Since the early discovery of Dahlstrom and Fuxe (1964), several lines of evidence indicate that the cell bodies of 5-HT neurons are concentrated in the midbrain. Of the 5- HT cell groups, the dorsal and median raphe nuclei have been most studied; at least half of all 5-HT neurons in the CNS are located in the dorsal raphe and almost 70% of the cells in the raphe contain 5-HT (Wiklund and Bjorklund, 1980; Wiklund et al., 1981). 5- HT neurons are medium sized cells with spiny dendritic arborization, and 5-HTcontaining dendrites synapse upon other 5-HT-containing dendrites. Serotonergic interconnections between various raphe nuclei have also been demonstrated (Chazal and Ralston, 1987). Besides dendrodendritic interactions, the activity of 5-HT neurons may also be regulated locally by recurrent axon collaterals (Harandi et al., 1987), and synaptic afferents found on 5-HT dendrites and cell bodies are often serotonergic. Few serotonergic nerve endings have synaptic specializations, suggesting that most 5-HT is released in a non-synaptic manner (Harandi et al., 1987). 5-HT neurons in the raphe fire spontaneously at 1-5 spikes/s (Vandermaelen and Aghajanian, 1983), and this activity is determined by at least three different processes: spontaneous, slow firing activity; 5-HT mediated autoinhibition; and afferent inputs mediated by non-5-HT receptors (Pineyro and Blier, 1999). Autoinhibition includes regulation of transmitter release; and synthesis and neural firing which are determined by various metabotropic (5-HTIA/IB/ID, 5-HT2A/2C, 5-HT7) and ionotropic (5-HT3) 5-HT receptors (Zifa and Pillion, 1992; Barnes and Sharp, 1999). Different 5-HT1 receptor subtypes are present on 5-HT neurons in the raphe and they inhibit 5-HT release (Davidson and Stamford, 1995; Sprouse and Aghajanian, 1987; Starkey and Skingle, 1994; Pineyro and Blier, 1996) although their regulatory functions are different. In the raphe nuclei, few data are available about the 5-HT receptors that mediate actions on neurons other than 5-HT neurons. Postsynaptic 5-HT receptors are the target of 5-HT released from dendritic varicosities into the synaptic gap or into the extracellular space, as both synaptic and non-synaptic 5-HT neurotransmission may occur (Chazal and Ralston, 1987; Harandi et al, 1987). 5-HT release in the raphe originates from the somata/dendrites and from recurrent axon collaterals, and the regulation of release from these sites may be different. Dendritic release originates from 5-HT neurons with vesicles in the dendrites (Hery et al., 1982) and may provide the neurochemical basis for synchronous firing among 5-HT cells. Activation of ascending 5-HT pathways inhibits 5-HT cells of the dorsal raphe, an effect mediated by 5-HT (Hajos et al., 1998); iontophoretically applied 5-HT also inhibits raphe neurons. The cycles of hyper- and depolarization of 5-HT cells are altered by excitatory and inhibitory neuronal projections and intemeurons besides 5-HT released from 5-HT neurons (Pineyro and Blier, 1996). The neural circuitry of the raphe converts incoming signals into output messages. The 5-HT projection neurons, GABA intemeurons and glutamatergic axon terminals form complex excitatory-inhibitory connections by which incoming excitatory signals are converted into inhibitory output and sent back to the areas where stimulation was generated. The frinctional state of this neural network is determined by neurotransmitters released both synaptically and non-synaptically (Vizi, 2000). Ascending 5-HT projections influence several fimctions in different forebrain structures (Parent et al., 1981; Steinbusch, 1981). The raphe-hippocampal pathway is one of numerous ascending 5-HT projections (El Mansari and Blier, 1996), and the caudate nucleus is another important projection field for 5-HT neurons of the raphe nuclei (Van der Kooy and Hattori, 1980; Vizi et al., 1981). 5-HT neurons of the raphe also form important projections to the median cerebral cortex (Kidd et al., 1991; Hajos et al., 1998), mediating inhibition in the cortical neural network (Steinbusch, 1981; O'Heam and Molliver, 1984). In many cases, forebrain structures that receive 5-HT innervation from the raphe send back reciprocal projections; for example, there is a glutamate pathway from the median prefrontal cortex to the raphe nuclei (Behzadi et al., 1990). GABA intemeurons in the raphe nuclei are the primary target for the cortico-raphe glutamate neurons, and their stimulation by glutamate leads to inhibition of 5-HT cells (Hajos et al., 1998). 5-HT neurons in the raphe are regulated by a number of efferent projections. The habenulo-raphe projection represents a major GABA inhibition of raphe neurons (Stem et al., 1981; Wang et al., 1992), this pathway also contains excitatory transmitters and the presence of substance P-containing and glutamatergic fibers has been reported. Another excitatory input to raphe 5-HT neurons is noradrenergic, acting through postsynaptic a2- adrenoceptors (Baraban and Aghajanian, 1981). This review discusses the fimctional relevance of somatodendritic 5-HT release and it contrasts the release properties between axon terminal and mindbrain raphe nuclei, and considers possible extrasynaptic targets.
ASJC Scopus subject areas