Short-term synaptic plasticity results from use-dependent activity, lasts on the timescale of milliseconds to minutes, and is thought to underlie working memory and neuronal information processing. Here, we focus on two forms of short-term plasticity: 1) post-tetanic potentiation (PTP), which is induced by high-frequency stimulation, and 2) presynaptic ionotropic receptor-activated synaptic enhancement, which can be produced by the activation of presynaptic glycine receptors. Potentiation of evoked and spontaneous responses is thought to arise from elevations in presynaptic residual Ca2+, which activates one or more molecular targets to increase neurotransmitter release. However, the Ca2+ sensor protein has not yet been identified. The overall goal of this work is to elucidate the Ca2+-dependent mechanisms of short-term plasticity.

Synapses vary widely in the probability of transmitter release. For instance, in response to an action potential the phasic synapses of the crayfish have a 100-1000-fold higher release probability than tonic synapses. The difference in release probability is attributed to differences in the exocytotic machinery such as the degree of “zippering” of the trans-SNARE (Soluble N-ethylmaleimide-sensitive factor Attachment protein REceptor) complex. I used physiological and molecular approaches to determine if the zippered state of SNAREs associated with synaptic vesicles and the interaction between the SNARE complex and Complexin influence the probability of release at the synapse. I used three Botulinum neurotoxins which bind and cleave at different sites on VAMP to determine whether these sites were occluded by SNARE interaction (zippering) or open to proteolytic attack. Under low stimulation conditions, the light-chain fragment of botulinum B (BoNT/B-LC) but not BoNT/D-LC or tetanus neurotoxin (TeNT-LC) cleaved VAMP and inhibited evoked release at both phasic and tonic synapses. In addition, a peptide based on the C-terminal half of crayfish VAMP’s SNARE motif (Vc peptide) designed to interfere with SNARE complex zippering at the C-terminal end inhibited release at both synapses. The susceptibility of VAMP to only BoNT/B-LC and interference by the Vc peptide indicated that SNARE complexes at both phasic and tonic synapses were partially zippered only at the N-terminal end with the C-terminal end exposed under resting conditions. I used a peptide containing part of the crayfish Complexin central α-helix domain to interfere with the interaction between Complexin and the SNARE complex. The peptide enhanced phasic evoked release and inhibited tonic evoked release under low stimulation but attenuated release at both synapses under intense stimulation. Therefore, Complexin appeared to exhibit a dual function under low synaptic activity but only promoted release under high synaptic activity. The results showed that the zippered state of the SNARE complex does not determine initial release probability as a similar zippered SNARE complex structure under resting conditions is common to both phasic and tonic synapses. However, Complexin may have a role in influencing the initial release probability of a synapse. Therefore, the interaction between the SNARE complex and Complexin is important for release but other factors contribute more significantly to synaptic strength.

Synapses vary widely in the probability of transmitter release. For instance, in response to an action potential the phasic synapses of the crayfish have a 100-1000-fold higher release probability than tonic synapses. The difference in release probability is attributed to differences in the exocytotic machinery such as the degree of “zippering” of the trans-SNARE (Soluble N-ethylmaleimide-sensitive factor Attachment protein REceptor) complex. I used physiological and molecular approaches to determine if the zippered state of SNAREs associated with synaptic vesicles and the interaction between the SNARE complex and Complexin influence the probability of release at the synapse. I used three Botulinum neurotoxins which bind and cleave at different sites on VAMP to determine whether these sites were occluded by SNARE interaction (zippering) or open to proteolytic attack. Under low stimulation conditions, the light-chain fragment of botulinum B (BoNT/B-LC) but not BoNT/D-LC or tetanus neurotoxin (TeNT-LC) cleaved VAMP and inhibited evoked release at both phasic and tonic synapses. In addition, a peptide based on the C-terminal half of crayfish VAMP’s SNARE motif (Vc peptide) designed to interfere with SNARE complex zippering at the C-terminal end inhibited release at both synapses. The susceptibility of VAMP to only BoNT/B-LC and interference by the Vc peptide indicated that SNARE complexes at both phasic and tonic synapses were partially zippered only at the N-terminal end with the C-terminal end exposed under resting conditions. I used a peptide containing part of the crayfish Complexin central α-helix domain to interfere with the interaction between Complexin and the SNARE complex. The peptide enhanced phasic evoked release and inhibited tonic evoked release under low stimulation but attenuated release at both synapses under intense stimulation. Therefore, Complexin appeared to exhibit a dual function under low synaptic activity but only promoted release under high synaptic activity. The results showed that the zippered state of the SNARE complex does not determine initial release probability as a similar zippered SNARE complex structure under resting conditions is common to both phasic and tonic synapses. However, Complexin may have a role in influencing the initial release probability of a synapse. Therefore, the interaction between the SNARE complex and Complexin is important for release but other factors contribute more significantly to synaptic strength.

