Regional distribution of calcium elevation during sensory transduction in spider mechanoreceptor neurons

Spider mechanosensory VS-3 neurons receive peripheral efferent synaptic modulation, with regional variations in the types of efferent synapses and transmitter receptors. VS-3 somata possess a voltage-activated calcium current, but the levels and time courses of calcium changes in other regions are unknown. The roles of calcium in these neurons are not completely understood, but could include modulation of both mechanosensitivity and response dynamics. Here, we measured calcium concentration rises caused by single, mechanically induced action potentials in VS-3 sensory dendrites, somata and axons, using Oregon Green BAPTA-1 fluorescence. Calcium concentration rose by approximately 1 nM following each action potential. Time courses of calcium rise and fall were similar in the three regions but the rise in amplitude was about 50% higher in the sensory dendrite than in the soma. Antibody to the Ca(V)3.1(alpha(1g)) isotype of T-type calcium channel labeled all three neuronal regions. Some Ca(V)3.1 labeling colocalized with synapsin labeling, suggesting that calcium channels play some part in efferent modulation. We conclude that mechanically stimulated action potentials start near sensory dendrite tips and pass rapidly through the neurons to the axons, activating low voltage activated calcium channels in all three regions and causing calcium concentration to rise rapidly in each region. These results suggest important roles for calcium in several stages of mechanosensation.     

Distribution of calcium channel CaV1.3 immunoreactivity in the rat spinal cord and brain stem

The function of local networks in the CNS depends upon both the connectivity between neurons and their intrinsic properties. An intrinsic property of spinal motoneurons is the presence of persistent inward currents (PICs), which are mediated by non-inactivating calcium (mainly Ca_V1.3) and/or sodium channels and serve to amplify neuronal input signals. It is of fundamental importance for the prediction of network function to determine the distribution of neurons possessing the ion channels that produce PICs. Although the distribution pattern of Ca_V1.3 immunoreactivity (Ca_V1.3-IR) has been studied in some specific central nervous regions in some species, so far no systematic investigations have been performed in both the rat spinal cord and brain stem. In the present study this issue was investigated by immunohistochemistry. The results indicated that the Ca_V1.3-IR neurons were widely distributed across different parts of the spinal cord and the brain stem although with variable labeling intensities. In the spinal gray matter large neurons in the ventral horn (presumably motoneurons) tended to display higher levels of immunoreactivity than smaller neurons in the dorsal horn. In the white matter, a subset of glial cells labeled by an oligodendrocyte marker was also Ca_V1.3-positive. In the brain stem, neurons in the motor nuclei appeared to have higher levels of immunoreactivity than those in the sensory nuclei. Moreover, a number of nuclei containing monoaminergic cells, for example the locus coeruleus, were also strongly immunoreactive. Ca_V1.3-IR was consistently detected in the neuronal perikarya regardless of the neuronal type. However, in the large neurons in the spinal ventral horn and the cranial motor nuclei the Ca_V1.3-IR was clearly detectable in first and second order dendrites. These results indicate that in the rat spinal cord and brain stem Ca_V1.3 is probably a common calcium channel used by many kinds of neurons to facilitate the neuronal information processing via certain intracellular mechanisms, for instance, PICs.     

Serotonin acts in the synaptic region of sensory neurons in Aplysia to enhance transmitter release

An important mechanism that contributes to sensitization in Aplysia is heterosynaptic facilitation of the synaptic connections between sensory neurons (SNs) and motor neurons (MNs). Heterosynaptic facilitation, in turn, is associated with broadening of the spike in the SN. Spike broadening is readily observed in recordings from somata of SNs, and from growth cones of SNs in culture, but broadening in synaptic terminals has only been inferred. Intracellular recordings were made from somata of SNs and from somata of follower MNs. Additional recordings were made from the axons of SNs as they enter the neuropil in the pedal ganglion. Serotonin (5-HT) broadened action potentials in axons of SNs and enhanced excitatory postsynaptic potentials (EPSPs) in the MNs, even after the axons of SNs were surgically separated from their somata. These results indicate that both heterosynaptic facilitation and spike broadening in the axon are due to the local action of 5-HT and can occur independently of modulation of membrane properties in the soma.     

