| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print October 4, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1,
* Center for Substance Abuse Research,
Fels Institute for Cancer Research and Molecular Biology,
Department of Pharmacology, Temple University School of Medicine, Philadelphia, Pennsylvania
1 Correspondence: Center for Substance Abuse Research, Temple University School of Medicine, 3400 N. Broad Street, Philadelphia, PA 19140. E-mail: baldeagl{at}temple.edu
ABSTRACT
Chemokines are a family of small proteins involved in cellular migration and intercellular communication. Although the chemokines and their receptors are located throughout the brain, they are not distributed uniformly. Among the chemokines and their receptors that are arrayed disproportionately in glia and neurons are monocyte chemotactic protein-1/CC chemokine ligand 2 (CCL2), stromal cell-derived factor-1/CXC chemokine ligand 12 (CXCL12), fractalkine/CX3C chemokine ligand 1, interferon-
-inducible-protein-10/CXCL10, macrophage inflammatory protein-1
/CCL3, and regulated on activation, normal T cell expressed and secreted/CCL5. In the brain, they are found in the hypothalamus, nucleus accumbens, limbic system, hippocampus, thalamus, cortex, and cerebellum. The uneven distribution suggests that there may be functional roles for the chemokine "system," comprised of chemokine ligands and their receptors. In addition to anatomical, immunohistochemical, and in vitro studies establishing the expression of the chemokine ligands and receptors, there is an increasing body of research that suggests that the chemokine system plays a crucial role in brain development and function. Our data indicate that the chemokine system can alter the actions of neuronally active pharmacological agents including the opioids and cannabinoids. Combined with evidence that the chemokine system in the brain interacts with neurotransmitter systems, we propose the following hypothesis: The endogenous chemokine system in the brain acts in concert with the neurotransmitter and neuropeptide systems to govern brain function. The chemokine system can thus be thought of as the third major transmitter system in the brain.
Key Words: opioids • heterologous desensitization • neuromodulators • interneuronal communication
INTRODUCTION
Chemokines are a family of small (8–12 kDa) proteins involved in cellular migration and intercellular communication, and their receptors are members of the G protein-coupled receptor (GPCR) superfamily. Most chemokines act on more than one receptor, although a few, such as CXC chemokine ligand 12 (CXCL12) and CX3C chemokine ligand 1 (CX3CL1) appear to act on only one [1 , 2 ]. Although it appears that some chemokines and their receptors are expressed by glia and neurons throughout the brain, others, such as CC chemokine ligand 2 (CCL2), CCL3, CCL5, CXCL10, CXCL12, and CX3CL1, are not distributed uniformly [3 , 4 ]. Such uneven distribution is characteristic of the neurotransmitter and neuropeptide systems and may hold a clue as to possible functional roles played by the different members of these systems in certain defined brain regions. Chemokines and their receptors are found in microglia and neurons in such brain areas as the hypothalamus, nucleus accumbens, limbic system, hippocampus, thalamus, cortex, and cerebellum [3 4 5 6 ]. Neural progenitor cells and mature neurons express multiple functional chemokine receptors and chemokines [7 , 8 ]. In addition, small numbers of resident immune cells are present in the brain [9 , 10 ] and are likely to contribute chemokine ligands to the brain tissue.
DEFINITIONS
A number of terms and classifications have been used in an attempt to categorize neuroactive compounds, chemicals that affect neuronal activity and play a functional role in regulating activity of the brain in normal, pathological, and homeostatic states. Among the most commonly used terms are neurotransmitter, neuromodulator, and neurohormone. Neuropeptides are often considered separately, and neurohormones (e.g., hypothalamic-releasing factors) are a type of neuropeptide. Some characteristics of these categories are summarized in Table 1 [11 ].
