Circumscribed lesion of the medial forebrain bundle area ... - Nature

Molecular Psychiatry (1998) 3, 397–404  1998 Stockton Press All rights reserved 1359–4184/98 $12.00

ORIGINAL RESEARCH ARTICLE

Circumscribed lesion of the medial forebrain bundle area causes structural impairment of lymphoid organs and severe depression of immune function in rats KI Pasternak1,2, C Timo-Iaria2, CJ Rodrigues2, DA Maria2, AJS Duarte2, L Paiva2, DH Pozzi2, BB de Mendonça2, M-L Wong1 and J Licinio1 1 Clinical Neuroendocrinology Branch, Intramural Research program, National Institute of Mental Health, NIH, Bethesda, MD 20892-1284, USA; 2Faculdade de Medicina da Universidade de Saõ Paulo, 01246-903 Saõ Paulo, SP, Brazil

Interactions between the immune system and the brain are a key element in the pathophysiology of diseases such as multiple sclerosis, neuroAIDS, and Alzheimer’s, which affect large numbers of individuals and are associated with a high social cost. However, the neuroanatomical basis of brain-immune interactions has not been elucidated. We report that in Wistar rats of either sex bilateral electrolytic lesion of the medial forebrain bundle reduces body weight by 28% 7 days after lesioning, and causes widespread infections, aphagia, adypsia, structural damage to the lymphoid organs and heavy depression of T lymphocytes cytotoxicity. The following alterations occur in the immune system after those lesions: the weight of the thymus, spleen and lymphonodes is reduced by 77.9%, 49.1% and 48.4%, respectively. The thymus is atrophied and contains fewer lymphoid cells in the cortex than in the medulla. In the spleen the white pulp is reduced and lymphoid cells from periarteriolar zones and at the chords are almost absent. In lymph nodes cortical small lymphocytes are depleted and primary and secondary nodules and germinal centers all but disappear. Cytotoxicity of lymphocytes is reduced by 86.2% in the thymus, 77.6% in the spleen and 70.2% in lymph nodes. The critical area of lesion is at the medialmost portion of the medial forebrain bundle, at the preoptic area and rostral part of the anterior hypothalamus. We suggest that this area contains neural circuits that are crucial for keeping the structure of lymphoid organs and the functional integrity of the immune system. Keywords: lymphoid organs; cytotoxicity; medial forebrain bundle; hypothalamus; thymus; spleen

Introduction The regulation of immune function is accomplished by a variety of complex mechanisms involving cytokines, cell-to-cell communication, hormones, and neural circuits. Peripheral immune mechanisms have been investigated to a great extent but the neural structures that play a role in such regulation have not been fully identified.1 Brain immune interactions have an important role in the pathophysiology of diseases such as multiple sclerosis, neuroAIDS, and Alzheimer’s, which are highly prevalent and have a profound social cost.2–9 Several studies have demonstrated that the central nervous system participates in the regulation of immune function.10–21 Immune deficiency may result from lesioning the hypothalamus and the reticular formation,22–24 the medial septal nucleus25 and nucleus basalis of Meynert.26 A complex pattern of neural regulation of the immune system was found by Belluardo

Correspondence: Dr K Pasternak, Rua Austria 141, Saõ Paulo, SP 01447, Brazil Received 2 February 1998; revised 6 April 1998; accepted 24 April 1998

et al,19 who observed that electrolytic lesions of individual nuclei in the mouse medial hypothalamus depresses lymphocyte cytotoxic activity, whereas lesions of all the nuclei cause only a slight reduction in lymphocyte cytotoxic activity. These data indicate that there may be a functional interaction between hypothalamic nuclei that produce a balance between activation and deactivation of immune function. To investigate the anatomical basis of neuroimmune interactions, we conducted hypothalamic lesions to determine whether lesions of specific areas of the hypothalamus affect the functioning of the peripheral immune system. We designed the experiments described below in order to accomplish the following: (1) to determine the critical hypothalamic area associated with the immune depression that followed lesioning; (2) to ascertain the effects of those lesions on the size and structure of lymphoid organs and on lymphocyte toxicity; and (3) to verify whether changes in the structure and function of lymphoid organs were secondary to food deprivation or to changes in corticosterone levels.

