|Additional related information may be found at:|
|Neuropsychopharmacology: The Fifth Generation of Progress|
General Overview of Neuropeptides
General Overview of Neuropeptides
Tomas Hökfelt, Zhi-Qing David Xu, Christian Broberger, Tiejun Sten Shi and
The development of the neuropeptide field was covered in several chapters in the three last editions of this series, i.e. "Psychopharmacology: A Generation of Progress" (96), "Psychopharmacology: The Third Generation of Progress" (108) and “Psychopharmacology: The Fourth Generation of Progress” (15). In the second volume (96) 11 chapters out of 149 dealt with neuropeptides. In the third volume (108) 6 papers discussed basic aspects on peptides, and in a further 11 articles possible involvement of peptides in various diseases including schizophrenia and Alzheimer’s disease were considered. In the concluding chapter by Bloom (13) peptides were also discussed, suggesting that peptides have secured a position in the field of psychopharmacology. Finally, in the Fourth Generation of Progress (15) 14 chapters directly dealt with neuropeptides, and it was stated that although many interesting concepts and hypotheses had been advanced, there was up till then “no major breakthrough in terms of a casual relation between a peptide and a disease or with regard to the use of drugs acting via peptidergic mechanisms for treatment of CNS disorders” (59). In contrast, we can now report that, in fact, such a breakthrough seems to have occurred. Thus, Kramer et al. (81) have shown that a substance P (NK-1) antagonist has as good antidepressant activity as paroxetine, a serotonin reuptake inhibitor, but with less side effects.
In the present revised chapter, as in the original one, we will, due to limited space, essentially only include references from 1987 and later. Also, we will not attempt to deal with all recent findings, for example with regard to novel peptide receptors and peptide antagonists, but try to focus on some, at least to us, interesting findings. We now also include a list of peptides and peptide families (without referring to the original papers) (Tabel 1). For more detailed information see, for example, reviews by Hökfelt (58), Kupfermann (84), Otsuka (133) and Herbert (54) and books by Burbach and de Wied (19) and Strand (158).
NEW METHODOLOGICAL APPROACHES
There is reason to believe that over the next few years further decisive data will emerge with regard to the involvement of peptides in various CNS functions, in one or the other direction, that is to confirm or refute existing hypotheses. The basis for this cautious optimism is that, in general terms, several major advancements have been made, since the publication of the third volume in 1987 (108). Thus, novel techniques for studies of peptides have been taken into use. The histochemical approach has benefitted in particular from various in situ hybridization procedures (see 183). Based on molecular biological methods numerous peptide receptors have been cloned, including three types of opioid receptors (see 10, 113, 148) and Table 2). Furthermore, after the introduction of the first peptide antagonists, which all were modified peptide molecules, there is now available a second generation of compounds which are small molecules of non-peptide nature and which pass the blood brain barrier (Table 3). As indicated in this table, the development of these antagonists has occurred within the frame-work of the pharmaceutical companies. Clearly, the resources of the industry and their commitment provide hope for further antagonists which can be available both for studying the role of peptides in experimental animals as well as testing in various disease states. An alternative approach with which to explore the functional role of peptides is the use of antisense probes. Thus over the last years numerous studies have been published in which oligonucleotide probes complementary to a sequence of a certain peptide mRNA have been used to block or attenuate translation of mRNA into protein. Interestingly, successful results have been obtained not only in in vitro studies but also e.g. after intraventricular administration (see 171). Moreover, it is considered that such oligonucleotides may become useful as therapeutic agents (see 157). Finally, an increasing number of transgenic mice have been created and are contributing to the understanding of peptide functions (see 14, 134).
In the 1970-ties the hunt for new peptides was intense, and an ever growing list could be compiled (Table 1). Since the publication of "The Third Generation of Progress” Volume (108), several important peptides have been discovered (Table 1), for example the atrio-natriuretic factor and the endothelins which have corresponding family members in the brain with interesting distribution patterns (45, 152, 161). In 1993, secretoneurin (77), a cleavage product from secretogranin II (or chromogranin C), built up of 33 amino acids was found with interesting distribution patterns both in the rat (101) and human (102) brain, as well as having potent biological effects, for example causing release of dopamine from striatal slices (144). The latter studies strengthen earlier knowledge that the large polypeptide precursor molecules may hide interesting biological activities, and suggest that further analysis of precursors may yield other peptides, as has been repeatedly shown in other studies on, for example, the opioid peptide precursors.
