Associative short-term synaptic plasticity mediated by endocannabinoids

Associative learning is important on rapid timescales, but no suitable form of short-term plasticity has been identified that is both associative and synapse specific. Here, we assess whether endocannabinoids can mediate such plasticity. In the cerebellum, bursts of parallel fiber (PF) activity evoke endocannabinoid release from Purkinje cell dendrites that results in retrograde synaptic inhibition lasting seconds. We find that the powerful climbing fiber (CF) to Purkinje cell synapse regulates this inhibition. Compared to PF stimulation alone, coactivation of PF and CF synapses greatly enhanced endocannabinoid-mediated inhibition of PF synapses. Retrograde inhibition was restricted to PFs activated within several hundred milliseconds of CF activation. This associative plasticity reflects two aspects of calcium-dependent endocannabinoid release. First, PF-mediated activation of metabotropic glutamate receptors locally reduced the dendritic calcium levels required for endocannabinoid release. Second, CF and PF coactivation evoked localized supralinear dendritic calcium signals. Thus, endocannabinoids mediate transient associative synaptic plasticity.     

Physiological properties of afferents and synaptic reorganization in the rat cerebellum degranulated by postnatal x-irradiation

Elimination of most granule, basket, and stellate interneurons in the rat cerebellum was achieved by repeated doses of low level x-irradiation applied during the first two weeks of postnatal life. Electrical stimulation of the brain stem and peripheral limbs was employed to investigate the properties of afferent cerebellar pathways and the nature of the reorganized neuronal synaptic circuitry in the degranulated cerebellum of the adult. Direct contacts of mossy fibers on Purkinje cells were indicated by short latency, single spike responses: 1.9 msec from the lateral reticular nucleus of brain stem and 5.4 msec from ipsilateral forlimb. These were shorter than in normal rats by 0.9 and 2.1 msec, respectively. The topography of projections from peripheral stimulation was approximately normal. Mossy fiber responses followed stimulation at up to 20/sec, whereas climbing fiber pathways fatigued at 10/sec. The latency of climbing fiber input to peripheral limb stimulation in x-irradiated cerebellum was 23 ± 8 (SD) msec. In x-irradiated rats, the climbing fiber pathways evoked highly variable extracellular burst responses and intracellular EPSPs of different, discrete sizes. These variable responses suggest that multiple climbing fibers contact single Purkinje cells. We conclude that each type of afferent retains identifying characteristics of transmission. However, rules for synaptic specification appear to break down so that: (1) abnormal classes of neurons develop synaptic connections, i.e., mossy fibers to Purkinje cells; (2) incorrect numbers of neurons share postsynaptic targets, i.e., more than one climbing fiber to a Purkinje cell; and (3) inhibitory synaptic actions may be carried out in the absence of the major inhibitory interneurons, i.e., Purkinje cell collaterals may be effective in lieu of basket and stellate cells.     

Climbing fiber activation of metabotropic glutamate receptors on cerebellar purkinje neurons

In the cerebellum, metabotropic glutamate receptors (mGluRs) are required for distinct forms of synaptic plasticity expressed at parallel fiber (PF) and climbing fiber (CF) synapses. At PF synapses, mGluR activation generates a slow synaptic current and triggers intracellular calcium release; at CF synapses, mGluR activation has not been observed. This has led some investigators to propose that mGluR-dependent changes in CF synaptic strength are induced heterosynaptically. Here we describe an mGluR-mediated response to CF stimulation consisting of two parallel signaling pathways: one leading to a slow synaptic conductance and the other leading to internal calcium release. This additional target for glutamate broadens the signaling capabilities of CF synapses and raises the possibility that changes in CF strength are homosynaptically triggered.     

Physiological properties of afferents and synaptic reorganization in the rat cerebellum degranulated by postnatal X-irradiation.