Hypothalamic lesions and lymphoid organs KI Pasternak et al

398

Materials and methods Adult Wistar rats of either sex, weighing 160–200 g, underwent electrolytic lesion of the lateral portion of preoptic and anterior hypothalamic areas. Animals were anesthetized with ketamine hydrochloride 50 mg kg−1 plus flunitrazepan 10 mg kg−1 (intraperitoneally injected) and fixed to a stereotaxic instrument for bilateral, symmetric lesions with 1 mA current during 5 s by means of a stainless steel semimicroelectrode (50 ␮m in diameter at the tip). The electrode was positioned according to the atlas of Koening and Klippel,27 initially aiming at lesioning from A:5.2 to 7.0, L:1.2 to 2.0, and H:2.0 to 3.0. The critical lesion was circumscribed to a tiny strip within the medial portion of the medial forebrain bundle at the preoptic area and a small strip of the anterior hypothalamus. When the lesioning procedure was completed, the animals received penicillin and were thereafter kept in individual cages. Because lesions caused complete aphagia, rats were intragastrically fed three times a day with a standard laboratory diet food suspended in milk (50% liquid, 50% solid mix). Control and sham-operated rats were also fed the same way. All animals were weighed daily. When infection occurred a combination of penicillin and streptomycin was used. Within 7–10 days of surgery, animals developed infection. Therefore, for our experimental purposes, rats were killed 6 days after lesioning. During autopsy we searched for signs of infection or inflammation. Thymus, spleen and inguinal lymph nodes were removed and their fresh weights were determined, and they were preserved for histologic examination. Brains were removed and processed for the identification of the extent and topography of lesions in 10-␮m, Nisslstained sections. The adrenal glands were also removed and weighed. In the control group lesions were made in the parietal cortex. In the sham-operated rats the semimicroelectrode was introduced in the brain but no current was passed; the electrode was positioned one millimeter above the critical area, because the target area was very small and its tunneling by the electrode could produce an effective lesion, which we had found to occur in preliminary experiments. An additional control group consisted of animals deprived of food and water for 6 days. Corticosterone levels were measured to determine if changes in immune function were associated with alterations in the levels of this stress-responsive hormone. Immune function tests Five groups were used to assess lymphocyte cytotoxicity function: normal (N), control lesion (C), sham lesion (S), medial forebrain bundle area lesion (L) and a control group, consisting of non-lesioned animals that were fasted for 7 days (F). Three days before being killed, the animals were intravenously injected with sheep red blood cells (SRBC) and treated according to the technique described by Miller et al.28 A 2.108 SRBC suspension in 199 medium supplemented with 2 nM

Hepes (Riedel de Haen AG D-3016 Seelze 1, Germany), 1% l-glutamine, and 5% inactivated bovine fetal serum at pH 7.4 was used to sensitize the animals. In the control group a sterilized saline solution was injected intravenously. Thymus, spleen and lymph nodes were removed for the lymphocyte cytotoxicity test. The lymphoid organs were cut into pieces, washed and placed in a Petri dish containing 199 medium added with 2 nM Hepes, 4 nM NaHCO3 (Mallinkrodt, USA) and 5.10−5 M 2 mercaptoethanol and thereafter passed through nylon wool columns (nylon fiber 38.1 mm type 200) for separation of adhering cells.29 One hour after incubation the material was centrifuged (1800 rpm) at room temperature for 10 min. Subsequently, non-adherent cells, T-lymphocytes, were resuspended in 199 medium supplemented with 2 nM Hepes, 4 nM NaHCO3 and 5 × 10−5 M mercaptoethanol and used as effector cells.29 Corticosterone levels Corticosterone was measured by radioimmunoassay (RIA). In deeply anesthetized animals, blood was collected by cardiac puncture in heparin 40 min after the anesthetic had been injected, to reduce stress to a minimum. Anticorticosterone antibodies (from Radioassay Systems Laboratories) and a tritiated (3H) antigen were used in the RIA. Histology Brains of the lesioned and sham-lesioned animals were also removed for Nissl-staining to assess the lesioned areas in the L group and electrode position in S. Sections of the lymphoid organs were stained with hematoxylin-eosine for histological examination. Lymphocyte mediated cytotoxicity (LMC) T cells were incubated in 96 wells in round-bottomed microtiter plates with ×105 51Cr-labeled SRBC,30 resulting in an E/T ratio of 50:1 in a total volume of 0.2 ml RPMI/FCS. Each culture was performed in triplicate. The plates were incubated for 18 h at 37°C. The cultures were then centrifuged for 1 min and 50 ␮l of the supernatant from each well were removed and the 51 Cr release was measured in a gamma counter. Spontaneous release of 51Cr was determined from target cells alone and maximum release was obtained by the addition of distilled water to the 51Cr-SRBC. Cytotoxicity was calculated as the relative (percentual) specific lysis (cpm of the test group as related to cpm of spontaneous release), according to the following formula:31–34 specific lysis = experimental cpm − spontaneous cpm × 100. maximum cpm − spontaneous cpm The percentage of maximum cpm relative to spontaneous cpm was over 90% in control experiments.