Neurotransmitter release at presynaptic terminals is a complex process involving calcium ion influx through voltage-gated calcium channels (CaV). In addition to their role as entry points through which calcium influx may occur, CaV are now understood to be fundamental components of a common release-site complex that is highly adapted for modulation. Consistent with this model, I investigated mechanisms of modulating a presynaptic calcium channel, CaV2.2, via a heterotrimeric G-protein pathway. Using the patch-clamp technique, I demonstrated in chick dorsal root ganglion (DRG) neurons that the slow kinetics of G-protein inhibition of CaV2.2 via GTPgammaS were limited by the rate of GDP dissociation from the G-protein nucleotide binding site. In addition, I investigated the role of G-protein regulation of CaV2.2 currents evoked by action potential-like stimuli. Here, I characterized an inhibited current that was advanced in time with respect to uninhibited controls. These currents exhibited a shorter latency to current activation and faster deactivation. These findings may have important physiological ramifications on signal transduction and timing. In addition to G-protein regulation, presynaptic CaV2.2 have been demonstrated to exhibit a resistance to voltage-dependent inactivation (VDI), a property thought to be important in determining channel availability and synaptic excitability. I demonstrated a role for dynamic palmitoylation in conferring resistance to VDI in presynaptic terminals of the chick ciliary ganglion. Using tunicamycin, an inhibitor of palmitoylation, I induced a hyperpolarizing shift in the steady-state-inactivation (SSI) profile of presynaptic CaV2.2. Finally, I examined the role of a CaV interacting protein, Munc18, as a potential regulator of CaV. I probed for alterations in CaV2.2 function in DRG neurons that had been transfected with Munc18 or Munc18 siRNA. Despite the intimate interaction between Munc18 and CaV2.2, no major effects on the fundamental characteristics of CaV2.2 function were observed. However, a hyperpolarizing shift in the inactivation profile of CaV2.2 was determined in DRG neurons in which Munc18 was knocked down. It is not clear if this was a direct consequence of Munc18 perturbation.

Neurotransmitter release at presynaptic terminals is a complex process involving calcium ion influx through voltage-gated calcium channels (CaV). In addition to their role as entry points through which calcium influx may occur, CaV are now understood to be fundamental components of a common release-site complex that is highly adapted for modulation. Consistent with this model, I investigated mechanisms of modulating a presynaptic calcium channel, CaV2.2, via a heterotrimeric G-protein pathway. Using the patch-clamp technique, I demonstrated in chick dorsal root ganglion (DRG) neurons that the slow kinetics of G-protein inhibition of CaV2.2 via GTPgammaS were limited by the rate of GDP dissociation from the G-protein nucleotide binding site. In addition, I investigated the role of G-protein regulation of CaV2.2 currents evoked by action potential-like stimuli. Here, I characterized an inhibited current that was advanced in time with respect to uninhibited controls. These currents exhibited a shorter latency to current activation and faster deactivation. These findings may have important physiological ramifications on signal transduction and timing. In addition to G-protein regulation, presynaptic CaV2.2 have been demonstrated to exhibit a resistance to voltage-dependent inactivation (VDI), a property thought to be important in determining channel availability and synaptic excitability. I demonstrated a role for dynamic palmitoylation in conferring resistance to VDI in presynaptic terminals of the chick ciliary ganglion. Using tunicamycin, an inhibitor of palmitoylation, I induced a hyperpolarizing shift in the steady-state-inactivation (SSI) profile of presynaptic CaV2.2. Finally, I examined the role of a CaV interacting protein, Munc18, as a potential regulator of CaV. I probed for alterations in CaV2.2 function in DRG neurons that had been transfected with Munc18 or Munc18 siRNA. Despite the intimate interaction between Munc18 and CaV2.2, no major effects on the fundamental characteristics of CaV2.2 function were observed. However, a hyperpolarizing shift in the inactivation profile of CaV2.2 was determined in DRG neurons in which Munc18 was knocked down. It is not clear if this was a direct consequence of Munc18 perturbation.

