A Molecular Basis for Bidirectional : Communication Between the Immune and Neuroendocrine Systems
J. EDWIN BLALOCK Department of Physiology and Biophysics, University of Alabama at Birmingham, Birmingham, Alabama
PRODUCTION OF LYMPHOKINES AND MONOKINES BY THE NEUROENDOCRINE SYSTEM
If bidirectional communication between the immune and neuroendocrine systems occurs as a result of shared signal molecules, then one might expect that not only should the immune system produce peptides that were previously restricted to the neuroendocrine system but that the converse should occur. That is, the neuroendocrine system should contain peptides and proteins that are classically associated with the immune system, i.e., lympho-kines and monokines. Although studies to test this idea are in their infancy, they nonetheless strongly point to such a situation. One of the first lympho-kines and/or monokines shown to be produced by the neuroendocrine system was IL 1. The synthesis of this molecule has been reported to occur in astro-cytes, glial cells, and in the brain of endotoxin-treated mice (51,52). Another lymphokine/monokine shown to be produced by neuroendocrine tissue (as-trocytes) was an IL 3-like factor. On the introduction of LPS into the media of cultured astrocytes, it was shown that the supernate contained a factor of a molecular weight of 30,000 that caused the growth of IL 3-dependent cells (53). More recently, it has been shown that glial maturation factor (GMF), produced by astrocytes, resembles IL 1 in its ability to induce the proliferation of fibroblasts, to synergize with other growth factors, and to induce glial cell differentiation (97).
Thymosin peptides are also present in the central nervous system. Discrete subcortical regions of the rat brain were found to contain immunologi-cally detectable thymosin ax. The highest concentrations were found in the arcuate nucleus and in the median eminence while the lowest were in the angulate cortex and striatum (62). Thymosin /?4 has also been identified in a number of tissues, including the brain (63). Although it has not yet been systematically evaluated, it is probably a safe assumption that IFNs are produced at a number of sites in the neuroendocrine system. This notion derives from the observation that virtually all nucleated cells can produce at least one type of IFN (134).
Thus it appears that the neuroendocrine system, like the immune system, can produce on stimulation hormones (lymphokines/monokines) that have both neuroendocrine and immune functions. The sharing of ligands between the two systems seems to provide a biochemical basis by which both systems can communicate with one another. However, for this line of communication to be complete, it requires the presence of receptors by which these ligands can interact with and transmit their messages.
PEPTIDE HORMONE RECEPTORS COMMON TO THE IMMUNE AND NEUROENDOCRINE SYSTEMS
Cells of the immune system have been shown to harbor specific binding sites for many neuroendocrine hormones (Table 5). Mouse spleen mononu-clear cells have both high- and low-affinity receptors for ACTH [dissociation constant (Kd) = 0.1 and 4.8 nm, respectively] with ~ 3,000 high- and 50,000 low-affinity sites per cell (73). Interestingly, ACTH receptors on rat adrenal cells possessed similar high- and low-affinity binding sites (Kd = 0.25 and
table 5. Neuroendocrine peptide receptors in the immune system
|Receptor||Cellular or Tissue Source||KifnU||References|
Human peripheral lymphocytes
|0.1, 4.1 0.04, 3.4||73 127|
|0-Endorphin (nonopiate)||Lymphoblastoid cells||3.0||68|
|Enkephalin (opiate)||Murine spleen||0.59||73|
|Substance P||Lymphoblastoid cells||0.69||110|
See text for definitions of abbreviations.
10.0 nm, respectively) with 3,000 high- and 30,000 low-affinity sites per cell (93). More recently, the mouse BCLi has been shown to have ACTH receptors (26). Human peripheral blood mononuclear leukocytes are not particularly unlike mouse splenocytes in that they also have both high- and low-affinity receptors for ACTH (Kd = 0.04 and 3.4 nm, respectively) with ~ 3,000 high-and 45,000 low-affinity sites per cell (127).