Some peptides have been found via one of the many orphan receptors, that is receptors for which the endogenous ligand is unknown. One such receptor is the "opioid receptor-like 1" (ORL1) which was identified independently by several groups using homology-based screening strategies. Subsequently, the structure of the endogenous agonist of this ORL1 receptor, a 17-amino acid peptide termed orphanin FQ or nociceptin, was elucidated (112, 138). A study focusing on another orphan receptor, hGR3 (also called GPCR) revealed that a prolactin releasing peptide in the brain represents an endogenous ligand for this receptor (57). Using cocaine and amphetamine stimulation a highly abundant transcript encoding several putative peptides has been discovered and termed cart (cocaine and amphetamine regulated transcript) (31). Several peptides involved in regulation of feeding behaviour have been identified such as the endogenous melanocortin antagonist agouti gene-related protein (AGRP) (132, 153) and the orexins (143), alternatively named hypocretins (27), and also their receptors (143). Some further peptides related to pain sensation have been discovered, that is two potent and selective endogenous agonists for the mu-opiate receptor called endomorphin-1 and -2 (184), and nocistatin which is part of the same precursor as orphanin FQ/nociceptin and which blocks the effect of orphanin FQ/nociceptin (131). Massot et al. (104) have described a new peptide, 5-hydroxytryptamine (5-HT)-moduline, which influences serotonergic transmission via interaction with 5-HT1B/1D receptors.
It is well established that peptides occur in the whole animal kingdom (as well as in plants). The hydra, a member of the coelenterates, is considered to have the most simple nervous system, but it still contains numerous peptides, mainly belonging to the FMRFamide family (see 51). FMRF amide-like peptides are expressed in ~ 10% of the neurons in Caenorhabditis elegans, a nematode, and are encoded by at least 14 genes which also are transcribed (125). It is assumed that each peptide family has its origin in a many million years old primordial gene that through various processes such as gene duplication and point mutations have developed into the presently known family member (see 65). This has been described in particular detail for the two hypothalamic magnocellular neurosecretory peptides oxytocin and vasopressin, in fact the first discovered of all neuropeptides (32). Although of high structural homology, they are produced in different neurons, have different physiological activities with two molecular lineages, the isotocyin-mesotocyn-oxytocin line associated with reproductive functions and the vasotosin-vasopressin line primarily concerned with water and electrolyte balance (see 1, 65). Not only in the phylogenetic perspective, but also in terms of discovery of peptides, the frog has turned out to be an important species through the work mainly of Vittorio Erspamer and his collaborators in Italy. They have discovered a large number of peptides, many of which subsequently could be correlated to mammalian peptides.
COEXISTENCE OF MESSENGERS
When discussing the role of peptides it is important to note that they virtually in all systems coexist with at least one classic transmitter (and often with other peptides) (see 58). Most attention has been focused on colocalization with biogenic amines and acetylcholine. For example, many 5-HT neurons in the raphe nuclei and noradrenaline neurons in the rat locus coeruleus synthesize galanin, and galanin can be visualized in most noradrenergic nerve terminals in cortex and hippocampus, and exerts actions both at the cell body and nerve terminal level to inhibit firing and classic transmitter release (178, 179) (Fig. 1) There is immunohistochemical evidence that neurons in addition may express an amino acid transmitter, such as GABA (9) or glutamate (72, 127). In fact, using triple-staining methodology, evidence for coexpression of glutamate-, 5-HT- and substance P-like immunoreactivities in bulbo-spinal neurons has been presented (128). In an elegant in vitro study, Johnson (70) has shown co-release of serotonin and glutamate from rat mesopontine neurons. In the peripheral nervous system certain neurons may release ATP, noradrenaline and neuropeptide tyrosine (NPY) (73, 154). Thus, neurons may release a 'cocktail' of messenger molecules providing a spectrum of biological actions, including different temporal information (fast, intermediate and slow signalling). The functional implications of coexistence have been discussed in several reviews (see e.g. 84, 98).
A major difference between principal transmitters and peptides is the mode of synthesis and replacement after release. Thus, classic transmitters often have a membrane reuptake mechanism, now known to be represented by specific transporter molecules, which allow reutilization of the transmitter. There is also mostly local synthesis in the nerve endings. With regard to peptides it has been a dogma that they can only be produced ribosomally in the cell bodies with packaging in the Golgi apparatus (see 150), and that replacement after release has to occur via axonal transport from the cell bodies to the nerve endings. However, contrary to earlier belief, occurrence of a high affinity uptake for cholecystokinin has been reported (116).