|
Irrespective of terminology such as that cited above, neuroscientists generally think of the systems of chemicals classified as neurotransmitters (e.g., acetylcholine, norepinephrine, dopamine, serotonin, -aminobutyric acid) or neuropeptides (e.g., ß-endorphin, enkephalins, dynorphin, cholecystokinin, substance P, endocannabinoids, somatostatin) as being responsible for neuronal transmission or modulation of neuronal events. The molar concentration of neuropeptides is two or three orders of magnitude lower than neurotransmitters, and their receptors respond to the lower concentrations. Neurotransmitters and neuropeptides may be co-localized in individual synaptic terminals and may be co-released. Like the neurotransmitter and neuropeptide systems, the chemokine system, composed of chemokine ligands and their respective receptors, is widely but unevenly distributed in the brain and has ligands and receptors located in neurons. Neurotransmitter and neuropeptide systems interact in the brain, and it is therefore logical to postulate that the chemokine system interacts functionally with both of these other neuronal systems.
HETEROLOGOUS DESENSITIZATION
Most reports on the subject of chemokines and their receptors in the brain have focused on their role in inflammatory processes or neurogenesis. The possibility that these substances play an even more fundamental role in neurotransmission has not been given adequate consideration. Work carried out with leukocytes in vitro first showed that opioid treatment suppressed chemokine-mediated chemotactic responses, and this was shown to be a result of (at least in part) the process of heterologous desensitization between opioids and some of the chemokine receptors (Fig. 1 ) [12 ]. Subsequently, results were reported that demonstrated that the heterologous desensitization process is bidirectional, and chemokine receptor activation was shown to inactivate opioid receptor activity in vitro [13 14 15 ]. Of potential functional significance was the finding that cross-desensitization of CC chemokine receptor 5 (CCR5) by opioids resulted in decreased susceptibility to R5 but not X4 strains of human immunodeficiency virus type 1 (HIV-1) [16 ].
|
-opioid receptors, suggesting that CXCR4 is a relatively potent desensitizer. Studies reported at this time suggest that heterologous desensitization can involve target receptor phosphorylation or the alteration of downstream signaling events such as the inhibition of phospholipase C activity. It is clear that cross-desensitization of GPCRs occurs in many cases as a result of activation of second messenger protein kinases [protein kinase A (PKA) or PKC]. These serine/threonine kinases can act directly on GPCRs and result in receptor desensitization by inhibiting the coupling of G proteins to the receptor. Target receptor phosphorylation is a component of the cross-desensitization of CXCR1 and CXCR2 by opioid receptors [12
], as is the desensitization of the µ-opioid receptor by CCL5 [13
, 15
]. Indeed, target cell phosphorylation is a common observation for cross-desensitization among several of the chemoattractant receptors [17
18
19
20
]. However, simple receptor phosphorylation does not provide an explanation for cross-desensitization in all cases, including the desensitization of FPR by C5aR or CXCR2 [17
]. The biochemical basis for the selectivity that exists in the cross-desensitization process remains undefined at this time. Although the evidence clearly demonstrated that heterologous desensitization between opioid and chemokine receptors occurred in vitro, an unanswered question was whether some of the chemokines might have the ability to desensitize opioid receptors in vivo. Such a finding was necessary to determine whether the in vitro findings are physiologically relevant. We predicted that the administration of a chemokine should be able to block or diminish the effect of an opioid without producing an effect of its own. We found that the analgesic response to opioids in a rat model, as determined by measuring the ability of a drug to increase the threshold to a noxious stimulus (the rat cold water tail-flick test [21 ]), was inhibited following chemokine administration. Animals were administered a range of doses of CXCL12 or CCL5 directly into the periaqueductal gray (PAG), the area primarily involved in the antinociceptive effects of opioids, 30 min prior to morphine or D-Ala2, NMe-Phe4-Glyol5 (DAMGO), a selective agonist at the µ-opioid receptor. These chemokines interfered with the normal analgesic response to the opioids [13 ].