weight loss in the infected group (28.0%). When lesions were restricted to the medial strip, some animals were able to eat and drink a little but loss of weight was similar to that of the animals with complete lesion of the medial forebrain bundle. A slight loss of body weight also occurred in control (C) and in sham-operated (S) rats, as compared to normals (N) (Table 1), probably due to the surgical procedures. There was no significant difference between the weight of the C and S groups but both were different (P ⬍ 0.01) from the lesioned (L) group.

Figure 1 A rat subjected to lesion of the medial forebrain bundle (below) is compared to a normal rat (above). Both had a similar weight at the beginning of the experimental period. Notice the piloerection and cachectic aspect of the lesioned animal.

Statistical analysis The significance of differences between groups was assessed using analysis of variance (ANOVA) followed by post hoc correction (Duncan test). Statistical significance was assumed for P ⬍ 0.05.

Results Body weight The animals in which the entire medial forebrain area was bilaterally lesioned (which included the critical area) soon exhibited the lateral hypothalamic syndrome. They adopted a curled posture, walked little and slowly and became apathetic, reacting weakly and only to strong sensory stimuli. Most animals did not eat or drink spontaneously. Loss of weight evolved rapidly towards cachexia within 5 or 6 days (Figure 1), when weight had dropped on the average from 194.42 ± 2.66 g (mean ± SEM) before operation to 139.93 ± 2.61 g (Table 1), which amounts to a reduction of 28% (P ⬍ 0.01). Loss of body weight could have been a consequence of infection but when the latter was prevented by antibiotics, no significant difference was found in weight between the groups with and without infection. In non-infected rats the relative loss of weight was 28.1%, therefore not different from the

Lymphoid organs Thymus, spleen and lymph nodes’ weights were compared among the three experimental groups, including L animals in which infection was prevented. Table 2 shows that in L rats thymus weight was sharply reduced (by 77.9%) as compared to the normal group (P ⬍ 0.001); in L, thymus weight was also significantly lower than in C and S groups (P ⬍ 0.01). Among C and S animals thymus weight was similar but it was lower than in N (P ⬍ 0.01); weight loss was, respectively, 20.3% and 21.2% lower in C and S than in N. Spleen weight was only affected in L rats, in which there was a weight loss of 49.1% in relation to group N (P ⬍ 0.01 in comparison to N, C and S). Lymph node weight was equally reduced (48.4% as related to N) in the rats with a lesion in the medial forebrain bundle area. The role of infection in weight loss was assessed by comparing the effects of lesioning of the medial forebrain bundle area among animals that were treated with antibiotics and those that were not. As shown in Table 3, infection was not the cause of weight loss, because its prevention did not change the proportion of either body or lymphoid organs’ weight reduction after lesioning. Lymphoid organs were structurally affected in the medial forebrain bundle lesioned animals but not in the other groups. Thymus was clearly atrophied, yellowish and hardened. The vascular connective tissue surrounding the lobules increased, subdividing the organ into many small lobules which communicated with medullary vessels. Figure 2 illustrates the deep structural changes undergone by the thymus in L animals as compared to normals. Demarcation between cortex and medulla in each lobule was poor and, contrary to what is normally found, medulla was more

Table 1 Body weight (in grams, mean ± SEM) just before (Wb) the experimental period began and when the animals were sacrificed (Ws)

Wb Ws

N

L

C

S

194.53 ± 3.44 (n = 13) 194.55 ± 3.41 (n = 13)

194.42 ± 2.66 (n = 31) 139.93 ± 2.61a (n = 31)

204.61 ± 3.53 (n = 31) 197.81 ± 3.49 (n = 31)

218.19 ± 5.39 (n = 21) 211.95 ± 5.80 (n = 21)

a Significantly different (P ⬍ 0.01) from Wb N (normal group) and from Ws L (lesioned), C (control) and S (sham-operated) rats. n = number of animals.