Elimination of most granule, basket, and stellate interneurons in the rat cerebellum was achieved by repeated doses of low level x-irradiation applied during the first two weeks of postnatal life. Electrical stimulation of the brain stem and peripheral limbs was employed to investigate the properties of afferent cerebellar pathways and the nature of the reorganized neuronal synaptic circuitry in the degranulated cerebellum of the adult. Direct contacts of mossy fibers on Purkinje cells were indicated by short latency, single spike responses: 1.9 msec from the lateral reticular nucleus of brain stem and 5.4 msec from ipsilpateral forelimb. These were shorter than in normal rats by 0.9 and 2.1 msec, respectively. The topography of projections from peripheral stimulation was approximately normal. Mossy fiber responses followed stimulation at up to 20/sec, whereas climbing fiber pathways fatigued at 10/sec. The latency of climbing fiber input to peripheral limb stimulation in x-irradiated cerebellum was 23 +/- 8 (SD) msec. In x-irradiated rats, the climbing fiber pathways evoked highly variable extracellular burst responses and intracellular EPSPs of different, discrete sizes. These variable responses suggest that multiple climbing fibers contact single Purkinje cells. We conclude that each type of afferent retains identifying characteristics of transmission. However, rules for synaptic specification appear to break down so that: (1) abnormal classes of neurons develop synaptic connections, i.e., mossy fibers to Purkinje cells; (2) incorrect numbers of neurons share postsynaptic targets, i.e., more than one climbing fiber to a Purkinje cell; and (3) inhibitory synaptic actions may be carried out in the absence of the major inhibitory interneurons, i.e., Purkinje cell collaterals may be effective in lieu of basket and stellate cells.     

Mechanosensory integration in the crayfish abdominal nervous system: Structural and physiological differences between interneurons with single and multiple spike initiating sites

Summary Cobalt backfills were used to demonstrate a population of approximately 50 paired interneurons in the 6th abdominal ganglion of the crayfish,Procambarus clarkii. Intracellular recordings from somata were used to study the response properties of individual interneurons, which were subsequently injected with Lucifer yellow. This report deals with 22 identified mechanosensory interneurons, which were each studied 2 to 20 times. (The total number of cells studied was 177). All but two of the interneurons could be assigned to one of two homogeneous classes, based on their receptive field sizes and four other consistent features: amplitude of soma spikes, duration of afterdischarge, presence of postsynaptic inhibition, and structure of the neuropilar processes. Unisegmental interneurons (Type I) (n=9) had excitatory receptive fields restricted to one segment, small soma spikes, little afterdischarge, and received extensive postsynaptic inhibition from contralateral and occasionally anterior sensory fields. All of these interneurons had a large diameter neuropilar segment (integrating segment) that was separated from the main axon by a constricted region. Multisegmental interneurons (Type II) (n=11) had excitatory receptive fields of at least six hemisegments (one half of the abdomen), large (sometimes overshooting) soma spikes, prolonged afterdischarge, and little evidence of postsynaptic inhibition. These interneurons lacked any expanded region of the dendritic tree that could be called an integrating segment. Anomalous interneurons (n=2) had multisegmental receptive fields, but in all other respects they resembled unisegmental interneurons, although their soma spikes were somewhat larger in amplitude.We hypothesize that the fundamental difference between the two main kinds of interneurons is that Type II interneurons have multiple spike initiating sites distributed throughout their dendritic trees, with any site being capable of initiating a spike that propagates to the main axon, while Type I interneurons have a single spike initiating site. The properties of anomalous interneurons are consistent with them having a single spike initiating site in each of several ganglia.     

Receptive-field modification in rat visual cortex induced by paired visual stimulation and single-cell spiking

Experience-dependent plasticity of visual cortical receptive fields (RFs) involves synaptic modifications in the underlying neural circuits, but the site and mechanism of these modifications remain to be elucidated. Using in vivo whole-cell recordings, we show that pairing visual stimulation at a given retinal location with spiking of a single neuron in developing rat visual cortex induces rapid RF modifications. The time course of the response to the visual stimulus at the paired RF location is altered, with an enhancement of the response preceding the spike time and a reduction following the spike. Such bidirectional modification is consistent with spike timing-dependent plasticity. Response modification also occurs at nearby locations, the direction and magnitude of which are correlated with the change at the paired location. In addition, changes at unpaired locations show a negative correlation with the initial strength of the response, which may facilitate rapid modification of the spatial RF profile.     