An important issue will be to understand the mechanisms underlying the release of these different classes of messenger molecules. There is early evidence that the release of classic transmitter and peptide can be differential and dependent on frequency and patterns of firing, presumably due to different subcellular storage sites (see 98). Verhage et al. (167) have analysed this question on isolated nerve endings, providing evidence that neuropeptide release is triggered by small elevations in the Ca2+ concentration in the bulk cytoplasm, whereas secretion of amino acids requires higher elevations, as produced in the vicinity of Ca2+ channels', i.e. near the active zone at synapses.
Interestingly, there are indications that RNA can be transported into axons (68, 120). Thus, in addition to studies with non-radioactive in situ hybridization showing neuropeptide-encoding mRNAs in neuronal processes beyond the perikaryon (12), there has over the last few years accumulated evidence that oxytocin and vasopressin mRNAs can be detected in the posterior lobe of the pituitary and in the median eminence, that is the projection areas of the magnocellular hypothalamic neurosecretory neurons producing these two peptides (68, 91, 94, 106, 118-120, 122, 166). This is true also for galanin mRNA (86). Moreover, evidence has been presented for uptake and expression of exogenous vasopressin mRNA after injection into the lateral hypothalamus of vasopressin-deficient Brattleboro rats (69) and that exogenous vasopressin mRNA can transiently correct diabetes insipidus in such rats (99). The function of mRNA within the axonal compartment still remains to be elucidated, but these findings raise several interesting questions concerning role(s) of mRNA and site of synthesis of peptides.
The existence of neuropeptide receptors remained obscure for a long period, in spite of many serious attempts to biochemically purify and to clone such receptors, particularly with regard to the opioid receptors. Thus, in spite of the demonstration of multiple types of binding sites for various opioid peptides in the early 1970’s as well as the strong evidence from many ligand binding studies, the receptor proteins were elusive. In fact, since peptides frequently seemed to coexist with classic transmitters, it could not be excluded that binding sites for peptides were located on the receptor molecules of classic transmitters. It was therefore important to establish the nature of peptide receptors, and this was first achieved for members of the tachykinin family in Japan by Nakanishi and his collaborators.
In 1987 the isolation of a cDNA clone for a bovine substance K receptor from a stomach cDNA library was reported (105), and this first cloned peptide receptor belonged to the family of G-protein-coupled receptors with seven membrane-spanning segments. Subsequently, the neuronal substance P (Fig. 2) and substance K (55, 182) and the neurotensin (163) receptors were cloned and shown to belong to the same family. Subsequent work has demonstrated that so far virtually all neuropeptide receptors are of this type, including the cloned d, µ and k opioid receptors (113) (Table 2). An exception has been found in that the peptide FMRFamide induces a fast excitatory depolarizing response via direct activation of an amiloride-sensitive sodium channel, the first example of a peptide gated ionotropic receptor (48, 95).
Today, the G-protein-coupled receptor families can be divided into the rhodopsin-like receptor family and the glucagon-vasoactive intestinal peptide/calcitonin receptor family (see 148). Neuropeptide receptors belonging to the rhodopsin-like family are listed in Table 4 (from ref. 148). This family represents receptors for a number of other types of ligands including retinal-opsins, odrans, monoamines, purines, some chemokines and glucoprotein hormones. The neuropeptide receptors of the second family are also listed in Table 2. The latter receptors are characterized by large N-terminal extracellular part with six conserved cysteine residues which presumably are involved in ligand binding. These ligands appear to fold into a common two helical confirmation.
From a conceptual point of view, the identification of these neuropeptide receptors has been of great importance, since it shows that the neuropeptides have their own targets and own second messenger systems, and thus, presumably, their own physiological roles. Even if there is at least one receptor cloned for almost all of the peptides so far discovered, it still remains to be elucidated how many receptors can be found for each of the peptides. So far five somatostatin receptors (see 64) and a similar number of NPY receptors (see 88) have been cloned, and it does not seem unreasonable to assume that there are several receptors also for many of the other peptides. However, their number may not approach those described for some of the classic transmitters, such as the 15 or so for serotonin. Needless to say, the multiplicity of receptors offers unique and important openings for the development of specific agonists and antagonists which may be useful for treatment of various disorders.