399


400

Table 2

Organ weight (in grams ± SEM) at the day of sacrifice

Organ

Spleen Thymus Lymph nodes Adrenal glands

Group N

L

C

S

0.914 ± 0.077 (n = 19) 0.217 ± 0.008 (n = 19) 0.031 ± 0.001 (n = 19) 0.038 ± 0.001 (n = 13)

0.466 ± 0.023a (n = 40) 0.048 ± 0.001a (n = 40) 0.016 ± 0.001a (n = 40) 0.043 ± 0.001 (n = 31)

0.864 ± 0.032 (n = 39) 0.173 ± 0.008 (n = 39) 0.034 ± 0.001 (n = 39) 0.041 ± 0.001 (n = 31)

0.911 ± 0.054 (n = 27) 0.171 ± 0.010 (n = 27) 0.032 ± 0.001 (n = 27) 0.034 ± 0.001a (n = 21)

Spleen, thymus and lymph nodes; asignificantly different (P ⬍ 0.01) from N, C and S. Adrenal glands: asignificantly different (P ⬍ 0.01) from N, L and C. Table 3 Body and organ weights (in grams, mean ± SEM) of rats with and without infection after lesioning

Initial weight (body) Final weight (body) Spleen Thymus Lymph nodes Adrenal glands

Infected (n = 16)

Non-infected (n = 22)

202.37 ± 6.84

195.90 ± 3.33

143.50 ± 4.58

140.81 ± 3.09

0.512 ± 0.051 0.042 ± 0.002 0.018 ± 0.001 0.041 ± 0.002

0.472 ± 0.035 0.047 ± 0.002 0.015 ± 0.008 0.043 ± 0.001

deeply stained than cortex. The latter was drastically reduced and small lymphoid cells depleted, being occupied by large cells with abundant and clear cytoplasm; a few polymorphonuclear cells and mastocytes were located in scattered sites within the outermost layer adjacent to connective tissue. Medulla showed relatively more lymphoid cells than the cortex, thus generating the inverted overall picture of the gland. In some lesioned rats the thymus presented a striking structural disruption. The spleen gross structure was fairly preserved but reduction of the white pulp was patent. Lymphoid cells from periarteriolar zones and the cords all but disappeared (Figure 3). Lymph nodes from L rats were depleted from small lymphocytes in the cortical area (Figure 4). Primary and secondary nodules disappeared and the germinal centers were almost absent. A moderate increase of plasma cells was seen in the medullary cords. A thin rim of secondary cortex was observed in several lymph nodes. Adrenal gland weight The weight of the adrenal gland was not affected in any of the experimental groups, except for the shamoperated rats, in which there was a 10.5% reduction in comparison to the normal group (P ⬍ 0.01). Corticos-

Figure 2 Microscopic structure of the thymus of a medial forebrain bundle lesioned rat (a) as compared to a normal organ (b). In (b), a normal Hassal body; the characteristic conspicuous lobules with a pale medullary core surrounded by densely packed lymphocytes can be seen. In (a), there appears a corticomedullary inversion, due to the intense cortical lymphocyte rarefaction. HE, × 170.


401

Figure 3 Microscopic structure of the spleen from a medial forebrain bundle lesioned (a) and from a normal animal (b). Notice in (b) the well-developed Malpighian bodies and in (a) the scarceness of lymphocytes that in (b) surrounds the central arteriole. HE, × 170 in (a) and × 80 in (b).

terone levels at the end of the experiments were statistically similar (P ⬎ 0.05) among all the experimental groups (Table 4). Cytotoxicity Table 5 summarizes data obtained from the cytotoxicity tests. Thymus lymphocyte cytotoxicity (expressed as mean ± SEM of the percentage of 51Cr released by the experimental vs control groups) was reduced by the critical lesion to only 13.8% of the normal value (P ⬍ 0.01), falling from 31.0 ± 3.6 in normal rats to 4.4 ± 1.2 in the L group. Among C and S rats no difference was found and neither group differed from normals. Spleen lymphocyte cytotoxicity fell to 22.4% of the normal value, decaying from 30.4 ± 3.8 in N rats down to 6.8 ± 2.2 in the lesioned rats (P ⬍ 0.01). Lymph node lymphocyte cytotoxicity was reduced in the L group to 29.8% as compared to N rats, falling from 27.6 ± 3.8 to 8.2 ± 2.5 (P ⬍ 0.01). Among control and sham-operated rats lymph node lymphocyte cytotoxicity was similar and did not differ from the N group. In the animals starved for 7 days lymphocyte cyto-