Cortical Presynaptic Control of Dorsal Horn C–Afferents in the Rat

Lamina 5 sensorimotor cortex pyramidal neurons project to the spinal cord, participating in the modulation of several modalities of information transmission. A well-studied mechanism by which the corticospinal projection modulates sensory information is primary afferent depolarization, which has been characterized in fast muscular and cutaneous, but not in slow-conducting nociceptive skin afferents. Here we investigated whether the inhibition of nociceptive sensory information, produced by activation of the sensorimotor cortex, involves a direct presynaptic modulation of C primary afferents. In anaesthetized male Wistar rats, we analyzed the effects of sensorimotor cortex activation on post tetanic potentiation (PTP) and the paired pulse ratio (PPR) of dorsal horn field potentials evoked by C–fiber stimulation in the sural (SU) and sciatic (SC) nerves. We also explored the time course of the excitability changes in nociceptive afferents produced by cortical stimulation. We observed that the development of PTP was completely blocked when C-fiber tetanic stimulation was paired with cortex stimulation. In addition, sensorimotor cortex activation by topical administration of bicuculline (BIC) produced a reduction in the amplitude of C–fiber responses, as well as an increase in the PPR. Furthermore, increases in the intraspinal excitability of slow-conducting fiber terminals, produced by sensorimotor cortex stimulation, were indicative of primary afferent depolarization. Topical administration of BIC in the spinal cord blocked the inhibition of C–fiber neuronal responses produced by cortical stimulation. Dorsal horn neurons responding to sensorimotor cortex stimulation also exhibited a peripheral receptive field and responded to stimulation of fast cutaneous myelinated fibers. Our results suggest that corticospinal inhibition of nociceptive responses is due in part to a modulation of the excitability of primary C–fibers by means of GABAergic inhibitory interneurons.     

Responses of cat auditory cortical neurons to tones of different frequencies and electrical stimulation of corresponding regions of the cochlea

Characteristic frequencies of neurons in the cat auditory cortex (area AI) whose receptive fields are located in different parts of the basilar membrane of the cochlea were determined in cats anesthetized with pentobarbital. The higher the characteristic frequency of a neuron in area AI, the nearer its receptive field lies to the base of the cochlea. Receptive fields of neurons with a characteristic frequency higher than 4 kHz lie on the first 10 mm of the basilar membrane. Receptive fields of neurons with a characteristic frequency below 4 kHz lie on the remaining 11–12 mm of the membrane. The effect of electrical stimulation of the center of the receptive field of a neuron corresponds to its response to a tone of characteristic frequency. The more the frequency of the acting tone differs from the characteristic frequency, or the further the point of stimulation from the center of the receptive field of the neuron, the less likely is the neuron to respond with an action potential. Neurons with a low characteristic frequency have wider receptive fields than neurons with a high characteristic frequency. Receptive fields of neurons with close characteristic frequencies on the basilar membrane overlap considerably. It was shown by the method of paired stimulation that excitation evoked in neurons in area AI by the action of a tone of a particular frequency is followed by long-lasting inhibition. This inhibition lasts longest and is most effective if a tone of the characteristic frequency is used.     

Sensory interneurons: some observations concerning the physiology and related structural significance of two cells in the crayfish brain

Two large interneurons in the crayfish brain which are sensitive to vibrational stimuli were injected with the fluorescent dye Procion Yellow. The dendritic branching profiles reflect the directional sensitivity of their respective mechanoreceptive fields on the cephalic appendages and integument. One interneuron branches exclusively on the contralateral side of the brain and receives monosynaptic input from the contralateral antenna; the second interneuron branches primarily on the ipsilateral side and is more sensitive to input from ipsilateral receptors although its receptive field is bilateral. The data suggest that these cells are primary and secondary sensory interneurons, respectively.     