The cloning of various peptide receptors has allowed studies of their regulation and of the distribution of cell bodies producing these receptors through in situ hybridization. The information from the cloning studies can also be used to produce antibodies against specific portions of the receptor protein, either raised against a short, unique peptide sequence or against longer peptides produced in various expression systems. For example, antibodies against the substance P (NK-1) receptor have been produced and used for analysis of distribution in the brain (e.g. 124), and the distribution of three opioid receptors has also been mapped (see 34).
Particularly interesting findings have been reported on internalization of the NK-1 receptor (46, 100). Manthy et al. (100) observed after peripheral nerve stimulation for a few minutes that the previously membrane-bound receptors were found in the cytoplasmin of second order spinal dorsal horn neurons , a process that was reversed within 30 minutes. Similar events have been observed for mu opioid receptors in some brain areas (74). This may be one mechanism underlying the powerful ‘desensitization’ often seen after application of peptides. In the case of NK-1 receptors in dorsal horn, could it even be a mechanism to prevent substance P to exert an action after release from central primary afferents in the dorsal horn, and thus direct the function of this peptide to the peripheral branches of the sensory neurons?
Recently, a novel principle for neuropeptide receptor regulation and ligand specificity has been described by McLatchie et al. (107, see also 104). They provide evidence that a single receptor (the calcitonin-receptor-like receptor) can have two alternative pharmacological profiles, which react with two different, though related, peptides depending on specific factors, so called receptor-activity-modifying proteins (RAMPs). Thus, in the case of CGRP and adrenomedullin, RAMP1 presents the receptor at the cell surface in the form of a mature glycoprotein sensitive to CGRP, whereas RAMP2-transported receptors are core-glycosylated and represent adrenomedullin receptors. This opens up a new field of receptor plasticity, and raises the question whether or not this is a mechanism operating for other types of ligands and receptors.
The neuropeptide antagonists, agonists and ENZYME INHIBITORS
Drugs active on peptidergic mechanisms can be classified into at least three categories, antagonists, agonists and peptidase inhibitors, the latter preventing peptide breakdown and thus strengthening the peptidergic transmission indicate (agonist effect). The early antagonists were of peptidergic nature, mostly D-substituted analogs. Of special importance was the finding by Peikin and collaborators that buturyl derivatives of cGMP antagonize the action of CCK, and this was followed by development of more potent CCK antagonist at the Rolla Laboratories (Italy), Merck Sharp and Dohme (USA and UK) and Parke-Davis (UK) (Table 3). In this work, various types of approaches have been taken, mainly extensive screening efforts resulting in lead compounds which have been modified into powerful and specific, non-peptide antagonists. However, also rational drug design starting out from the peptide molecule itself has been employed (see 66).
Many of these compounds pass the blood brain barrier and can thus be used to probe for central peptide functions under normal circumstances and after various types of manipulations, and potentially in disease states. In a recent review article by Betancur et al. (10) well over a 100 such compounds were listed for a number of peptide receptors, with, for example, around 20 compounds for both the CCKB and NK-1 (substance P) receptors, and almost 30 for the angiotensin AT1 receptor. The binding characteristics of such antagonists are now being worked out (see 148).
Development of potent and specific agonists has been more complicated. However, peptide modification has resulted in, for example, a potent CCKB agonist that passes the blood brain barrier (33), and a potent analgesic opioid has been extracted from a synthetic combinatorial library containing 52.128.400 D-amino acid hexapeptides (30). An alternative to peptide agonists is represented by peptidase inhibitors. For example, inhibitions of enkephalin catabolism represent a potential drug treatment for pain and opioid addiction. In fact, drugs have been developed which inhibit two enzymes, neutral endopeptidase 24.11 and aminopeptidase N, involved in enkephalin degradation (see 142).
Why are peptide antagonists suitable as drugs? In general terms, the effects of peptides seem to be much less pronounced than those of classic transmitters such as glutamate and GABA and still less than those of monoamine transmitters. Moreover, peptides may preferentially be released, at least in some systems, when neurons are strongly activated or under pathological conditions. It is thus only under these circumstances that an antagonist can exert an effect, and these characteristics together should lead to less pronounced side effects. The recent demonstration of clinical efficacy of an NK-1 antagonist in depression (81) may represent an example of this. Moreover, the discovery that peptides often have more than one receptor provides possibilities to design antagonists for such receptor subtypes which may be involved in specific functions. This principle has been important in the monoamine field where already many subtype specific agonists and antagonists have been developed.
