~1'22. Interesting as these data are, they do not permit us to decide whether such effects reflect the existence of sympathetic immunoregulation since they are based on manipulation of sympathetic mechanisms. Therefore, we analysed whether sympathetic nerve activity in lymphoid organs, as reflected by their NA level, changes during the immune response. After antigen challenge (SRBC) we found a marked decrease (40-70 %) in NA content in rat spleen on days 3 and 4". The degree and persistence of this NA decrease is inversely related to the magnitude of the response 23. No change in the NA content of a non- lymphoid organ, the heart, was discerned.

The magnitude of NA decrease is somehow surprising considering that NA tissue levels are very stable even under extreme conditions. This stability is assured by a refined feedback mechanism which antagonizes NA fluc- tuation and operates at the level of the rate-limiting step of NA synthesis, the activity of tyrosine hydroxylase. This situation by itself makes it very unlikely that one is dealing

344

with a purely local phenomenon such as higher NA consumption or degradation that would account for the NA decrease.

Since animals are constantly developing immune responses to environmental antigens, the basal activity of sympathetic nerves may be influenced by the degree of immunological activity. Germ-free (GF) animals, which are subjected to a minimal degree of external antigenic challenge, and specific pathogen-free (SPF) animals, which are constantly infected and protected only against pathogens, offer an excellent model to study the basal levels of catecholamines in lymphoid organs. According to the results and rationale described above, the prediction was that the immunologically more engaged SPF rats should have lower NA levels in their lymphoid organs than their GF counterparts. As predicted, lymphoid tissues of SPF rats such as spleen, thymus and lymph nodes had about half the NA content of GF animals. The NA content of rat stomach and intestine (duodenum) was similar in SPF and GF conditions2L We explain these results by postulating that antigenic exposure of conven- tional or SPF animals during their lifespan contributes to, or even causes, the described lower NA content in the lymphoid organs. In the same experiments the adrenaline (A) and NA contents of the adrenal gland were also deter- mined. Total adrenal NA and A content was significantly lower in SPF animals.

In these studies two facts deserve our special attention. First, the thymus, an organ only marginally involved in the immune response, has less NA in SPF than in GA animals. Second, the adrenal medulla, which is centrally controlled, has less NA and A in SPF animals. Again, both situations suggest that the effect of the immune response on the NA content of lymphoid organs is not only based on higher rates of NA consumption and degradation. All these findings attest, in our view, to sympathetic participation in immunoregulation.

In the following we wish to discuss another immunoregulatory mechanism, the one mediated by the adrenal cortex, which directly or indirectly is under brain control. Glucocorticoid hormones exert well-known multifaceted effects on immunity. However, it remains unclear whether these hormones are relevant for immunoregulation under physiological conditions. The following data have provided the primary evidence to support the view that they are. After antigenic challenge with three different antigens (SRBC, HRBC, TNP-haemocyanin) in two species (rats, mice), increased glucocorticoid blood levels were noted at about the time of the peak of the immune response 25. This increase in blood glucocorticoid level was directly related to the magnitude of the immune response (unpublished observations). The increased hormone levels are known to be immunosuppressive and furthermore, as we have shown earlier, they are capable of suppressing the response to unrelated antigens. In fact, inhibition of this hormone increase by adrenalectomy can overcome sequential antigenic competition 26.

The fact that the immune response is able to modify some brain functions and can induce responses in brain- linked autonomic and endocrine mechanisms already shows that the brain is receiving information originally derived from immunological cells. The examples men-


tioned, and predictably there are more to come, thus provide a strong body of evidence for the conclusion t h a t the brain does indeed 'know what the immune system is doing'.

How do the immune system and the brain communicate with each other?

Immunological cells are receptive to brain messages mediated by neurotransmitters, hormones and nerves. On the other hand, our knowledge that the immune response can elicit electrical and biochemical changes in the brain shows that messages derived from immunological cells are directly or indirectly perceived by the brain. Some such messengers which can affect brain mechanisms are already known to be produced by certain immunological cells, for example histamine, serotonin, prostaglandins, and interferon which shares common amino acid sequences with adrenocorticotropic hormone (ACTH) and fl-endorphin 27. It is not yet known whether under physiological conditions these messengers are released in sufficient quantity during the immune response to elicit brain or brain-linked responses.