Interactions between glutamate transporters and metabotropic glutamate receptors at excitatory synapses in the cerebellar cortex

Five glutamate transporter genes have been identified; two of these (EAAT3 and EAAT4) are expressed in neurons and are predominantly confined to the membranes of cell bodies and dendrites. At an ultrastructural level, glutamate transporters have been shown to surround excitatory synapses in hippocampus and cerebellum [J. Neurosci. 18 (1998) 3606; J. Comp. Neurol. 418 (2000) 255]. This pattern of localization overlaps the well-described perisynaptic distribution of Group I metabotropic glutamate receptors or mGluRs [Neuron 11 (1993) 771; J. Chem. Neuroanat. 13 (1997) 77]. Both of the principal excitatory synaptic inputs to cerebellar Purkinje neurons, the parallel fiber (PF) and climbing fiber (CF) synapses, express mGluR-dependent forms of synaptic plasticity [Nat. Neurosci. 4 (2001) 467]. Prompted by the colocalization of postsynaptic glutamate transporters and mGluRs, we have examined whether glutamate uptake limits mGluR-mediated signals and mGluR-dependent forms of plasticity at PF and CF synapses in cerebellar slices. We find that, at PF and, surprisingly also at CF synapses, mGluR activation generates a slow synaptic current and triggers intracellular calcium release. At both PF and CF synapses, mGluR responses are strongly limited by glutamate transporters under resting conditions and are facilitated by short trains of stimuli. Nearly every Purkinje neuron expresses an mGluR-mediated synaptic current upon inhibition of glutamate transport. Global applications of glutamate achieved by photolysis of chemically caged glutamate yield similar results and argue that the colocalized transporters can effectively limit glutamate access to the mGluRs even in the face of such a large amount of transmitter. We hypothesize that neuronal glutamate transporters and Group I mGluRs located in the perisynaptic space interact to sense and then regulate the amount of glutamate escaping excitatory synapses. This hypothesis is currently being tested using electrophysiological methods and the introduction of optically tagged glutamate transporter proteins. In the brain, synaptic signals are terminated mainly by neurotransmitter transporters. Families of genes encoding transporters for the major neurotransmitters (dopamine, GABA, glutamate, glycine, norepinephrine and 5-HT) have been identified. Although transporters serve as targets for important classes of therapeutic drugs (e.g. selective serotonin reuptake inhibitors) and drugs of abuse (amphetamine, cocaine), little is known about how they operate at a molecular level or contribute to synaptic transmission.     

The octopamine-containing buccal neurons are a new group of feeding interneurons in the pond snail Lymnaea stagnalis

In the pond snail, Lymnaea stagnalis, the paired buccal ganglia contain 3 octopamine-immunoreactive neurons, which have previously been shown to be part of the feeding network. All 3 OC cells are electrically coupled together and interact with all the known buccal feeding motoneurons, as well as with all the modulatory and central pattern generating interneurons in the buccal ganglia. N1 (protraction) phase neurons: Motoneurons firing in this phase of the feeding cycle receive either single excitatory (depolarising) synaptic inputs (B1, B6 neurons) or a biphasic response (hyperpolarisation followed by depolarisation) (B5, B7 motoneurons). Protraction phase feeding interneurons (SO, N1L, NIM) also receive this biphasic synaptic input after OC stimulation. All of protraction phase interneurons inhibit the OC neurons. N2 (retraction) phase neurons: These motoneurons (B2, B3, B9, B10) and N2 interneurons are hyperpolarised by OC stimulation. N2 interneurons have a variable (probably polysynaptic) effect on the activity of the OC neurons. N3 (swallowing) phase: OC neurons are strongly electrically coupled to both N3 phase (B4, B4cluster, B8) motoneurons and to the N3p interneurons. In case of the interneuronal connection (OC<->N3) the electrical synapse is supplemented by reciprocal chemical inhibition. However, the synaptic connections formed by the OC neurons or N3p interneurons to the other members of the feeding network are not identical. CGC: The cerebral, serotonergic CGC neurons excite the OC cells, but the OC neurons have no effect on the CGC activity. In addition to direct synaptic effects, the OC neurons also evoke long-lasting changes in the activity of feeding neurons. In a silent preparation, OC stimulation may start the feeding pattern, but when fictive feeding is already occurring, OC stimulation decreases the rate of the fictive feeding. Our results suggest that the octopaminergic OC neurons form a sub-population of N3 phase feeding interneurons, different from the previously identified N3p and N3t interneurons. The long-lasting effects of OC neurons suggest that they straddle the boundary between central pattern generator and modulatory neurons.     