Plasticity in peptide expression
Studies both in the peripheral and central nervous system have demonstrated that peptide levels may vary considerably during different conditions, including endogenous variations and after experimental manipulations. This is per se not surprising, since peptides in general are assumed to be produced ribosomally and since replacement after release only seems to occur via new synthesis in cell bodies (however, see below). This is in marked contrast to, for example, catecholamines which can be locally synthezised in nerve endings and also be replaced by efficient reuptake mechanisms. Moreover, the rate of catecholamine synthesis can also be regulated by phosphorylation of synthetic enzymes. Thus, it is possible to keep classic transmitter levels constant under various conditions. The quite dramatic regulation of peptide synthesis has been particularly evident when using the in situ hybridization technique (see 183) for analysis of peptide mRNA levels in neuronal somata under various experimental conditions. Here we will focus on two systems, primary sensory neurons and neurons in the rat striatum. We are convinced that similar regulations occur in many other systems, which are important from a psychopharmacological point of view. We include primary sensory neurons, because they represent, just as the striatum, an easily accessible model system.
Primary sensory neurons and the dorsal horn
Primary sensory neurons were one of the first systems to be shown to contain substantial amounts of peptides in the mid 1970’s and have ever since represented a model for analysis of peptidergic mechanisms. It is now clear that peptides in primary sensory neurons can be divided into at least two groups (62), those which are present in substantial amounts under normal circumstances and which in all probability facilitate transmission in the dorsal horn. They include substance P, CGRP and somatostatin and are mainly down-regulated after axotomy (Table 5). In contrast, peptides such as VIP/peptide histidine isoleucine (PHI), galanin and NPY are normally expressed at low levels or can not be detected at all, but are dramatically increased after experimental manipulation, especially axotomy (60, 149, 172) (Table 5). This has been shown to be an oversimplification since after axotomy both substance P (129) and CGRP (117) have been shown to be upregulated in subpopulations of DRG neurons. Interestingly, galanin is strongly expressed during early embryogenesis and then seems to be strongly downregulated (177). There are also impressive regulations of the mRNAs for cholecystokinin (CCK)B (185) and NPY Y1 (187) and Y2 (186) receptors in dorsal root ganglion neurons after axotomy. Provided that the increase or decrease in mRNA levels results in corresponding changes in receptor protein and incorporation into the neuronal membrane, these findings suggest that changes in sensitivity to a certain peptide represent a further principle to adapt to a lesion. Taken together, primary sensory neurons change their phenotype both with regard to messengers, receptors and function after peripheral nerve injury, the implications being that DRG neurons adapt to the new situation by suppressing excitatory transmitters, enhancing inhibitory mechanisms and promoting survival and regenerative mechanisms. This is in agreement with the general view of the reaction of neurons in response to injury, mainly based on studies on motoneurons (see 82), emphasizing that the synthetic machinery of the neuron is reprogrammed from transmitter synthesis to production of molecules of importance for survival and recovery. It is an interesting hypothesis that similar changes may occur when neurons in the brain are injured and eventually degenerate. For further information, (see Refs. 62, 175).
Peptide regulation in the rat striatum
The striatum occupies a central position in basal ganglia, not at least due to its rich dopaminergic innervation and its involvement in several serious central nervous system disorders such as Parkinson's disease and Huntington's chorea. Although much interest has been focused on dopamine and its functional role, it has become clear that also neuropeptides play an important role in this system (see 44, 47). In addition to the fact that many dopamine neurons themselves contain peptides such as CCK, several neuron populations in the striatum express peptides at higher or lower levels. Two major populations of striatal projection neurons exist, one containing substance P and dynorphin directed to the substantia nigra zona reticulata, whereas enkephalin immunoreactive neurons mainly seem to project to the globus pallidus (44, 47). Double staining techniques have demonstrated that many of these peptide neurons utilize GABA as their principal transmitter (136).