The first part of my thesis involves an investigation into mechanisms underlying the presynaptic regulation of transmitter release from myelinated hindlimb sensory afferents in rodents. The central hypothesis is that in addition to chemical transmission in spinal neuronal networks, electrical synapses formed by connexins are critically involved in presynaptic inhibition of large diameter sensory afferents. Subsequent sections of the thesis present a detailed examination of the distribution of connexins in the rodent spinal cord with a particular emphasis on the neuronal connexin, Cx36. Connexin36 (Cx36) is widely believed to be the protein forming the neuronal gap junctions that create electrical synapses between mammalian neurons in many areas of the central nervous system (Condorelli et al 1998). The first part of thesis concerns a previously unknown role of neuronal connexins in interneurone pathways involved in presynaptic control of synaptic transmission in the lumbar spinal cord of rodents. As far as we are aware, the idea that electrical contacts between spinal neurons contribute to spinal presynaptic inhibition is a novel hypothesis. Evidence will be presented: 1) that Cx36 is present in regions of the spinal cord containing interneurons involved in presynaptic inhibition, 2) that the lack of Cx36 in Cx36-/- knockouts mice results in a severe impairment of presynaptic inhibition, and 3) that blocking gap junctions pharmacologically in wild type mice impairs presynaptic inhibition. The exploration of this hypothesis will involve a combination of electrophysiological and immunohistochemical approaches in juvenile wild-type and knockout mice lacking Cx36, as well as immunohistochemical observations in adult rodents. This first section of the thesis begins with the development of a preparation in which several measures of presynaptic inhibition described in the in vivo adult cat preparation can be examined in vitro in young mice. The following sections of the thesis describe the distribution and features of Cx36 on neurons in mice and rats of different ages in four parts. The first will show that Cx36 is the only connexin associated with spinal neurons and refutes claims in the literature about the existence of a variety of connexions on spinal neurons. The second part will show that while gap junctions between some spinal neurons are only a transient developmental phenomenon, they persist in abundance in adult animals. The third part will present evidence of a previously unsuspected III association of Cx36 gap junctions at the chemical synapse between muscle afferent fibres and motoneurons. Specifically, an association between Cx36 and the glutamate transporter used in primary afferents, Vglut1 will be described. To our knowledge these results are the first to suggest the existence of mixed (electrical and chemical) synapses between primary afferents and motoneurons in the mature mammalian spinal cord. The final part of the thesis will describe the presence of Cx36 gap junctions on adult sacral motoneurons involved in control of sexual, urinary and defecation functions in the rodent.

The first part of my thesis involves an investigation into mechanisms underlying the presynaptic regulation of transmitter release from myelinated hindlimb sensory afferents in rodents. The central hypothesis is that in addition to chemical transmission in spinal neuronal networks, electrical synapses formed by connexins are critically involved in presynaptic inhibition of large diameter sensory afferents. Subsequent sections of the thesis present a detailed examination of the distribution of connexins in the rodent spinal cord with a particular emphasis on the neuronal connexin, Cx36. Connexin36 (Cx36) is widely believed to be the protein forming the neuronal gap junctions that create electrical synapses between mammalian neurons in many areas of the central nervous system (Condorelli et al 1998). The first part of thesis concerns a previously unknown role of neuronal connexins in interneurone pathways involved in presynaptic control of synaptic transmission in the lumbar spinal cord of rodents. As far as we are aware, the idea that electrical contacts between spinal neurones contribute to spinal presynaptic inhibition is a novel hypothesis. Evidence will be presented: 1) that Cx36 is present in regions of the spinal cord containing interneurones involved in presynaptic inhibition, 2) that the lack of Cx36 in Cx36-/- knockouts mice results in a severe impairment of presynaptic inhibition, and 3) that blocking gap junctions pharmacologically in wild type mice impairs presynaptic inhibition. The exploration of this hypothesis will involve a combination of electrophysiological and immunohistochemical approaches in juvenile wild-type and knockout mice lacking Cx36, as well as immunohistochemical observations in adult rodents. This first section of the thesis begins with the development of a preparation in which several measures of presynaptic inhibition described in the in vivo adult cat preparation can be examined in vitro in young mice. The following sections of the thesis describe the distribution and features of Cx36 on neurones in mice and rats of different ages in four parts. The first will show that Cx36 is the only connexin associated with spinal neurons and refutes claims in the literature about the existence of a variety of connexions on spinal neurons. The second part will show that while gap junctions between some spinal neurons are only a transient developmental phenomenon, they persist in abundance in adult animals. The third part will present evidence of a previously unsuspected III association of Cx36 gap junctions at the chemical synapse between muscle afferent fibres and motoneurones. Specifically, an association between Cx36 and the glutamate transporter used in primary afferents, Vglut1 will be described. To our knowledge these results are the first to suggest the existence of mixed (electrical and chemical) synapses between primary afferents and motoneurones in the mature mammalian spinal cord. The final part of the thesis will describe the presence of Cx36 gap junctions on adult sacral motoneurones involved in control of sexual, urinary and defecation functions in the rodent.