Modulation of T-type calcium channels by bioactive lipids

T-type calcium channels (T-channels/Ca_V3) have unique biophysical properties allowing a calcium influx at resting membrane potential of most cells. T-channels are ubiquitously expressed in many tissues and contribute to low-threshold spikes and burst firing in central neurons as well as to pacemaker activities in cardiac cells. They also emerged as potential targets to treat cancer and hypertension. Regulation of these channels appears complex, and several studies have indicated that Ca_V3.1, Ca_V3.2, and Ca_V3.3 currents are directly inhibited by multiple endogenous lipids independently of membrane receptors or intracellular pathways. These bioactive lipids include arachidonic acid and ω3 poly-unsaturated fatty acids; the endocannabinoid anandamide and other N-acylethanolamides; the lipoamino-acids and lipo-neurotransmitters; the P450 epoxygenase metabolite 5,6-epoxyeicosatrienoic acid; as well as similar molecules with 18–22 carbons in the alkyl chain. In this review, we summarize evidence for direct effects of these signaling molecules, the molecular mechanisms underlying the current inhibition, and the involved chemical features. The impact of this modulation in physiology and pathophysiology is discussed with a special emphasis on pain aspects and vasodilation. Overall, these data clearly indicate that T-current inhibition is an important mechanism by which bioactive lipids mediate their physiological functions.     

Localization of L-type calcium channel CaV1.3 in cat lumbar spinal cord – with emphasis on motoneurons

Voltage-dependent persistent inward currents (PICs) which underlie the plateau potentials are an important intrinsic property of spinal motoneurons. Electrophysiological experiments have indicated that a subtype of the low threshold L-type calcium channel, Ca_V1.3, mediates this current. In mouse and turtle lumbar spinal cord it has been shown that these channel proteins are mainly found on motoneuron dendrites. In the present study we have used immunohistochemistry to locate these channels in lumbar spinal neurons, especially motoneurons, of the cat. The results indicate that Ca_V1.3 immunoreactivity was unevenly distributed among the laminae of the spinal grey matter. The small neurons in superficial dorsal horn (laminae I–III) were sparsely and weakly labelled, while large neurons in ventral horn were frequently and densely labelled. Groups of motoneurons in lamina IX that were immunoreactive to choline acetyltransferase also co-expressed Ca_V1.3. The immmunoreactivity was mainly associated with neuronal somata and proximal dendrites. Double staining with antibodies against Ca_V1.3 and MAP2 (a dendritic marker) showed that some fine fibres, which may include distal dendrites, were also labelled. These results in the cat spinal cord show some differences from studies in mouse and turtle motoneurons where the immunoreactivity against this channel was mainly localized to the dendrites.     

Beta amyloid peptide plaques fail to alter evoked neuronal calcium signals in APP/PS1 Alzheimer's disease mice

Alzheimer's disease (AD) is a multifactorial disorder of unknown etiology. Mechanistically, beta amyloid peptides (Aβ) and elevated Ca²⁺ have been implicated as proximal and likely interactive features of the disease process. We tested the hypothesis that proximity to Aβ plaque might exacerbate activity-dependent neuronal Ca²⁺ signaling in hippocampal pyramidal neurons from APP_SWE/PS1_M146V mice. Using combined approaches of whole cell patch clamp recording and 2-photon imaging of neuronal Ca²⁺ signals with thioflavin-S plaque labeling in hippocampal slices, we found no correlation between thioflavin-S labeled Aβ plaque proximity and Ca²⁺ responses triggered by ryanodine receptor (RyR) activation or action potentials in either dendrites or somata of AD mice, regardless of age. Baseline and RyR-stimulated spontaneous excitatory postsynaptic potentials also showed little difference in relation to Aβ plaque proximity. Consistent with previous studies, RyR-evoked Ca²⁺ release in APP_SWE/PS1_M146V mice was greater than in nontransgenic controls. Within the soma, RyR-evoked Ca²⁺ release was elevated in older APP_SWE/PS1_M146V mice compared with younger APP_SWE/PS1_M146V mice, but was still independent of plaque proximity. The results indicate that early Ca²⁺ signaling disruptions can become yet more severe with age through mechanisms independent of Aβ plaques, suggesting that alternative pathogenic mechanisms might contribute to AD-associated dysfunction.     