Our own approach in the search of immune-derived messengers consisted of stimulating in vitro rat spleen cells or human peripheral blood cells and injecting supernatants from such cultures into normal animals. The prediction was that stimulation of a few cells will release into the culture medium soluble mediators capable of inducing brain and neuroendocrine changes. Since these animals are receiving preformed messengers, these changes should mimic within hours those produced within days during the immune response.

Thus, in collaboration with C. Honegger and R. Burri, of the Neurobiology Unit at The University of Basel, supernatants containing activated immune cell products were assayed for their capacity to induce, as the immune response does, a decrease in hypothalamus NA levels in rats. Two hours after injection of supernatants from 106 rat spleen cells stimulated with concanavalin A (Con A), the NA concentration in the hypothalamus was reduced when compared with the supernatants obtained from non-stimulated lymphoid cells. This NA decrease corresponds to more than 50 % of the maximum possible decrease induced by a-methyl-p-tyrosine, which blocks NA synthesis for several hours. Therefore, 106 spleen cells can produce sufficient quantities of mediators capable of strongly inhibiting NA synthesis in the hypothalamus and also in the brain stem of an intact animal 21. This product is thus most likely the messenger which mediates the decrease in hypothalamic synthesis of NA observed during the immune response.

We have also explored the possibility that 'factors' released by immunological cells can affect blood levels of glucocorticoids as is observed during the immune response. Supernatants of Con A-stimulated rat spleen ceils (106) or of human peripheral leukocytes (107) were injected into rats. The outcome of these experiments was unequivocal. Increased blood corticosterone levels were obtained which, in their magnitude, were similar to those observed in antigen-challenged rats 28. However, as expected, the hormone increase occurred within 30-120 minutes after application of preformed mediators instead


of after several days. Therefore, injection of the mediators derived from immunological cells can also in this instance mimic the events occurring after antigen injection.

Since in other experiments supernatants obtained from 10 6 human peripheral lymphocytes (containing less than 0.1% monocytes) have also produced these effects, it is likely that lymphocytes are the cell source of the messenger(s). Our results were confirmed by several groups which showed that similar supernatants, after injection into humans, produced a several-fold increase in blood cortisol ~'a6. Experimental evidence shows that the im- munohormonal messengers which increased blood glucocorticoid levels do not act directly on the adrenal cortex but via the adenohypophysis. Thus we found that blockade of A C T H output by surgical (hypophysectomy) or chemical means (dexamethasone) prior to injection of the soluble immunological cell-derived mediators prevents entirely the output of glucocorticosteroid. Our data are in agreement with the observation that similar material does not affect the function of the adrenal cortex in vitro ~1 .

In summary, the question posed above can also be answered in the affirmative: brain messengers such as hormones and neurotransmitters, which mediate central signals, can influence the immune response; and uniden- tified immunological signals can affect brain, autonomic and endocrine functions. Thus there is no lack of means by which the immune system and the brain can communicate with each other.

Are the brain- immune system interactions linked in regulatory feedback circuits?

Immunoregulation, as any regulation, requires a dynamic flow of signals between the parameters to be regulated and the regulatory agencies. These signals should be integrated in circuits which guide the immune system towards a given program. Attempts to analyse external immunoregulation must therefore be based on knowledge of the circuitry between the immune system and the brain and neuroendocrine structures.

As is common in studies of regulatory circuits, experi- mentation permits only successive stages of approxima- tion. In this sense, the available but still very incomplete information that during the immune response an exchange of messages occurs between the brain and immunological cells is subject to similar considerations in the context of immunoregnlatory feedback circuits. For instance the functional changes, which we have observed in the brain during the immune response, may be integrated in a central loop which may give rise to peripheral immunoregulatory signals. It is well known that noradrenergic neurones projecting into the hypothalamus decrease the firing rates of several neuronal types. It is therefore conceivable that the inhibition of central noradrenergic neurones, which is likely to be mediated by messengers derived from the immune response, may result in a release of hypothalamic neurones from restraint. This may explain the observed increase in firing rates. The increased activity of neurones may in turn induce neuroendocrine and autonomic changes which, by affecting immunological cells, will then close the circuit.

With regard to a putative, sympathetically mediated

345

immunoregulatory circuit, our data show that a decrease in sympathetic nerve activity in lymphoid organs occurs in association with immune reactions. In the hypothalamus and the brain stem there is also a reduction in NA content which, as previously discussed, is mediated by immunological cell products.