Leukocytes have also been shown to possess receptors for other neuroendocrine peptides. Opiate-binding sites were initially and indirectly demonstrated by showing that morphine and Met-enkephalin inhibited and enhanced, respectively, T-cell rosetting of SRBC (145). This inhibition and enhancement could be reversed by naloxone, whereas an inactive enantiomer, levomoramide, had no effect on rosetting. By competitive binding experiments with radiolabeled ligand, Met-enkephalin receptors with a single high-affinity binding site of ~0.59 nM were directly shown on mouse splenocytes (73). Stereospecific dihydromorphine binding to human phagocytic leukocytes was also observed (88). Both granulocytes and monocytes had specific receptors with apparent iJTds of 10 and 8 nM, respectively. Hazum et al. (68) reported a high-affinity /?-endorphin-binding site on transformed human lymphocytes with an apparent Kd of 3 nM plus a lower-affinity site (68). The authors felt these were not classic opiate receptors, since the binding was not affected by opiate agonists and antagonists. Because binding to opiate receptors classically is through the NH2-terminal end of the peptide, this nonopiate receptor binding seems to be through the COOH-terminus of i8-endorphin (68). This leads to the interesting possibility that a molecule such as jS-endorphin could bridge two subtypes of lymphocytes by binding through its NH2-terminus to an opiate receptor on one cell and through its COOH-terminus to a nonopiate receptor on another lymphocyte. Other specific binding sites that are present on cells of the immune system are listed in Table 5. These include, but are not limited to, neurotensin (6), substance P (110), GH (78), VIP (42,107), and insulin (83). Interestingly, the VIP receptor on lymphocytes has been shown to be coupled to the adenylate cyclase system (103).
Receptors and their binding sites can be defined by at least two methods, pharmacological or biochemical. Pharmacological techniques, as were employed in most of the aforementioned reports, are very limited with respect to the study of receptor structure. For example, an accurate molecular weight of the receptor cannot be determined nor can its subunit composition be defined (if applicable). Moreover, if the receptor is multimeric, chains not responsible for recognizing and binding the given ligand would go unnoticed. Thus, by employing pharmacological techniques to compare “receptors” from two different tissues, the results may indicate similar affinities and specificities for a given ligand or ligands without necessarily defining the structures to be identical. Thus, for the above reasons, biochemical methodologies have been employed to compare receptors from the immune and neuroendocrine systems. To this end, the molecular characteristics of ACTH receptors have recently been explored in our laboratory using a new technology (25). The ACTH receptors were immunoaffinity purified from mouse Y-l adrenal cells, and the molecular weight and subunit structure were determined using polyacrylamide gel electrophoresis. To summarize our results, the total molecular weight of the adrenal ACTH receptor was 225,000, and it was composed of four polypeptide chains of 83,000, 64,000, 52,000, and 22,000. The 83,000 and 52,000 chains were shown to be disulfide linked, with the 83,000 chain containing the ACTH-binding site (24). When ACTH receptors were isolated from mouse splenic or human peripheral blood mononuclear cells, similar if not identical molecular weights were found for the whole receptor and the four subunits. Thus, at a molecular level, ACTH receptors from adrenal and immune cells seem to be identical. The idea that the ACTH receptor on adrenal and lymphoid cells might be identical gene products seems to have been recently confirmed by an “experiment of nature.” Specifically, it was shown that an individual who apparently lacked a functional adrenal ACTH receptor also did not possess an ACTH-binding site on their peripheral mononuclear leukocytes (127).
Recently, a similar apprbach to that for the ACTH receptor was employed for the opiate receptor(s) on neural and lymphoid tissues (34, 35). In this system, the opiate receptor complex from both mouse splenocytes and brain was composed of four polypeptide chains of molecular weights of 70,000, 56,000, 46,000, and 31,000 with the 56,000 chain containing the opiate-binding site (34). In addition to having an apparently identical subunit structure, the opiate receptors from neurons and splenocytes are apparently coupled to similar secondary messenger systems. That is, binding of opiates to the receptor on splenocytes leads to both an inhibition of adenylate cyclase and the closing of a potassium channel. Thus, in terms of activation of an opiate receptor, perhaps the only readily apparent difference between neural and lymphoid cells is the final readout system, i.e., postsynaptic vs. im-munologic.