CaV2.1 channels are modulated by muscarinic M1 receptors through phosphoinositide hydrolysis in neostriatal neurons

In adult neostriatal projection neurons, the intracellular Ca²⁺ supplied by Ca_V2.1 (P/Q) Ca²⁺ channels is in charge of both the generation of the afterhyperpolarizing potential (AHP) and the release of GABA from their synaptic terminals, thus being a major target for firing pattern and transmitter release modulations. We have shown that activation of muscarinic M₁-class receptors modulates Ca_V2.1 channels in these neurons in rats. This modulation is reversible, is not membrane delimited, is blocked by the specific M₁-class muscarinic antagonist muscarine toxin 7 (MT-7), and is neither mediated by protein kinase C (PKC) nor by protein phosphatase 2B (PP-2B). Hence, the signaling mechanism of muscarinic Ca_V2.1 channel modulation has remained elusive. The present paper shows that inactivation of phospholipase C (PLC) abolishes this modulation while inhibition of phosphoinositide kinases, PI-3K and PI-4K, prevents its reversibility, suggesting that the reconstitution of muscarinic modulation depends on phosphoinositide rephosphorylation. In support of this hypothesis, the supply of intracellular phosphatidylinositol (4,5) bisphosphate [PI(4,5)P₂] blocked all muscarinic modulation of this channel. The results indicate that muscarinic M₁ modulation of Ca_V2.1 Ca²⁺ channels in these neurons involves phosphoinositide hydrolysis.     

Temperature sensitivity of transduction and action potential conduction in a spider mechanoreceptor

Previous work has suggested that the activation energy of mechanotransduction is higher than expected from the simple electrochemistry of ion channels, but the temperature sensitivity of mechanically activated receptor current has not been measured directly before. We used the single-electrode voltage-clamp technique to measure receptor currents in sensory neurons of the VS-3 slit-sense organ in the spider, Cupiennius salei. Receptor currents were generated by deforming the cuticular slits. Conduction velocity in afferent axons from the same organ was also measured by recording action potentials at two locations in the leg during mechanical stimulation of the slits. Activation energies of mechanotransduction and conduction velocity were estimated by making the measurements at a range of temperatures. The mean activation energy for receptor current was 23.1 kcal/mol (96.6 kJ/mol), corresponding to a Q10 value of 3.2. Conduction velocity in the afferent axons was approximately equal to 5 m/s at room temperature and it was much less temperature sensitive, with an activation energy of 6.3 kcal/mol (26.3 kJ/mol), corresponding to a Q10 value of 1.5. These results provide the first direct measurements of the activation energy of mechanically activated currents and support previous suggestions that a high thermal energy barrier is involved in mechanotransduction.     

Subcellular localization of the voltage-gated potassium channels Kv3.1b and Kv3.3 in the cerebellar dentate nucleus of glutamic acid decarboxylase 67-green fluorescent protein transgenic mice