Figure 4 Microscopic structure of a lymph node from a medial forebrain bundle lesioned rat (a) and from a normal animal (b). The dense cortical follicules, patent in (b), are absent in )a). HE, × 120. Table 4 Corticosterone levels at the day of sacrifice (in ␮g per 100 ml of plasma, mean ± SEM) N 45.58 ± 7.79 (n = 13)

L

C

S

51.90 ± 5.66 (n = 23)

45.00 ± 5.24 (n = 27)

45.45 ± 6.42 (n = 11)

toxicity increased significantly (P ⬍ 0.05) in the thymus (49.8 ± 11.5) and spleen (P ⬍ 0.05, 47.5 ± 6.4) but was not affected in lymph nodes. Topography of the lesion Impairment of lymphoid organs’ structure and function as described above occurred following destruction of a thin strip of the medial forebrain bundle area at the preoptic and anteriormost hypothalamic regions. Figure 5a illustrates the site of the large lesion, at the level of the preoptic region, and Figure 5b shows a small lesion confined to a thin portion of the medial border of the medial forebrain bundle area. Immune


402

Table 5 Cytotoxicity (in % of 51Cr released by the sample in comparison to release by the control, mean ± SEM) of lymphocytes from lymphoid organs in the four experimental groups Organ

Thymus Spleen Lymph nodes

Group N

L

C

S

31.0 ± 3.6 (n = 10) 30.4 ± 3.8 (n = 10) 27.6 ± 3.8 (n = 10)

4.4 ± 1.2a (n = 12) 6.8 ± 2.2a (n = 12) 8.2 ± 2.5a (n = 12)

26.1 ± 3.3 (n = 10) 30.5 ± 5.5 (n = 10) 35.7 ± 4.5 (n = 10)

24.2 ± 3.0 (n = 8) 32.0 ± 3.4 (n = 8) 24.3 ± 3.1 (n = 8)

Significantly different (P ⬍ 0.01) from N, C and S.

a

depression occurred only when the lesions were bilateral.

Discussion Bilateral electrolytic lesion of a thin strip of the medial forebrain bundle area at the preoptic and anteriormost hypothalamic regions caused body weight loss, thymus, spleen and lymph nodes hypotrophy and reduction of lymphocyte cytotoxic activity. The literature on the effects of brain lesions on the immune system in general has been reviewed in detail by Cross.35 In our studies, we observed reduction of body weight by about one third in 1 week, despite intragastric feeding and hydration, which maintained the animals adequately fed and hydrated. When lesion of the medial forebrain bundle was completed at the preoptic and anterior hypothalamic levels, the animals could not eat or drink at all but when only its medialmost portion was destroyed they occasionally ate and drank. Depression of cytotoxic T cells activity, although more pronounced when the entire medial forebrain bundle area was destroyed, was severe in either case. Recovery from the failure to eat provoked by lateral hypothalamic lesions was first reported by Teitelbaum and Stellar.36 They observed that the crucial lesion that induces aphagia and adypsia in the rat is located 1 mm above the floor of the brain and 2 mm off the midline on each side, extending all along parallel to the ventromedial nuclei. Left to eat spontaneously all the animals died within a few weeks but no catastrophic infections have been ever described in studies of the so-called lateral hypothalamic syndrome. If properly fed, rats bearing lesions in that region will start eating by themselves in about 10 days, mainly if offered food with a high fat content, and all the animals do eventually resume eating of ordinary food and water.37 However, in preliminary experiments in which the entire medial forebrain bundle area had been destroyed, we found that all rats died within less than 15 days after lesioning. In our experiments, recovery of eating and drinking after whole medial forebrain bundle lesion at the preoptic and anterior hypothalamic levels never occurred, presumably due to the early death of the experimental animals; when the lesions were restricted to the