BIOCHEMICAL INTEGRATION OF NEUROENDOCRINE AND IMMUNE SYSTEM
CIRCUITRY BY COMMON RECEPTORS AND LIGANDS
Collectively, all of the above data and observations would seem to point to a biochemical basis for bidirectional communication between the immune and neuroendocrine systems. Put most simply, these two systems contain and use the same set of signal molecules in the form of hormones, lymphokines, and monokines for inter- and intrasystem communication and regulation. Furthermore, they harbor the same array of receptors for the shared ligands. Thus, in retrospect, it would seem virtually impossible for there not to be “cross talk” between the immune and neuroendocrine systems. Furthermore, with time, it seems that the two systems will be further interwoven with respect to shared ligands and receptors. One example of this is the recent demonstration that pituitary cells have an IL 2 receptor that, similar to that on T lymphocytes, can be upregulated by IL 2 (132).
IN VIVO IMMUNOREGULATION BY THE NEUROENDOCRINE SYSTEM
At this point, it seems worthwhile to note a few examples that suggest that there are possible in vivo correlates of the previous in vitro studies. This section is not intended to be an exhaustive review but merely to point out a few avenues of in vivo research that may be related. For instance, Jankovic and Isakovic (72) showed that electrical stimulation in the hypothalamus enhanced the Arthus reaction (an in vivo measure of immediate hypersensi-tivity). This group of investigators additionally determined that lesions of the anterior hypothalamus had effects on cellular immunity by showing impairments of T-cell-mediated delayed-type hypersensitivity and graft versus host reaction. Stein and co-workers (136) found that lesions in the anterior hypothalamus but not in the median or posterior hypothalamus yielded significant protection against lethal anaphylaxis after challenge antigen, and this same protection was not seen in sham-operated or nonoper-ated animals.
Roszman and co-workers (for review, see Ref. 118) extended the lesioning studies to include not only the anterior hypothalamus but other regions of the brain and limbic system, such as the liippocampus (HC) and the amygdala (AM) as well as the mammillary bodies (MB). The overall findings of their studies were that lesions in the anterior hypothalamus are immunosup-pressive, and this immunosuppression is not due to increased levels of corti-costeroids but rather to increased suppressor activity of macrophages. They also found that lesions in the HC, AM, and MB enhanced the immune response as measured by splenocyte and thymocyte activation and blastogenic responsivity to the mitogen ConA. Similar lesioning studies were then conducted before and after hypophysectomy. These revealed that the removal of the pituitary gland abrogated many of the inhibitory effects induced by anterior hypothalamic lesions and all facilitory effects of HC and AM lesioning. Thus many of the immunologic alterations that are induced by central nervous system lesioning are mediated through the pituitary gland or other endocrine components. These sorts of studies would seem to set the stage for future dissection of the possible in vivo involvement of pituitary peptides in immunoregulation.
A more specific example of possible immunomodulation by the neuroen-docrine system is the involvement of substance P in arthritis (for review, see Ref. 85). The finding of a greater density of substance P-containing nocicep-tive afferents in a joint that develops more severe arthritis (ankle) suggested a role of substance P in joint injury. Direct evidence that the proinflamma-tory factor released from these nociceptors is substance P is provided by the finding that the injection of substance P into a joint that normally develops less severe arthritis (knee) increases the severity of arthritis in that joint. Although it remains to be determined whether certain aspects of such inflammation are indirectly mediated through substance P effects on the local immune system, this seems a distinct possibility considering the functional responsivity of leukocytes to this neuropeptide (60, 111). It seems that a greater understanding of such molecular aspects of neuroendocrine immune interactions will have an impact on the immunologic manifestations of human conditions such as stress and bereavement (135).
children’s ALLERGY CENTER online
JL TAMAN BENDUNGAN ASAHAN 5 JAKARTA PUSAT, JAKARTA INDONESIA 10210
PHONE : (021) 70081995 – 5703646
email : firstname.lastname@example.org\
Information on this web site is provided for informational purposes only and is not a substitute for professional medical advice. You should not use the information on this web site for diagnosing or treating a medical or health condition. You should carefully read all product packaging. If you have or suspect you have a medical problem, promptly contact your professional healthcare provider.
Copyright © 2009, Children Allergy Center Information Education Network. All rights reserved.