Potentiation of mossy fiber EPSCs in the cerebellar nuclei by NMDA receptor activation followed by postinhibitory rebound current

Behavioral and computational studies predict that synaptic plasticity of excitatory mossy fiber inputs to cerebellar nuclear neurons is required for associative learning, but standard tetanization protocols fail to potentiate nuclear cell EPSCs in mouse cerebellar slices. Nuclear neurons fire action potentials spontaneously unless strongly inhibited by Purkinje neurons, raising the possibility that plasticity-triggering signals in these cells differ from those at classical Hebbian synapses. Based on predictions of neuronal activity during delay eyelid conditioning, we developed quasi-physiological induction protocols consisting of high-frequency mossy fiber stimulation and postsynaptic hyperpolarization. Robust, NMDA receptor-dependent potentiation of nuclear cell EPSCs occurred with protocols including a 150-250 ms hyperpolarization in which mossy fiber stimulation preceded a postinhibitory rebound depolarization. Mossy fiber stimulation potentiated EPSCs even when postsynaptic spiking was prevented by voltage-clamp, as long as rebound current was evoked. These data suggest that Purkinje cell inhibition guides the strengthening of excitatory synapses in the cerebellar nuclei.     

Endocannabinoids control the induction of cerebellar LTD

The long-term depression (LTD) of parallel fiber (PF) synapses onto Purkinje cells plays a central role in motor learning. Endocannabinoid release and LTD induction both depend upon activation of the metabotropic glutamate receptor mGluR1, require postsynaptic calcium increases, are synapse specific, and have a similar dependence on the associative activation of PF and climbing fiber synapses. These similarities suggest that endocannabinoid release could account for many features of cerebellar LTD. Here we show that LTD induction is blocked by a cannabinoid receptor (CB1R) antagonist, by inhibiting the synthesis of the endocannabinoid 2-arachidonyl glycerol (2-AG), and is absent in mice lacking the CB1R. Although CB1Rs are prominently expressed presynaptically at PF synapses, LTD is expressed postsynaptically. In contrast, a previously described transient form of inhibition mediated by endocannabinoids is expressed presynaptically. This indicates that Purkinje cells release 2-AG that activates CB1Rs to both transiently inhibit release and induce a postsynaptic form of LTD.     

Long-term synaptic changes induced in the cerebellar cortex by fear conditioning

To better understand learning mechanisms, one needs to study synaptic plasticity induced by behavioral training. Recently, it has been demonstrated that the cerebellum is involved in the consolidation of fear memory. Nevertheless, how the cerebellum contributes to emotional behavior is far from known. In cerebellar slices at 10 min and 24 hr following fear conditioning, we found a long-lasting potentiation of the synapse between parallel fibers and Purkinje cells in vermal lobules V-VI, but not in the climbing fiber synapses. The mechanism is postsynaptic, due to an increased AMPA response. In addition, in hotfoot mice with a primary deficiency of the parallel fiber to Purkinje cell synapse, cued (but not contextual) fear conditioning is affected. We propose that this synapse plays an important role in the learned fear and that its long-term potentiation may represent a contribution to the neural substrate of fear memory.     

GABAergic inhibition regulates developmental synapse elimination in the cerebellum

Functional neural circuit formation during development involves massive elimination of redundant synapses. In the cerebellum, one-to-one connection from excitatory climbing fiber (CF) to Purkinje cell (PC) is established by elimination of early-formed surplus CFs. This process depends on glutamatergic excitatory inputs, but contribution of GABAergic transmission remains unclear. Here, we demonstrate impaired CF synapse elimination in mouse models with diminished GABAergic transmission by mutation of a single allele for the GABA synthesizing enzyme GAD67, by conditional deletion of GAD67 from PCs and GABAergic interneurons or by pharmacological inhibition of cerebellar GAD activity. The impaired CF synapse elimination was rescued by enhancing GABA(A) receptor sensitivity in the cerebellum by locally applied diazepam. Our electrophysiological and Ca2+ imaging data suggest that GABA(A) receptor-mediated inhibition onto the PC soma from molecular layer interneurons influences CF-induced Ca2+ transients in the soma and regulates CF synapse elimination from postnatal day 10 (P10) to around P16.     