Considering all these data together, it seems that immunological cells release one or several powerful products capable of inhibiting NA synthesis in peripheral and central neurones. This central and peripheral response may be integrated in a complex sympathetic reflex mechanism triggered by the immune response itself.

What could be the immunoregulatory meaning of this mechanism? We have demonstrated that surgical dener- ration of the spleen increased the immune response. Also, chemical sympathectomy of permanent character enhanced the immune response ~'1z22. On this basis it can be postulated that the sympathetic reflex mechanism, that is triggered before the peak of the immune response, may release immunological cells from inhibition and thereby create conditions which favor the take-offofthe response. In addition, the observed decrease in NA may at least in part be responsible for the increased blood flow shown to occur in the spleen during the immune response, a process which affects lymphoid cell recirculation.

As regards the glucocorticoid-associated immunoregulatory circuit, the above data permit us now to propose how such a circuit may operate. During the immune response immunological cells release still unknown soluble mediators which most likely act via the hypothalamus. It is in general accepted that noradrenergic neurones projecting into the hypothalamus inhibit corticotrophin- releasing factor (CRF)-producing cells. As NA synthesis in the hypothalamus is decreased during the immune response, increased production of CRF takes place which then acts on the adenohypophysis. As a consequence, A C T H is released and blood levels of glucocorticoid hormone increase to immunosuppressive levels.

What could be the regulatory meaning of this glucocorticoid-associated circuit? It is known that apart from metabolic effects glucocorticoid at concentrations within the physiological range inhibit cells producing inter- leukins 1 and 2. On this basis, the increased glucocorticoid blood levels can control the clonal expansion of those cells with a high affinity for the antigen. On the other hand, since resting immunological cells are much more sensitive to glucocorticoids than activated cells, we postulate that this circuit may have the function of preventing the excessive expansion of cells with low affinity for the antigen or just the polyclonal influence ofinterleukins. In this sense it is conceivable that this circuit plays a role in preventing autoimmune and lymphoproliferative diseases.

We have summarized here evidence for the existence of centrally mediated immunoregulatory circuits. Since immunological processes are continuously in action we believe that these circuits are in permanent operation. Admittedly, for conceiving such circuits we have, aside from firm data, also used much indirect evidence and speculation. However, we view this as the appropriate way to look for experimental models to challenge new working hypotheses.

346

It is now necessary to identify common pathways and/ or alternate sequencies of the external and internal signals involved in immunoregulation. In this sense it would be misleading to assume that these two categories of signals constitute two different immunoregulatory mechanisms which function quite independently of each other. In fact, many examples are available which support the above concept that externally and internally derived signals act at common levels. To name just a few: hormonal control of monokine and lymphokine production; increased number of hormonal receptors during lymphoid cell transformation (insulin, glucocorticoid, fJ-adrenergic); changes in IgM-IgG switch induced by glucocorticoids. Furthermore, some hormones and neurotransmitters control intracellular levels of cyclic nucleotides involved in immune cell activation. Even at the genetic level links were observed, for example the major histocompatibility complex (MHC) controls testosterone blood levels (for bibliography see Ref. 32).

In summary, we visualize immunoregulation as a dynamic and complex process based on a sequence of exchanges of signals between immunological cells themselves and between these cells and brain and brain- associated mechanisms.

The data showing that messengers from the immune system can alter brain activity could also be considered in a broader context. One of the major functions of the brain is to process information on changes in the external and internal environments detected by receptor organs. The evidence discussed above shows that the brain is informed about the intrusion of antigenic macromolecules and possibly also about modified self-antigens. Immunologi- cal cells which express a huge repertoire of specific receptors for antigenic agents may, by way of releasing appropriate but different combinations of immunohor- monal messengers, be the ultimate source of this information for the brain. The immune system may thus act as the peripheral receptor organ for this type of external or internal stimulus.

A c k n o w l e d g e m e n t s This work was supported by the Swiss National Science Foundation,

Grant No. 3.603.80.