medialmost part of this brain region, ability of the animals to eat and drink was partially recovered but the alterations in immune function were not. Rostrally located lesions in the anteroposterior regions caused alterations in immune function; those lesions also caused aphagia and adypsia. There is evidence that in the rat this region lodges glucoreceptors sensitive to cytoglucopenia, whose lesion may impair glucose metabolism and might therefore induce abnormalities in feeding behavior.38 Morgane39,40 reported aphagia, that did not evolve to spontaneous feeding, after prolonged instrumental feeding and inexorable death in rats subjected to localized lesion of the far-lateral region of the lateral hypothalamus, just medial to the internal capsule; however, no changes in immune function were reported in those studies. All animals were adequately fed and hydrated throughout the study; therefore, it is unlikely that either starvation or dehydration induced alterations in immune function in our experiments. We also documented that complete food and water deprivation of normal rats for 6 days caused a weight loss that was parallel to but was less prominent than that which occurred in the lesioned rats; nevertheless, it did not provoke changes in the structure or function of lymphoid organs. The severe impairment of structure and function of the lymphoid organs that we observed in this study is evidence of interruption of a normal mechanism that directly maintains their trophism. Thymus involution has been previously reported after central lesion41 but the type and intensity of the structural changes we found in this organ revealed a highly disrupting effect of the lesion. The alterations in spleen and lymph nodes we detected, however, are similar to those reported by Isakovic and Jankovic.41 The hypothalamic area they lesioned was more medially located as compared to our critical lesion. Acute thymus involution has been found to occur under increased endogenous secretion or administration of androgenic, estrogenic and adrenocortical steroid hormones, stress and starvation.42 The deeper impairment of thymus in our study, in comparison to spleen and lymph nodes, is expressed by its structural loss, which appears as an almost complete destruction


Figure 5 Schematic representation of the lesions which provoked immune depression in rats. (a) A larger lesion, embodying almost entirely the medial forebrain bundle area at the preoptic level (in black). This lesion includes portions of the medial forebrain bundle about 0.3 mm rostral and caudal to this level (not shown). (b) A smaller lesion at the medialmost portion of the medial forebrain bundle area (in black), mostly at the suprachiasmatic level, around 0.3 mm wide and deep and slightly less than 1.0 mm along the vertical axis. Abbreviations: ac, anterior commissure; amy, amygdala; c, cingulum; cc, corpus callosum; cg, cingulate gyrus; cp, caudate-putamen; f, fornix; gp, globuls pallidus; Ipo, lateral preoptic nucleus; mfb, medial forebrain bundle; mpo, medial preoptic nucleus; oc, optic chiasm; pm, magnocellular preoptic nucleus; sf, fimbrial spetalnucleus; sl, lateral septal nucleus; st, interstitial nucleus of stria terminalis.

Spleen and lymph node weight was reduced in L rats by approximately the same amount (near 50%) in comparison to normal rats, which is much less than weight reduction of the thymus (86.2% as related to normal animals). The medial forebrain bundle is not a single fiber tract but a mesh of autochthonous neurons, that entertain complex connections among themselves and with distant neurons, and long projection fibers that course in both ascending and descending directions.43 The local neurons originate and/or mediate many behavioral and homeostatic functions (such as feeding and sexual behavior, sleep, alertness, blood glucose regulation), requiring the activation of several distributed systems. This pattern of organization probably contributes to the impairment of immune regulation by the lesioning of several structures that are connected with the medial forebrain bundle (septal area, hippocampus, frontal cortex, nucleus of Meynert). In fact, the medial forebrain bundle is a central link in an intricate network, which involves these structures and other, hypothalamic and preoptic connections, related to the control of secretion of the pituitary hormones. Several pituitary hormones and endogenous peptides do either inhibit (adrenocorticotropic hormone, ␣-endorphin, leu- and met-enkephalin) or enhance (thyrotropic and growth hormone) immune function. The medial forebrain bundle could potentially influence the immune system through its connections with the preoptic and hypothalamic nuclei involved in pituitary function, particularly hormones regulating hypothalamic-pituitaryadrenal (HPA) axis function.44,45 However, corticosterone levels were not abnormal in our study. Further studies are needed to ascertain the mechanisms by which the medialmost portion of the medial forebrain bundle at the preoptic and rostral anterior hypothalamic areas contribute to regulate immune function. In conclusion, our experiments suggest the existence, in the medialmost portion of the medial forebrain bundle at the preoptic and rostral anterior hypothalamic areas, of neural circuits that are essential for the maintenance of the structure and function of lymphoid organs. Acknowledgements The authors are indebted to Miss Neusi O Rolim for technical assistance and to Dr Jacyr Pasternak for his valuable suggestions. This research has been supported by Fundaçaõ Faculdade de Medicina, LIM-HC-FMUSP, FAPESP and CNPq. M-L Wong is the recipient of a Young Investigator Award from NARSAD.

References of the cortex, and weight reduction to only about one fifth of that prevailing in the normal organ. Thymus lymphocyte cytotoxicity was reduced to 13.8% of the normal value. Whether atrophy of the organ and reduction of cytotoxicity in thymus T lymphocytes are independent variables or the latter depends upon the former remains an open question.