Deep cerebellar dentate nuclei are in a key position to control motor planning as a result of an integration of cerebropontine inputs and hemispheric Purkinje neurons signals, and their influence through synaptic outputs onto extracerebellar hubs. GABAergic dentate neurons exhibit broader action potentials and slower afterhyperpolarization than non-GABAergic (presumably glutamatergic) neurons. Specific potassium channels may be involved in these distinct firing profiles, particularly, Kv3.1 and Kv3.3 subunits which rapidly activate at relatively positive potentials to support the generation of fast action potentials. To investigate the subcellular localization of Kv3.1b and Kv3.3 in GAD- and GAD+ dentate neurons of glutamic acid decarboxylase 67-green fluorescent protein (GAD67-GFP) knock-in mice a preembedding immunocytochemical method for electron microscopy was used. Kv3.1b and Kv3.3 were in membranes of cell somata, dendrites, axons and synaptic terminals of both GAD- and GAD+ dentate neurons. The vast majority of GAD- somatodendritic membrane segments domains labeled for Kv3.1b and Kv3.3 (96.1% and 84.7%, respectively) whereas 56.2% and 69.8% of GAD- axonal membrane segments were immunopositive for these subunits. Furthermore, density of Kv3.1b immunoparticles was much higher in GAD- somatodendritic than axonal domains. As to GAD+ neurons, only 70.6% and 50% of somatodendritic membrane segments, and 53.3% and 59.5% of axonal membranes exhibited Kv3.1b and Kv3.3 labeling, respectively. In contrast to GAD- cells, GAD+ cells exhibited a higher density labeling for both Kv3 subunits at their axonal than at their somatodendritic membranes. Taken together, Kv3.1b and Kv3.3 potassium subunits are expressed in both GAD- and GAD+ cells, albeit at different densities and distribution. They likely contribute to the distinct biophysical properties of both GAD- and GAD+ neurons in the dentate nucleus.     

Short- and long-term roles of phosphatidylinositol 4,5-bisphosphate PIP2 on Cav3.1- and Cav3.2-T-type calcium channel current

T-type calcium (Ca²⁺) channels play important physiological functions in excitable cells including cardiomyocyte. Phosphatidylinositol-4,5-bisphosphate (PIP₂) has recently been reported to modulate various ion channels’ function. However the actions of PIP₂ on the T-type Ca²⁺ channel remain unclear. To elucidate possible effects of PIP₂ on the T-type Ca²⁺ channel, we applied patch clamp method to investigate recombinant Ca_V3.1- and Ca_V3.2-T-type Ca²⁺ channels expressed in mammalian cell lines with PIP₂ in acute- and long-term potentiation. Short- and long-term potentiation of PIP₂ shifted the activation and the steady-state inactivation curve toward the hyperpolarization direction of Ca_V3.1-I_Ca.T without affecting the maximum inward current density. Short- and long-term potentiation of PIP₂ also shifted the activation curve toward the hyperpolarization direction of Ca_V3.2-I_Ca.T without affecting the maximum inward current density. Conversely, long-term but not short-term potentiation of PIP₂ shifted the steady-state inactivation curve toward the hyperpolarization direction of Ca_V3.2-I_Ca.T. Long-term but not short-term potentiation of PIP₂ blunted the voltage-dependency of current decay Ca_V3.1-I_Ca.T. PIP₂ modulates Ca_V3.1- and Ca_V3.2-I_Ca.T not by their current density but by their channel gating properties possibly through its membrane-delimited actions.     

Determinants of the voltage dependence of G protein modulation within calcium channel β subunits

Ca_Vβ subunits of voltage-gated calcium channels contain two conserved domains, a src-homology-3 (SH3) domain and a guanylate kinase-like (GK) domain with an intervening HOOK domain. We have shown in a previous study that, although Gβγ-mediated inhibitory modulation of Ca_V2.2 channels did not require the interaction of a Ca_Vβ subunit with the Ca_Vα1 subunit, when such interaction was prevented by a mutation in the α1 subunit, G protein modulation could not be removed by a large depolarization and showed voltage-independent properties (Leroy et al., J Neurosci 25:6984–6996, 2005). In this study, we have investigated the ability of mutant and truncated Ca_Vβ subunits to support voltage-dependent G protein modulation in order to determine the minimal domain of the Ca_Vβ subunit that is required for this process. We have coexpressed the Ca_Vβ subunit constructs with Ca_V2.2 and α₂δ-2, studied modulation by the activation of the dopamine D2 receptor, and also examined basal tonic modulation. Our main finding is that the Ca_Vβ subunit GK domains, from either β1b or β2, are sufficient to restore voltage dependence to G protein modulation. We also found that the removal of the variable HOOK region from β2a promotes tonic voltage-dependent G protein modulation. We propose that the absence of the HOOK region enhances Gβγ binding affinity, leading to greater tonic modulation by basal levels of Gβγ. This tonic modulation requires the presence of an SH3 domain, as tonic modulation is not supported by any of the Ca_Vβ subunit GK domains alone. Andriy V. Dresviannikov and Karen M. Page contributed equally to this work.     