To examine temporal aspects of the desensitization, the system was probed further. The question posed was whether the desensitization would still occur if the time interval between administration of the chemokine and the opioid were increased. It was found that if the CCL5 treatment was given 60 min prior to the opioid, analgesia was only partially blocked, but if the interval was increased to 2 h, the ability of CCL5 to desensitize the opioid receptor was lost. It is interesting that if the time interval was 2 h or greater, and the chemokine was readministered, the inhibition of the analgesic activity of DAMGO was restored [13 ]. A logical explanation is that the cross-desensitization of the opioid receptor is a reversible process. A testable hypothesis is that the ligand is metabolized as it is released from the receptor, and the reinstatement of the inhibition of opioid receptor function occurs with restoration with fresh chemokine ligand. It is also possible that new receptors are expressed on the cell surface, permitting additional ligand to bind. Formal proof regarding mechanisms has not been obtained, but we have found that there is no detectable internalization of the chemokine or opioid receptor following cross-desensitization in primary leukocytes or in a number of leukocyte and nonleukocyte cell lines [13 , 22 ]. Finally, it has also been shown that the cross-desensitization process is associated with target receptor phosphorylation [22 ]. Furthermore, the heterologous desensitization between CCR5 and the µ-opioid receptor is also associated with the formation of heterodimers composed of these two receptors [22 ].
Although the time course was slightly different with CXCL12 (the inhibition of opioid receptor function lasts for 
60 min longer), the kinetic patterns for reversible inhibition of opioid receptor function were observed for CXCL12 and CCL5 [13
]. The experiments cited thus far demonstrated that the µ-opioid receptor can be desensitized by treatment with certain chemokines. To determine if the -opioid and the 
-opioid receptors could also be cross-desensitized in vivo, D-Pen2, D-Pen5-enkephalin, a selective -agonist, and dynorphin A 1–17, the endogenous opioid, which is a selective -agonist, were also tested. The - and -opioid receptors were cross-desensitized by treatment with CCL5 or CXCL12, indicating that desensitization of all three opioid receptors is achieved with activation of CCL5 and CXCL12 chemokine receptors [13
] (and unpublished results from our laboratories). These chemokine receptors are expressed on neurons in multiple areas of the brain, including the PAG [6
].
To demonstrate that the chemokine actions are mediated via the chemokine receptors, the effect of AMD 3100, an antagonist at the receptor for CXCL12, was tested. The antagonist abolished the desensitization, allowing for full analgesic activity of DAMGO at the µ-opioid receptor (unpublished observations). These results confirm that the chemokine-induced desensitization was mediated through the chemokine receptor.
Another chemokine that is expressed in the brain is CX3CL1. Bajetto et al. [2
] reported that this was the only chemokine found to be expressed in higher concentrations in the central nervous system (CNS) than in the immune system and peripheral tissues. Like the other chemokines tested, CX3CL1 itself had no antinociceptive effect. Nevertheless, activation of the fractalkine receptor was able to partially block the antinociceptive action of DAMGO, the µ-opioid receptor agonist. Similarly, CX3CL1 also partially blocked the analgesic effects of a -selective agonist and a -selective agonist (unpublished observations). The basis for the inability of CX3C chemokine receptor 1 (CX3CR1) to inhibit cannabinoid receptor function in the PAG is currently under investigation.
Although CCL5, CXCL12, and CX3CL1 desensitized opioid receptors, the desensitization did not occur with all chemokines. For example, CCL2 did not induce the cross-desensitization of the opioid receptors. The reason probably lies in the fact that the CCL2 receptor, CCR2, is not expressed by neurons in the PAG matter at a significant level [4 ], and all injections were directly into the PAG.
We have suggested [13 , 14 , 23 ] that the implication of these results is that in situations where there is an elevation of chemokine levels in the brain (including most neuroinflammatory diseases such as multiple sclerosis and HIV encephalitis), there is a resulting loss of opioid receptor function, leading to greater sensitivity to painful stimuli. These results provide evidence of significant functional modulation of neural processes by chemokines and their receptors and support the notion that there is a clear role for chemokines in neural communication. Indeed, evidence based on the expression and functional activity of the chemokines and/or their receptors provides convincing support for our hypothesis that the chemokine system plays a major role in brain function (Table 2 ).
|
The systems of neurotransmitters and neuropeptides in the brain are accepted as playing the major role in the functioning of the brain in maintaining homeostasis and reacting to perturbations of that homeostasis. We propose that the endogenous chemokine system in the brain, consisting of ligands and receptors, is a third major system of the brain. It is the third leg in the stool that supports brain function.