1 Jankovic BD. Neuroimmunomodulation: facts and dilemmas. Immunol Lett 1989; 21: 101–118. 2 Cacabelos R, Alvarez XA, Fernandez-Novoa L et al. Brain interleukin-1 beta in Alzheimer’s disease and vascular dementia. Meth Find Exp Clin Pharmacol 1994; 16: 141–151. 3 Pachter JS. Inflammatory mechanisms in Alzheimer disease: the role of beta-amyloid/glial interactions. Mol Psychiatry 1997; 2: 91–95.

403


404

4 Vitkovic L, da Cunha A, Tyor WR. Cytokine expression and pathogenesis in AIDS brain. Res Publ Assoc Res Nerv Ment Dis 1994; 72: 203–222. 5 Chiang CS, Powell HC, Gold LH, Samimi A, Campbell IL. Macrophage/microglial-mediated primary demyelination and motor disease induced by the central nervous system production of interleukin-3 in transgenic mice. J Clin Invest 1996; 97: 1512–1524. 6 Issazadeh S, Lorentzen JC, Mustafa MI, Hojeberg B, Mussener A, Olsson T. Cytokines in relapsing experimental autoimmune encephalomyelitis in DA rats: persistent mRNA expression of proinflammatory cytokines and absent expression of interleukin-10 and transforming growth factor-beta. J Neuroimmunol 1996; 69: 103– 115. 7 Michelson D, Stone L, Galliven E et al. Multiple sclerosis is associated with alterations in hypothalamic-pituitary-adrenal axis function. J Clin Endocrinol Metabol 1994; 79: 848–853. 8 Sippy BD, Hofman FM, Wallach D, Hinton DR. Increased expression of tumor necrosis factor-alpha receptors in the brains of patients with AIDS. J Acquir Immune Defic Syndr Hum Retrovirol 1995; 10: 511–521. 9 Rudick RA, Ransohoff RM, Peppler R, VanderBrug MS, Lehmann P, Alam J. Interferon beta induces interleukin-10 expression: relevance to multiple sclerosis. Ann Neurol 1996; 40: 618–627. 10 Ader R, Cohen N. CNS-immune system interactions. Conditioning phenomena. Behav Brain Sci 1985; 8: 379–394. 11 Barneóud P, Neveu PJ, Vitiello S, Morme`de P, Le Moal M. Brain neocortex immunomodulation in rats. Brain Res 1988; 474: 394– 398. 12 Felten DL, Felten SY, Carlson SL, Olschowska JA, Livnat S. Noradrenergic and peptidergic innervation of lymphoid tissue. J Immunol 1985; 135: 755–765. 13 Williams JM, Peterson RG, Shea PA, Schmedtje JF, Bauer DC, Felten DL. Sympathetic innervation of murine thymus and spleen: evidence for a functional link between the nervous and immune systems. Brain Res Bull 1981; 6: 83–94. 14 Felten DL, Ackerman KD, Wiegand SJ, Felten SY. Nerve fibers associate with lymphocytes and macrophages in specific compartments of the splenic white pulp. J Neurosci Res 1987; 18: 28–36. 15 Felten SY, Olschowska J. Noradrenergic sympathetic innervation of the spleen. II. Tyrosine hydroxylase (TH)-positive nerve terminals form synapticlike contacts on lymphocytes in the splenic white pulp. J Neurosci Res 1987; 18: 37–48. 16 Felten SY, Bellinger DL, Collier TJ, Coleman PD, Felten DL. Decreased sympathetic innervation of spleen in aged Fischer 344 rats. Neurobiol Aging 1987; 8: 159–165. 17 Felten DL, Felten SY, Bellinger DL et al. Noradrenergic sympathetic neural interaction with the immune system: structure and function. Immun Rev 1987; 100: 225–260. 18 Carlson SL, Felten DL. Involvement of hypothalamic and limbic structures in neural-immune communication. In: Goetzl EJ, Spector NH (eds). Neuroimmune Networks: Physiology and Diseases. Allan R Liss: New York, 1989, pp 219–226. 19 Belluardo N, Mudo´ G, Cella S, Santoni A, Forni G, Bindoni M. Hypothalamic control of certain aspects of natural immunity in the mouse. Immunology 1987; 62: 321–327. 20 Neveu PJ. Minireview: cerebral neocortex modulation of immune functions. Life Sci 1988 42: 1917–1923. 21 Roszman TL, Brooks WH. Neural modulation of immune function. J Neuroimmunol 1985; 10: 59–69. 22 Tyrey L, Nalbandov AC. Influence of anterior hypothalamic lesions on circulating antibody titers in the rat. Am J Physiol 1992; 222: 179–185.