Cysteines in the loop between IS5 and the pore helix of CaV3.1 are essential for channel gating

The role of six cysteines of Ca_V3.1 in channel gating was investigated. C241, C271, C282, C298, C313, and C323, located in the extracellular loop between segment IS5 and the pore helix, were each mutated to alanine; the resultant channels were expressed and studied by patch clamping in HEK293 cells. C298A and C313A conducted calcium currents, while the other mutants were not functional. C298A and C313A as well as double mutation C298/313A significantly reduced the amplitude of the calcium currents, shifted the activation curve in the depolarizing direction and slowed down channel inactivation. Redox agents dithiothreitol (DTT) and 5,5′-dithiobis(2-nitrobenzoic acid) (DTNB) shifted the current activation curve of wild-type channels in the hyperpolarizing direction. Activation curve for all mutated channels was shifted in hyperpolarizing direction by DTT while DTNB caused a depolarizing shift. Our study reveals that the cysteines we studied have an essential role in Ca_V3.1 gating. We hypothesize that cysteines in the large extracellular loop of Ca_V3.1 form bridges within the loop and/or neighboring channel segments that are essential for channel gating.     

Light and electron microscopic analysis of the somata and parent axons innervating the rat upper molar and lower incisor pulp

The morphology of intradental nerve fibers of permanent teeth and of continuously growing rodent incisors has been studied in detail but little information is available on the parent axons that give rise to these fibers. Here we examined the axons and somata of trigeminal neurons that innervate the rat upper molar and lower incisor pulp using tracing with horseradish peroxidase and light and electron microscopic analysis. The majority (∼80%) of the parent axons in the proximal root of the trigeminal ganglion that innervated either molar or incisor pulp were small myelinated fibers (<20 μm² cross-sectional area). The remaining ∼20% of the fibers were almost exclusively large myelinated for the molar pulp and unmyelinated for the incisor pulp. The majority of neuronal somata in the trigeminal ganglion that innervated either molar (48%) or incisor pulp (62%) were medium in size (300–600 μm² cross-sectional area). Large somata (>600 μm²) constituted 34% and 20% of the trigeminal neurons innervating molar and incisor pulp, respectively, while small somata (<300 μm²) constituted 17% of the molar and 18% of the incisor neurons. The present study revealed that the morphology of parent axons of dental primary sensory neurons may differ from that of their intradental branches, and also suggests that the nerve fiber function may be carried out differently in the molar and incisor pulp in the rat.     

The central nervous system efferent control of the organs of balance and equilibrium.

The vestibular labyrinth is innervated by both primary afferent nerves and efferent axons with cell bodies located in the central nervous system. Efferent terminals are found on both hair cells and on primary afferent axons. Acetylcholine is the major efferent transmitter, but enkephalin and calcitonin gene-related peptide (CGRP) have also been localized to efferent terminals and somata. The efferent vestibular nuclei are bilaterally organized in the majority of species. Semicircular canal primary afferents have been classified by their sensitivity and phase in response to rotation. Electrical activation of efferents in monkey and fish increases afferent resting discharge and reduces afferent gain to adequate stimulation. Effects are most profound on high-gain, phase-advanced (re. velocity) afferents. Experiments in alert animals indicate that multiple sensory modalities can activate the efferent system.     