Floccular complex spike response to transparent retinal slip.

The flocculus of the rabbit is involved in the plasticity of compensatory eye movements. It is generally assumed that the climbing fiber input to floccular Purkinje cells encodes "retinal slip," which in turn would be a measure for the oculomotor performance error. To test this, we used transparent motion stimuli, creating a retinal slip signal that broke up this relation. We recorded the ensuing oculomotor behavior and complex spike activity of floccular Purkinje cells. Complex spike modulation in response to transparent stimulation was identical to that of a single optokinetic pattern, despite considerably different retinal slip. These results suggest that the climbing fiber code may be effectively related to the eye movement performance error, rather than to retinal slip.     

Identification of an inhibitory circuit that regulates cerebellar Golgi cell activity

Here we provide evidence that revises the inhibitory circuit diagram of the cerebellar cortex. It was previously thought that Golgi cells, interneurons that are the sole source of inhibition onto granule cells, were exclusively coupled via gap junctions. Moreover, Golgi cells were believed to receive GABAergic inhibition from molecular layer interneurons (MLIs). Here we challenge these views by optogenetically activating the cerebellar circuitry to determine the timing and pharmacology of inhibition onto Golgi cells and by performing paired recordings to directly assess synaptic connectivity. In contrast to current thought, we find that Golgi cells, not MLIs, make inhibitory GABAergic synapses onto other Golgi cells. As a result, MLI feedback does not regulate the Golgi cell network, and Golgi cells are inhibited approximately 2?ms before Purkinje cells, following a mossy fiber input. Hence, Golgi cells and Purkinje cells receive unique sources of inhibition and can differentially process shared granule cell inputs.     

Characteristics of the receptive fields of neurons of the posterior temporal area of the cortex of the cat

In records of 219 single units in the posterotemporal cortical area (field 21) of nonanaesthetized cats, 51% of cells reacted to visual stimulation. The neurones had receptive fields (RFs) with central (0-10 degrees) or peripheral (10-52 degrees) localization in the visual field, their size increasing with eccentricity. Carting of RFs by a light bar scanning the visual field revealed a considerable variability of RFs shape, size and orientation in different cells. RFs sizes of the majority of recorded cells (100-1000 grad) were very large and exceeded the size of large RFs of neurones in the primary projection zone of the visual cortex.     

Suppression of inhibitory synaptic potentiation by presynaptic activity through postsynaptic GABA(B) receptors in a Purkinje neuron

At inhibitory synapses on a cerebellar Purkinje neuron, the depolarization caused by heterosynaptic climbing fiber activation induces long-lasting potentiation accompanied by an increase in GABA(A) receptor responsiveness. Here we show that activation of a presynaptic inhibitory interneuron during the conditioning postsynaptic depolarization suppresses the potentiation. The suppression is due to postsynaptic GABA(B) receptor activation by GABA released from presynaptic terminals. The results suggest that GABA(B) receptor activation decreases the activity of cAMP-dependent protein kinase through the G(i)/G(o) proteins. The presynaptic activity-dependent suppression of synaptic plasticity is a novel regulatory mechanism of synaptic efficacy at individual synapses and may contribute to the learning and computational ability of the cerebellar cortex.     

Long-lasting increases in intrinsic excitability triggered by inhibition

Although experience-dependent changes in neural circuits are commonly assumed to be mediated by synaptic plasticity, modifications of intrinsic excitability may serve as a complementary mechanism. In whole-cell recordings from spontaneously firing vestibular nucleus neurons, brief periods of inhibitory synaptic stimulation or direct membrane hyperpolarization triggered long-lasting increases in spontaneous firing rates and firing responses to intracellular depolarization. These increases in excitability, termed firing rate potentiation, were induced by decreases in intracellular calcium and expressed as reductions in the sensitivity to the BK-type calcium-activated potassium channel blocker iberiotoxin. Firing rate potentiation is a novel form of cellular plasticity that could contribute to motor learning in the vestibulo-ocular reflex.