23 Jankovic BD. Structural correlates of immune microenvironment. In: Jankovic BD, Jankovic K (eds). Microenvironmental Aspects of Immunity. Plenum Press: New York, 1971, pp 1–4. 24 Jankovic BD, Isakovic K. Neuro-endocrine correlates of immune response. I. Effects of brain lesions on antibody production, Arthus reactivity and delayed hypersensitivity in the rat. Arch Allergy 1973; 45: 360–372. 25 Labeur M, Nahmod VE, Finkielman S, Arzt E. Lesions of the septal nucleus produce a long-lasting inhibition of T. lymphocyte proliferation. Neurosci Lett 1991; 125: 129–132. 26 Radosevic-Stasic B, Cuk M, Mrakovcic-Sutic Z et al. Immunological consequences of lesions of nucleus basalis in rats. Int J Neurosci 1990; 60: 325–327. 27 Koenig JFR, Klippel RA. The Rat Brain. A Stereotaxic Atlas of the Forebrain and Lower Parts of the Brain Stem. Williams & Wilkins: Baltimore, 1963. 28 Miller TE, MacKanes GB, Lagrange L. immunopotentiation with BCG. II. Modulation of the response to sheep blood cells. J Natl Cancer Inst 1973; 51: 1669–1676. 29 Hudson L, Hay FC. Practical Immunology. Third edn. Blackwell: Oxford, 1989. 30 Kay HD, Bonnard GD, West WH, Herberman RB. A functional comparison of human Fc-receptor-bearing lymphocytes active in natural cytotoxicity and antibody-dependent cellular cytotoxicity. J Immunol 1977; 118: 2058–2066. 31 Herberman RB, Holden HT. Natural-cell mediated immunity. Adv Cancer Res 1978; 27: 305–377. 32 Herberman RB, Reynolds CW, Ortaldo JR. Mechanism of cytotoxicity by natural killer. Annu Rev Immunol 1986; 4: 651–680. 33 Cerottini JC. Lymphoid cells as effectors of immunologic cytolysis. Hosp Pract 1977; 12: 57–68. 34 Calder EA, Penhale WJ, Barnes EW, Irvine WJ. Cytotoxic lymphocytes in Hashimoto thyroiditi. An ‘in vivo’ assay system using 51Crlabeled chicken red blood cells coated with thyroglobulin. Clin Exp Immunol 1973; 14: 19–27. 35 Cross RJ. Hypothalamic-immune interactions: modulation of immune function by electrolytic and chemical lesions in the central nervous system. In: Frier S (ed). The Neuroendocrine Immune Network. CRC Press: Boca Raton, Florida, 1990. 36 Teitelbaum P, Stellar E. Recovery from the failure to eat produced by hypothalamic lesions. Science 1954; 120: 894–895. 37 Teitelbaum P, Epstein AN. The lateral hypothalamic syndrome: recovery of feeding and drinking after lateral hypothalamic lesions. Psychol Rev 1962; 69: 74–90. 38 Timo-Iaria C. Glucoreceptor systems: from control of glycemia to feeding behavior. News Physiol Sci 1990; 5: 46–49. 39 Morgane PJ. Medial forebrain bundle and ‘feeding centers’ of the hypothalamus. J Comp Neurol 1962; 117: 21–25. 40 Morgane PJ. Alterations in feeding and drinking behavior of rats with lesions of globi pallidi. Am J Physiol 1961; 201: 420–428. 41 Isakovic K, Jankovic BD. Neuro-endocrine correlates of immune response. II. Chanes in the lymphatic organs of brain-lesioned rats. Arch Allergy 1973; 45: 373–384. 42 Dougherty TF. Effects of hormones on lymphatic tissue. Physiol Rev 1952; 32: 379–401. 43 Millhouse OE. A Golgi study of the descending medial forebrain bundle. Brain Res 1969; 15: 341–363. 44 Sternberg EM, Chrousos GP, Wilder RL, Gold PW. The stress response and the regulation of inflammatory disease. Ann Intern Med 1992; 117: 854–866. 45 Sternberg EM, Licinio J. Overview of neuroimmune stress interactions. Implications for susceptibility to inflammatory disease. Ann N Y Acad Sci 1995; 771: 364–371.