Efferent projections from the median preoptic nucleus to sleep- and arousal-regulatory nuclei in the rat brain

The median preoptic nucleus (MnPO) has been implicated in the regulation of hydromineral balance and cardiovascular regulation. The MnPO also contains neurons that are active during sleep and in response to increasing homeostatic pressure for sleep. The potential role of these neurons in the regulation of arousal prompted an analysis of the efferent projections from the MnPO. Anterograde and retrograde neuroanatomical tracers were utilized to characterize the neural connectivity from the MnPO to several functionally important sleep- and arousal-regulatory neuronal systems in the rat brain. Anterograde terminal labeling from the MnPO was confirmed within the core and extended ventrolateral preoptic nucleus. Within the lateral hypothalamus, labeled axons were observed in close apposition to proximal and distal dendrites of hypocretin/orexin immunoreactive (IR) cells. Projections from the MnPO to the locus coeruleus were observed within and surrounding the tyrosine hydroxylase-IR cell cluster. Labeled axons from the MnPO were mostly observed within the lateral division of the dorsal raphé nucleus and heavily within the ventrolateral periaqueductal gray. Few anterogradely labeled appositions were present juxtaposed to choline acetyltransferase-IR somata within the magnocellular preoptic area. The use of retrogradely transported neuroanatomical tracers placed within the prospective efferent terminal fields supported and confirmed findings from the anterograde tracer experiments. These anatomical findings support the hypothesis that MnPO neurons function to promote sleep by inhibition of orexinergic and monoaminergic arousal systems and disinhibition of sleep regulatory neurons in the ventrolateral preoptic area.     

Central organization of the efferent supply to the labyrinthine and lateral line receptors of the dogfish

Neurons that provide the efferent innervation of the inner ear and lateral line were located in the brain by applying horseradish peroxidase to appropriate cranial nerves. The efferent neurons are found in a rhombencephalic nucleus, called here the octavolateralis efferent nucleus, which lies at the rostral pole of the visceromotor column. Up to 40% of these neurons are contralateral. The location of efferent cells is not topographically related to the sense organs they innervate. Their axons leave the brain together with other motor axons in the facial and glossopharyngeal nerves and there is evidence that some neurons innervate both the ear and the lateral line. The efferent nucleus receives direct sensory input from the labyrinth, but not from the lateral line. The organization of the efferent neurons and the distribution of their axons indicates that their effect on the sense organs is probably widespread and nonspecific.     

Temperature neurons in the crotaline trigeminal ganglia.

1. Intrasomal recordings were made with microelectrodes from 153 warm (infrared) neurons in the trigeminal ganglia of 36 crotaline snakes, Trimeresurus flavoviridis. Background discharges were observed at room temperature. The 153 warm neurons were classified into two groups: 81 were sensitive to less than or equal to 10 mg of von Frey hair mechanical stimulation (warm T + M neuron), and 72 were insensitive to up to 100 mg or more of mechanical stimulation (warm T neuron). For T + M and T neurons the receptive fields were all located in the pit organ. The mechanically sensitive field of warm T + M neurons located within the infrared receptive field on the pit membrane was less than 1 mm in diameter, and there was only one field per neuron. 2. Electrophysiological parameters were measured. These measurements included membrane potential, action potential amplitude, time of peaking, time duration at the resting membrane potential level, afterhyperpotential (AHP) height and AHP time to half-decay, and maximum rates of depolarization and repolarization. No difference in action potential parameters between the means of these two submodality groups was observed. 3. Intracellular horseradish peroxidase (HRP) labeling was used for defining the warm neuron profile. The somata of warm T and warm T + M neurons and T- or Y-shaped bifurcations of the axon were observed in the ganglion. At the bifurcation point, nodes of Ranvier were observed, but without broad triangular expansion. Diameters of the central axons were thinner than those of the peripheral or stem axons. There were no differences between the mean diameters of the two submodalities. 4. The central axons of warm T and T + M neurons projected to the lateral descending nucleus of the trigeminal nerve (LTTD). Their synaptic boutons were found in the LTTD. No branching of the axons to the principal sensory nucleus or the descending nucleus of the trigeminal nerve was found. These results were the same for six warm T and eight warm T + M neurons. 5. Conduction velocities of the peripheral fibers were measured by stimulating superficial branches of the maxillary nerve electrically. Three groups of conduction velocity were identified in the compound potentials. The conduction velocity of the peak action potential of the warm T fibers was 6.9 +/- 1.2 (SD) m/s (n = 18), that of the T + M fibers 6.7 +/- 0.9 m/s (n = 23). These fell into the second group of the compound potentials.(ABSTRACT TRUNCATED AT 400 WORDS)     