Diss. (sammanfattning) Stockholm : Karol. inst., 2003. / Härtill 4 uppsatser.

xv, 102 p. : ill. (some col.) / This dissertation describes research to elucidate the early steps in the process of synapse formation in the zebrafish spinal cord. One question is how presynaptic proteins are trafficked and recruited to nascent synapses. Previous work has suggested two possible models of presynaptic transport, either (1) most presynaptic proteins are transported together or (2) two types of transport packets, synaptic vesicle (SV) protein transport vesicles (STVs) and Piccolo-containing active zone precursor transport vesicles (PTVs), transport the necessary components separately. We tested these models using in vivo imaging in zebrafish spinal cord and found that the recruitment of at least three distinct transport packets during presynaptic assembly of a glutamatergic synapse occurs in an ordered sequence. First, STVs are stabilized at future synaptic sites, then PTVs, followed by a third transport packet type carrying Synapsin, a cytosolic protein that can tether SVs to actin. These results identify an order to the assembly of the presynaptic terminal in vivo, suggesting that a single synaptogenic interaction may precipitate the cascade of recruitment steps. We next examined the Cadm/SynCAM family of cell adhesion molecules, a family of proteins that has been shown to be able to induce synapse formation in vitro and was thought to play a role in recruitment of presynaptic proteins. As the role of these proteins in vivo was not well understood, we chose to examine the role of the cadms in zebrafish spinal cord. We found that zebrafish possess six cadm genes, and all are expressed throughout the nervous system both during development and in the adult. We then looked at the role of one of the Cadms, Cadm2a, in vivo in the zebrafish spinal cord. We found that knockdown of cadm2a significantly decreases the ability of zebrafish embryos to respond to touch. We also found that there is a significant reduction in the number of synapses, as shown by immunohistochemistry, formed between Rohon-Beard and CoPA neurons, the first two cell types in the touch response circuit. These data suggest that Cadm2a plays an important role in synapse formation in vivo. This dissertation contains both my previously published and unpublished co-authored material. / Committee in charge: Monte Westerfield Chairperson; Philip Washbourne, Advisor; Judith Eisen, Member; Tory Herman, Member; Mike Wehr, Outside Member