There is now strong evidence that the three peptides mentioned above, enkephalin, substance P and dynorphin, are regulated by the dopaminergic input from the substantia nigra. Thus, various manipulations attenuating dopamine transmission in the striatum, for example lesioning the nigrostriatal pathway with 6-hydroxydopamine or treatment with dopamine receptor antagonists, result in increased expression of enkephalin, and a decrease in dynorphin and substance P, whereas dopamine agonists increase levels of dynorphin and substance P but not enkephalin (see 43).
More recent studies suggest that also neurotensin in the striatum is dopamine-regulated. This peptide has been shown to have interesting both clinical and behavioural interactions with dopamine (78, 126). It has been shown that drugs that attenuate dopamine transmission, such as haloperidol, 6-hydroxydopamine and reserpine upregulate neurotensin in many cell bodies in the rat striatum (see 29), an effect mediated via D2 receptors. However, also dopamine receptor stimulation (with methamphetamine) causes upregulation of neurotensin in basal ganglia, in this case via a D1 receptor (4, 93, 109-111, 169). D1 regulated neurotensin is present in striato-nigral neurons, whereas the D2 regulated neurotensin occurs in neurons projecting to the globus pallidus (22, 23)]. All these studies are in agreement also with the general view that striato-nigral neurons are mainly under the control of the D1 receptors and that the striato-pallidal neurons contain D2 receptors (42, 53, 89, 90, 140, 141). Schiffmann and Vanderhaeghen (147) have recently demonstrated upregulation of neurotensin mRNA in the striatum after chronic injection of caffeine.
Of the many central peptides that recently have received increased attention, we would here like to only mention some involved in the central control of feeding and body weight. These peptides include CCK, NPY and galanin as well as the orexins, melanocortins, MCH, AGRP and CART peptides, which either stimulate or attenuate food intake(see 92, 176, 71, see also 83, 85). This field has been greatly stimulated by the discovery of several receptor subtypes, especially for NPY (see 5, 88), which could provide selectivity for functions such as food intake regulation, as well as by the discovery of leptin, which is an adipose tissue derived signaling factor encoded by the obesity (OB) gene (188). The circuitries involved in regulation of feeding behavior are now being worked out (see 18, 35, 146).
Neuropeptide knock-out mice
One approach to understanding the functional role of proteins in general is the creation of knock-out mice, that is deletion of specific genes, using modern molecular biological tools. This has been practiced only to a limited extent in the neuropeptide field, but more recently several knock-out animals have been described including such where the gene for NPY and some of its receptors receptors have been deleted (6, 36, 103, 134, 135). There are also knock-out mice where substance P (21, 190) as well as its NK-1 (24) receptor have been deleted. The phenotypic changes observed are not as dramatic as might have been expected on the basis of pharmacological and morphological studies, possibly due to compensatory mechanisms, but are still very interesting. Thus, the substance P/NK1 receptor knockouts show altered nociception (21, 24, 190). Perhaps surprisingly, little effect on weight was observed in the NPY knock-out mouse (36). However, leptin-deficient ob/ob mice rendered NPY-deficient by crossing with NPY-KO mice have a much attenuated obese phenotype as compared to their NPY-expressing control (see 134). This elegant experiment suggests that combinatorial mutants may provide the greatest use for neuropeptide knockouts. In the Y1R knock-out daily food intake, as well as NPY-stimulated feeding, were only slightly diminished, whereas fast induced refeeding was markedly reduced (135). The Y1-R knock-out had increased body fat with no change in protein content. The Y5-R knock-out showed a late onset obesity characterized by increased body weight, food intake and adiposity (103). Whereas Y1 knock-out mice had a lower metabolic rate resulting in increased fat deposit, the Y5 knock-out were rather hyperphagic. The reaction of NPY-deficient mice in an epileptic seizure model was interesting, in that kainic acid induced essentially quantitatively similar seizures in both knock-out and wild-type mice, but that seizures did not terminate in the knock-out mice and virtually all of them died (6). This suggests an important role of NPY in the control of electrical activity in brain (6). Finally, NPY-deficient mice are more prone to ethanol consumption (164).
An alternative route to knock-out by homologous recombinantion was taken by Piccioli et al. (137) who generated transgenic mice expressing antibodies to substance P and showed a marked inhibition of neurogenic inflammation and motor deficits.