There is certainly no longer any question that chemokines have important actions in the brain and that they are neuroactive compounds that have direct and indirect actions on neurons. Bajetto et al. [28 ] suggested that chemokines might act as neuromodulators, and Tran and Miller [31 ] asked, "Do chemokines play some unexpected role in the physiology of the normal brain?" Given the evidence that has accumulated in the past 5–7 years, it is reasonable to use the above hypothesis relating to the chemokine system in the brain to begin to make testable predictions.
The importance of the chemokine system in the brain and its role in the functioning of the brain in homeostatic and perturbed situations is just beginning to be appreciated. This is apparent in studies about the role of CXCL12 and CXCR4 in brain development. CXCL12 and its receptor CXCR4 are required for normal brain development, based on the abnormal neuronal organization in the cerebellum of CXCR4- or CXCL12-deficient mice [32 33 34 ]. The absence of this chemokine or the receptor appears to result in a premature migration of external granule cells into the cerebellum [32 , 35 ], and it has been suggested that CXCL12 may normally act to prevent excessive migration of these granular cells by chemoattracting them away from the inner cerebellum and toward the pia matter overlying the cerebellum [32 , 35 36 37 ]. In addition, the absence of CXCR4 leads to an absence of proliferating cells in the dentate gyrus, and neurons in this anatomic site appear to differentiate prematurely before completing migration [33 ]. It is likely that a role for CXCL12 persists through adult life by virtue of the role of this chemokine in promoting neuronal survival and proliferation.
Figure 2 illustrates the concept being proposed in this paper. It is known that there are glia-to-glia and glia-to-neuron methods of communication, and it is known that glia can produce and release chemokines. In light of the more recent knowledge that chemokines and their receptors are also present in neurons, we hypothesize that there is also neuron-to-glia communication involving the chemokines. It is even more important that we hypothesize that neuron-to-neuron communication involves an additional system, the chemokine system of ligands and receptors. Thus, in addition to neuropeptides and neurotransmitters performing the vital function of transmission between neurons, we propose that chemokines participate in this function. Given the complexity of neuronal function under normal conditions and when perturbed, recognition of the chemokine system as potentially playing a vital role in brain functioning would allow another layer of interacting molecules to explain how the neuronal system achieves such magnificent control of discrete physical, behavioral, and chemical action and to envision new approaches to treat perturbations in homeostasis and enhance brain activity in nonpathological states.
|
This paper has put forth the postulate that chemokines function as transmitters in neuron-to-neuron, glia-to-neuron, and neuron-to-glia communication. Glia-to-glia communication for chemokines is well established based on experimental results from a large number of laboratories. For obvious reasons, we will not endeavor to review all of the criteria, which have established that neurochemical systems of the brain, such as those involving acetylcholine, dopamine, and serotonin, are transmitters. Most modern physiology, neurochemistry, and neuropharmacology texts discuss these criteria. Despite the fact that neuropeptides have been studied for over 30 years, not all the criteria have as yet been met [11 ] to allow neuropeptides to be called neurotransmitters. Nevertheless, neuroscientists universally accept their important role in neuron-to-neuron communication. As it is a system that has only recently been found to be involved in brain function, much remains to be discovered about the chemokine system. For example, the mechanism of release from neurons, the storage mechanisms, and the inactivation mechanisms are not known. Specifically, experiments to determine if selected chemokines can induce or alter neuronal responses (e.g., using patch-clamp technology) would be desirable and would provide needed information regarding a possible role for the chemokines in neurotransmission. Experiments should also be conducted to determine the consequences of administration of transmitter combinations following in vivo administration, as this will more accurately reflect the normal physiology. Finally, experiments that take advantage of the growing list of chemokine and chemokine receptor-deficient mouse strains will likely illuminate the potential role of those chemokine receptors and their ligands in the response of the brain to various stimuli. One notable example of this strategy is the recent work with CX3CR1-deficient mouse strains. As noted above, CX3CR1 and CX3CL1 are expressed in the brain, and CX3CR1 is present on microglia and certain neurons, suggesting a possible mode of communication between these two cell populations. However, studies with knockout mice do not currently suggest that this chemokine receptor pair plays a critical role in microglial cell development or activation [38 , 39 ].