Antecedendo movimentos voluntariamente gerados, existe atividade neuronal encefálica que se inicia alguns segundos antes da execução deste movimento. Esta atividade preparatória é responsável pela elaboração de um plano de execução que alcança a via final comum para realização de um ato motor voluntário, os motoneurônios. Entretanto, na última década, evidências apontam para a participação de circuitos neuronais na medula espinhal apresentando padrão de atividade similar aos padrões observados em áreas encefálicas e que, possivelmente, estaria relacionado a uma atividade preparatória para o movimento voluntariamente gerado. Por este motivo, o presente trabalho teve por objetivo verificar a atividade de circuitos neuronais na medula espinhal durante diferentes instantes de proximidade da ação voluntariamente gerada em paradigma de tarefa motora com período de instrução. Para isso, inicialmente, 15 sujeitos saudáveis, sem histórico de doença neuromuscular foram submetidos ao protocolo experimental. O protocolo experimental constituiu-se do processo de recrutamento dos sujeitos, sua preparação para o ensaio dentro do ambiente experimental, bem como as orientações necessárias para execução dos procedimentos e paradigmas. Os procedimentos referem-se às etapas realizadas para captação do reflexo H, bem como desta captação sob a influência de técnica de condicionamento por inibição pré-sináptica. Essa captação ocorreu em janelas de aquisição em que o sujeito encontrava-se em repouso e em três instantes de expectativa para a execução de ação voluntária, estando o músculo sóleo atuando como agonista (flexão plantar) ou antagonista (dorsiflexão), em paradigma de tarefa motora voluntária com período de instrução. Após os registros, por meio de processamento dos sinais coletados, foi possível se calcular a amplitude pico-a-pico do reflexo H nas diferentes condições experimentais de proximidade da execução (1000, 600 e 200 milissegundos) e de atuação do músculo sóleo (agonista e antagonista) que foi usado para: (1) análise da variação da excitabilidade reflexa, em porcentagem da onda M máxima, (2) análise da ocorrência de inibição pré-sináptica e (3) análise da variação da inibição pré-sináptica, em porcentagem de inibição. Os resultados mostram que a porcentagem da onda M máxima aumentou significativamente nos três instantes de proximidade com os sujeitos estando em expectativa da execução da tarefa motora quando o músculo sóleo atuaria como agonista da contração, quando comparados com os registros obtidos nas mesmas condições em repouso. Contudo, somente a 200 ms da execução é que foi observado aumento da porcentagem da onda M máxima quando o músculo sóleo atuaria como antagonista. Inibição pré-sináptica ocorreu em todas as condições experimentais, contudo aumento significativo da porcentagem de inibição pré-sináptica foi somente observado a 200 ms da execução da tarefa motora em que o músculo sóleo atuaria como antagonista. Diferenças entre agonista e antagonista com relação ao padrão de excitabilidade reflexa foi somente observado a 600 ms de proximidade da execução da tarefa e essas diferenças com relação à porcentagem de inibição pré-sináptica foi somente detectada a 200 ms. Nossos resultados nos permitem concluir que circuitos neuronais na medula espinhal apresentam atividade no período preparatório para a execução de tarefa motora voluntária que podem estar relacionadas ao comportamento de expectativa da realização de uma ação motora eminente, bem como relacionada ao planejamento motor para a ação a longa proximidade da execução de movimentos. / There is brain activity preceding voluntary movements a few seconds before the execution of the movement. This preparatory activity is responsible for the execution plan that reaches the final common pathway, i.e., the motoneurons. In the last decade, there have been reports indicating the involvement of spinal cord circuits in the preparatory activity for movement. The present work has the objective of verifying the activity of spinal cord neuronal circuits at different times preceding a voluntary action, under an instructed delay period paradigm. Fifteen healthy subjects participated in the study. The protocol included an explanation of the experimental tasks. Electrophysiological recordings of the H reflex with and without presynaptic inhibition conditioning were employed. The epochs of H reflex recording were associated either with a resting period or with one of three pre-action periods. The subject received a cue at an appropriate time about the type of contraction: plantarflexion or dorsiflexion. Peak to peak H reflex values were computed in the control resting period and at 1000 ms, 600 ms and 200 ms before the action. Percent values of H amplitude with respect to maximum M values were computed as well as the level of presynaptic inhibition. The results have shown that the relative H reflex value increased significantly at the three premovement times for the soleus under an agonist contraction (i.e., plantarflexion) when compared to control. However, when the soleus was an antagonist to the contraction (i.e., dorsiflexion) there was a statistical difference in the H amplitude only at 200 ms before movement. Presynaptic inhibition occurred in all experimental conditions, however only at 200 ms before contraction there was a significant increase. Differences in reflex excitability between agonist and antagonist activity were only observed at 600 ms before action. On the other hand, differences in presynaptic inhibition were only found at 200 ms before contraction. The results indicated that spinal cord neuronal circuits are activated during the preparatory period preceding a voluntary action. These may be correlated with an expectancy behavior for the execution of an imminent motor action and also with the planning of a motor action at larger times preceding movement execution.