Trophic effects of peptides
As described above, it is likely that peptides participate in signalling at non-synaptic and synaptic sites, primarily in slow signalling. However, increasing evidence indicate that they also exert trophic actions (58, 159). Although most results on the latter effects have been obtained in the periphery, it is likely that similar peptide actions also occur in the central nervous system. To mention a few recent studies, VIP has been shown to have a dramatic effect on growth of whole fetuses in vitro (49) which may be due to the fact that it shortens the G1 and S phases of the neural cell cycle (50). CGRP may represent an anterograde factor which after release at the motor end-plate regulates gene expression of acetylcholine receptors (40). CGRP has in another model system been reported to induce a dopaminergic phenotype in olfactory bulb neurons, thus mimicking the olfactory epithelium neurons and perhaps representing a differentiation factor for dopamine neurons (28).
It was early noticed that several peptides, for example somatostatin, can only be seen during the embryonic period in certain systems and then disappear, suggesting a developmental role (see 165). More recently, de Felipe et al. (25) have provided involvement for substance P and NK1 receptors during embryogenesis in the spinal cord, where transiently expressed substance P is released from pioneer neuronal pathways and via the NK1 receptor in floor plate cells regulates the release of a chemoattractant to guide the permanent commissural axons to the midline.
Even in the beginning of the peptide era it was predicted that peptides are involved in disease, and in particular that drugs interfering with peptidergic mechanisms could have therapeutic effects. For example, the realization that substance P is present in small diameter fibers of presumptive nociceptive sensory neurons, and the discovery of endogenous ligands for the opiate receptors suggested a rapid progress in understanding and treatment of various pain states. To date this prediction has not materialized. The insight that most hypothalamic releasing and inhibitory factors are peptides clearly suggested neuroendocrine therapeutical applications. In fact, some of the hypothalamic releasing peptides such as TRH are used diagnostically to probe for pituitary function. Moreover, the demonstration that the growth hormone inhibiting hormone, somatostatin, has a wide spread distribution and powerful inhibitory effects, initiated work to produce more powerful and long-lasting somatostatin analogues. In fact, such analogues are now used to treat prolactinoma as well as other peripheral endocrine tumors.
Little information is also available on the role of neuropeptides in humans. Nevertheless, it is unquestionable that morphine, acting on opioid peptide receptors, is the most efficient analgesic drug so far known, giving witness to the importance of peptide receptors as drug targets, in this case to alleviate pain. More recently focus has also been on involvement of opioid receptors in alcohol dependence. Thus, treatment with the opioid receptor antagonist naltrexone suggests that these types of receptors may be of importance for alcohol dependency and that this drug can be used successfully in the treatment of alcoholism (130, 168).
Much focus has been on possible involvement on CCK and CCK receptors in anxiety, and it has been shown in man that CCK-4-induced panic attacks can be significantly attenuated by the CCKB antagonist CI-988 (17), although several not published studies using such CCKB antagonists given to patients with anxiety appear to have failed to reveal significant effects.
An apparent breakthrough has, however, as mentioned above, occurred in the field of substance P. Thus, the Merck NK-1 receptor antagonist MK869 has demonstrated excellent antidepressant and anxiolytic activity when given alone and orally once daily (81). In fact, it is at least as efficient as a serotonin uptake inhibitor prototype (paroxetine), and has fewer side effects, in particular with regard to sexual dysfunctions and nausea. This appears to be first evidence for derangement of a central neuropeptide system in one of the major mental diseases and the possibility to orally treat this with a peptide antagonist of the non-peptide type which passes the blood brain barrier. It is possible that this represents only the first example of such a drug and is likely to stimulate pharmaceutical companies to further efforts in this direction.
Major progress has been made in the field of neuropeptides since the publication of "Psychopharmacology. The Third Generation of Progress" (108). The peptides have found their receptors, powerful drugs have been developed and novel insights into the regulation of peptide synthesis have been obtained, including the provocative finding of mRNA in axonal processes. Still, the physiological role of neuropeptides are not well defined, and both transmitter-like functions, modulation and trophic actions have to be considered. With the improved tools now available it should be possible to clarify many of these open questions. The use of peptide agonists and antagonists should help to elucidate the function of neuropeptides in systems of importance for psychopharmacology.
These studies were supported by the Swedish MRC (04X-2807), Marianne and Marcus Wallenberg’s Foundation, Knut and Alice Wallenberg’s Foundation, and a Bristol-Myers Squibb Unrestricted Neuroscience Grant. We thank Professor S. Nakanishi, Kyoto University, Kyoto, Japan for providing Fig. 2.