Little information exists as to specific functional systems in the brain linked to individual chemokines. However, this last issue also pertains to neuropeptides. Despite the paucity of data in terms of some of these matters, current evidence is consistent with the hypothesis that the chemokine system acts as a major system in the brain to facilitate or be an actual chemical system of ligands and receptors, which transmits information in a manner analogous to neurotransmitters and neuropeptides.
POTENTIAL THERAPEUTIC SIGNIFICANCE
In conclusion, several novel, potential therapeutic applications can be envisioned by coupling the capacity for heterologous desensitization of the opioid and chemokine systems with knowledge about the neuronal function of the chemokine system. First, it may be possible to improve the treatment of pain associated with neuroinflammatory states by blocking heterologous desensitization of opioid receptors. Again, the chemokine-mediated cross-desensitization of the opioid receptors appears to account for diminished efficacy of opioids in inflammatory pain. Second, the reverse strategy may be used in which certain unwanted side-effects of opioids may be inhibited by using chemokine desensitization of selected opioid pathways after determining neuronal localization of chemokine receptors. Third, we have already reported that the opioid-induced cross-desensitization of CCR5, a major coreceptor for entry of HIV in the CNS or periphery, results in reduced HIV susceptibility [16 ]. Moreover, the cross-desensitization of CCR5 and CXCR4 by other GPCRs, including the FPRs, has been shown to result in the loss of HIV coreceptor function [14 , 18 19 20 ]. Fourth, one may use the cross-desensitizing effects of chemokines to modify the potency of drugs of abuse. A similar strategy may be used to provide improved control of the activity of certain behavior-modifying, therapeutic agents used to treat certain mental disorders. Finally, it is conceivable that new approaches to prevention and treatment of neurodegenerative diseases may be achieved through blocking or enhancing the functional activity of the chemokines by selectively inducing or inhibiting, repectively, the cross-desensitization of these receptors. Each of these strategies will require a much greater understanding of the biochemical basis for the process of heterologous desensitization. Furthermore, any workable, therapeutic strategy will require intervention in a highly selective manner to avoid the undesirable consequences of widespread cross-desensitization of a large number of GPCRs. Perhaps the necessary degree of selective targeting of the GPCR of interest can be developed once a greater understanding of the biochemistry is achieved. The molecular basis for the heterologous desensitization process is a major focus of our current work.
ACKNOWLEDGEMENTS
The research reported here from the laboratories of Temple University Center for Substance Abuse Research (CSAR) was supported by numerous grants, especially DA06650, DA07237, DA14230, and DA-P30-13429 from the National Institute on Drug Abuse. The authors acknowledge the contribution of numerous members of the CSAR in conducting experiments and helping to develop the ideas that led to the hypothesis about the role of the chemokine system in brain function. Foremost among those are Dr. Toby K. Eisenstein and Ms. Ellen B. Geller. In addition, the scientific input of the laboratories of Drs. Lee-Yuan Liu-Chen, Ronald J. Tallarida, Alan Cowan, Ellen M. Unterwald, Xiaohong Chen, and Khalid Benamar and the technical assistance of Margaret Deitz are gratefully acknowledged.
Received April 24, 2005; revised August 26, 2005; accepted September 9, 2005.
REFERENCES
and µ opiate receptor-mediated heterologous desensitization J. Exp. Med. 188,317-325This article has been cited by other articles:
|
|
 |
|
  S. Bruzzone, I. Moreschi, C. Usai, L. Guida, G. Damonte, A. Salis, S. Scarfi, E. Millo, A. De Flora, and E. Zocchi Abscisic acid is an endogenous cytokine in human granulocytes with cyclic ADP-ribose as second messenger PNAS, April 3, 2007; 104(14): 5759 - 5764. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