R e f e r e n c e s 1 Besedovsky, H. O. and Sorkin, E. (1977) Clin. Exp. Immunol. 27, 1-12 2 Reilly, D., McCuskey, A., Miller, L. et al. (1979) Tissue Cell 11, 121-126


3 Williams, J. M. and Fehen, D. L. (1981) Anat. Rec. 199, 531-542 4 Besedovsky, H. O., del Rey, A. and Sorkin, E. (1983) in lmmunoregulation

(Fabris, W., Garaci, E., Hadden, J, and Mitchison, N. A., eds), pp. 315-339, Plenum Press, London

5 Hadden, J. w. (1983) in Immunoregulation (Fabris, W., Garaci, E., Hadden, J. and Mitchison, N. A., eds), pp. 201-230, Plenum Press, London

6 Talal, N. (1977) Autoimmum'ty (Talal, N., ed.), pp. 194-197, Academic Press, New York

7 Billingham, R. E., Krohn, P. L. and Medawar, P. B. (1951) Br. Med. J. 2, t049-1053

8 Dougherty, T. F., Berliner, M. L , Schneebeli, G. O. and Berliner, D. L. (1964) Ann. NYAcad. So'. 113, 825-843

9 Wolstenholme, G. E. W. and Knight, J. (eds) (1970) Hormones and the Immune Response, Ciba Foundation Study Group No. 36, J. and A. Churchill, London

10 Claman, H. N. (1975)J. AUergy Clin. ImmunoL 55, 145-151 11 Besedovsky, H. O., del Rey, A., Sorkin, E. etaL (1979) Cell. Immunol. 48,

346-355 12 Hall, N. R. and Goldstein, A. L. (1981) in Psychoneuroirnmunology (Ader,

R., ed.), pp. 521-543, Academic Press, New York 13 Marat, B. A. (1974) Bull. Exp. Biol. Med. 76, 971-973 14 Strom, T. B., Sytkowski, A. J , Carpenter, C. B. and Merill, J. P. (1974)

Proc. Natl Acad. Sci. USA 71, 1330-1333 15 Korneva, E. A. and Khai, L. M. (1967) FizioL Zh. SSSR im I M Sechmova

53, 27-42 16 Jankovic, B. D. and Isakovic, K. (1973) Int. Arch. Allergy Appl. Immunol.

45, 360-372 17 Stein, M., Schiavi, P. C. and Camerino, M. (1976) Science 191,435-440 18 Ader, R. and Cohen, N. (1981) in Psychoneuroimmunology (Ader, R., ed.),

pp. 281-319, Academic Press, New York 19 Riley, V., Fitzmaurice, M. A. and Spackman, D. H. (1981) in

Psychoneuroimmunology (Ader, R., ed.), pp. 31-102, Academic Press, New York

20 Besedovsky, H. O., Sorkin, E., Felix, D. and Haas, H. (1977) Eur. J. Immunol. 7, 323 325

21 Besedovsky, H. O., del Rey, A., Sorkin, E. et ul. (1983) Science 221, 564-566

22 Miles, K., Quint~.ns, J., Chelmicka-Schorr, E. and Arnason, B. G. W. (1981),/. Neuroimmunology 1, 101-105

23 del Rey, A., Besedovsky, H. O., Sorkin, E. eta/. (1982) Am. J . Physiol. 242 (Regulatory Integrative Comp. Physiol. 11), R30-R33

24 del Rey, A., Besedovsky, H. O., Sorkin, E. etaL (1981) Cell. lmmunol. 63, 329-334

25 Besedovsky, H. O., Sorkin, E., Keller, M. and Miiller, J. (1975) Pr0c. Sot. Exp. Biol. Med. 150, 466-470

26 Besedovsky, H. O., del Rey, A. and Sorkin, E. (1979) Clin. Exp. Immunol. 37, 106-113

27 Smith, E. M. and Blalock, J. E. (1981) Proc. Nail Acad. Sci. USA 78, 7530-7534

28 Besedovsky, H. O., del Rey, A. and Sorkin, E. (1981),/. Immunol. 126, 385-387

29 Bindon, C., Czerniecki, M., Ruell, P. et al. (1983) Br. J. Cancer 47, 123-133

30 Pulley, M. S., Dumonde, D. C., Carter, G. et al. (1982) in Human Lymphokines (Khan, A. and Hill, N. O., eds), pp. 651-664, Academic Press, New York

31 Vahouny, G. V., Kyeyune-Nyombi, E., McGillis, J. P. et al. (1983)J. lmmunol. 130, 791-794

32 Besedovsky, H. O. and Sorkin, E. (1981) in Psychoneuroimmunology (Ader, R., ed.), pp. 545-574, Academic Press, New York

Documents

What do the immune system and the brain know about each other?