Synaptic organization of immunocytochemically identified GABA neurons in the monkey sensory-motor cortex

Neurons in the monkey somatic sensory and motor cortex were labelled immunocytochemically for the GABA synthesizing enzyme, glutamic acid decarboxylase (GAD), and examined with the electron microscope. The somata and dendrites of many large GAD-positive neurons of layers III-VI receive numerous asymmetric synapses from unlabelled terminals and symmetric synapses from GAD-positive terminals. Comparisons with light and electron microscopic studies of Golgi-impregnated neurons suggest that the large labelled neurons are basket cells. Small GAD-positive neurons generally receive few synapses on their somata and dendrites, and probably conform to several morphological types. GAD-positive axons from symmetric synapses on many neuronal elements including the somata, dendrites and initial segments of pyramidal cells, and the somata and dendrites of non-pyramidal cells. Synapses between GAD-positive terminals and GAD-positive cell bodies and dendrites are common in all layers. Many GAD-positive terminals in layers III-VI arise from myelinated axons. Some of the axons form pericellular terminal nests on pyramidal cell somata and are interpreted as originating from basket cells while other GAD-positive myelinated axons synapse with the somata and dendrites of non-pyramidal cells. These findings suggest either that the sites of basket cell terminations are more heterogeneous than previously believed or that there are other classes of GAD-positive neurons with myelinated axons. Unmyelinated GAD-positive axons synapse with the initial segments of pyramidal cell axons or form en passant synapses with dendritic spines and small dendritic shafts and are interpreted as arising from the population of small GAD-positive neurons which appears to include several morphological types.     

Morphology of cortically projecting basal forebrain neurons in the rat as revealed by intracellular iontophoresis of horseradish peroxidase

The intracellular horseradish peroxidase technique was employed to study the morphology of basal forebrain neurons that were identified as cortically projecting by antidromic invasion from the cerebral cortex. Four neurons were examined in detail; they were located at different rostrocaudal levels within the basal forebrain. Their somata were large, 30-50 microns in longest dimension, and gave rise to three to eight primary dendrites, which ramified into third- to fifth-order dendrites. The longest observed dendrite in each neuron terminated at a distance of 600-900 microns from the soma. The sizes of soma and dendritic field of the two most rostrally located cells were smaller than those of the other two cells located more caudally. Dendritic spines were seen in all four cortically projecting basal forebrain neurons. Spines had shafts of variable lengths, and usually had spherical or elongated heads. The density of spines varied among the four neurons; one neuron, a type II cortically projecting basal forebrain neurons as defined physiologically by Reiner et al., had a much greater number of dendritic spines than the other three neurons, which were type I neurons. No somatic spines were observed. Presumptive axons were identified in three of the four cortically projecting basal forebrain neurons. These axons originated from either the soma or a primary dendrite, and two of them gave off local collaterals, which displayed occasional bouton-like swellings. The above observations confirm and extend previous findings that cortically projecting neurons in the basal forebrain are large multipolar cells, and provide evidence to support the conclusion that these cells, although somewhat variable in size, generally have extensive dendrites which display frequent spines.