- Email: [email protected]
Meaningless minis? Mechanisms of neurotransmitter-receptor clustering
Research Update
9 Tokumaru, H. et al. (2001) SNARE complex oligomerization by synaphin/complexin is essential for synaptic vesicle exocytosis. Cell 104, 421–432 10 Itakura, M. et al. (1999) Transfection analysis of functional roles of complexin I and II in the exocytosis of two different types of secretory vesicles. Biochem. Biophys. Res. Commun. 265, 691–696 11 Chen, X. et al. (2002) Three-dimensional structure of the complexin/SNARE complex. Neuron 33, 397–409
TRENDS in Neurosciences Vol.25 No.8 August 2002
12 Pabst, S. et al. (2000) Selective interaction of complexin with the neuronal SNARE complex. J. Biol. Chem. 275, 19808–19818 13 Pabst, S. et al. (2002) Rapid and selective binding to the synaptic SNARE complex suggests a modulatory role of complexins in neuroexocytosis. J. Biol. Chem. 277, 7838–7848 14 Xu, T. et al. (1999) Inhibition of SNARE complex assembly differentially affects kinetic components of exocytosis. Cell 99, 713–722
383
Karla E. Marz Phyllis I. Hanson* Dept of Cell Biology and Physiology, Washington University School of Medicine, 660 South Euclid Avenue, Box 8228, St Louis, MO 63110, USA. *e-mail: [email protected]
Meaningless minis? Mechanisms of neurotransmitterreceptor clustering Patrik Verstreken and Hugo J. Bellen Initiation and maintenance of the postsynaptic neurotransmitter-receptor field are important steps during synapse formation and maturation, as they play a determinative role in regulating synaptic strength. However, the mechanisms directing neurotransmitter-receptor clustering and maintenance are poorly understood. Recently, two models explaining glutamate-receptor clustering at the Drosophila neuromuscular junction have been proposed. One model postulates that release of an agent via single vesicle fusion events (minis) is required for the initiation of postsynaptic glutamatereceptor clustering, and that glutamate is not responsible for initiation or maintenance of the postsynaptic receptor field. The other model rules out a role for minis in initiation of clustering, and suggests a role for non-vesicular release of glutamate in receptor-field maintenance. Here, we compare and discuss the data underlying both models.
For neurons to communicate effectively with one another it is crucial that the presynaptic density, the site at which neurotransmitters are released, is aligned with the postsynaptic receptor field. In addition, the abundance of the receptors in the postsynaptic field is known to limit the postsynaptic response. Hence, alignment and size of the postsynaptic receptor field, as well as those of the presynaptic density, are important determinants in regulating synaptic strength [1]. Numerous vertebrate neurons communicate with each other using glutamate as an excitatory neurotransmitter. Glutamate is also utilized as a neurotransmitter at the http://tins.trends.com
fruit-fly neuromuscular junction (NMJ). The postsynaptic clustering of receptors is believed to be mediated by the presynaptic neuron, because glutamate-receptor clustering at the postsynaptic site depends on presynaptic activity [2–4]. The presynaptic signal seems to induce a postsynaptic cascade of events involving many proteins, including kinases and vesicle trafficking proteins in vertebrate cells, that eventually leads to the regulation of glutamate-receptor insertion into the membrane and clustering at the synapse [5]. The nature of the presynaptic signal resulting in glutamate-receptor clustering is currently unknown in Drosophila, but a candidate, neuronal activity regulated pentraxin (NARP), has been identified in the mouse [6]. As in mammalian neurons, the trigger to cluster glutamate receptors at the NMJ of Drosophila embryos is believed to originate presynaptically (because the clustering is an activity-dependent process, requiring a contact between the motor nerve ending and the muscle). Glutamate receptors are known to be present in muscles before synapse formation, but do not cluster at the postsynaptic density before nerve contact or in the absence of electrical activity [4,7,8]. Hence, nerve–muscle contact and electrical activity are essential for initial clustering. However, many crucial questions remain unanswered, or are at least the subject of intense debate. Which substances induce and maintain clustering? Are different substances involved in induction and maintenance? Are minis required for clustering? Do different types of vesicles control the processes of membrane addition to the
synapse, neurotransmitter release and release of clustering agents? Do these forms of vesicular release all depend on the same protein machinery? Are there non-vesicular forms of release? In the next sections we compare data presented in two studies, one by Saitoe et al. [9] and one by Featherstone et al. [10]. We outline the proposed models that try to answer some of these questions. Initiation of receptor clustering
The availability of Drosophila mutants with blocked neurotransmitter release has allowed investigation of the role of synaptic transmitter release in receptor clustering. Some mutations that block evoked release [e.g. synaptobrevin, cysteine string protein (CSP)] do not block the spontaneous fusion events that are measured as minis, whereas other mutations block both evoked and spontaneous release (e.g. shibire or syntaxin) [11,12,13]. Saitoe et al. [9] showed that glutamate receptors clustered normally when only evoked activity was blocked in synaptobrevin or CSP mutants. However, when both evoked and spontaneous neurotransmitter release were blocked, using syntaxin or shibire mutants, the authors reported no clustering of postsynaptic glutamate receptors. These data indicate that spontaneous neurotransmitter release, usually considered merely the leakage of transmitter from single vesicle fusions, plays a paramount role in inducing postsynaptic receptor clustering [9,11] (Fig. 1a). Although conceptually appealing (minis, thus, having a meaning), these data were recently challenged by Broadie and colleagues (University of Utah,
0166-2236/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0166-2236(02)02197-5
Research Update
384
(a)
TRENDS in Neurosciences Vol.25 No.8 August 2002
(b)
Initiation of receptor clustering; Saitoe et al. (2001) Synaptic vesicle Glutamate receptor subunit Endosome
Regulation of receptor cluster size; Featherstone et al. (2002) Excitatory amino acid transporter Pre-synaptic glutamate
2 1 1 2
Initiation: Minis are required, evoked release is not required Regulation of cluster size: Glutamate most likely has no role (at least) it does not act through glutamate receptors
Initiation: Neither minis nor evoked release is required Regulation of cluster size: Non vesicular glutamate release is a negative regulator of receptor cluster size TRENDS in Neurosciences
Fig. 1. Boutons at the neuromuscular junction of Drosophila. (a) Saitoe et al. have proposed a model in which spontaneous vesicle fusion (1) induces postsynaptic receptor clustering (2) [9]. According to the model, the fusion of a single vesicle not only results in liberation of neurotransmitter into the synaptic cleft, but possibly also leads to release of a factor inducing clustering of glutamate receptors. This factor has yet to be identified. Maintenance of receptor cluster size is, according to Saitoe et al., not dependent on glutamate in the cleft. (b) In contrast to Saitoe et al., Featherstone et al. have proposed a model in which glutamate-receptor cluster size is controlled by the abundance of presynaptic cytosolic glutamate [10]. In their model, Featherstone et al. propose that glutamate from a non-vesicular source is a signaling molecule that negatively regulates receptor clustering. (1) Glutamate could be released into the synaptic cleft by reversal of the polarity of excitatory amino-acid transporters, but to date no in vivo evidence supports this transport mechanism. (2) The presynaptic signal could be a regulator of the glutamatereceptor cycle. Glutamate receptors are continuously internalized and transported to the endosome. They are re-inserted into the plasma membrane (arrows) and they cluster at the active zone [18]. Regulation of cycling rates and regulation of receptor breakdown by trafficking to the lysosome (not shown) would allow for tight control of receptor cluster size at the postsynaptic density.
UT, USA) [10]. Featherstone et al. showed that glutamate-receptor clustering did occur in the absence of evoked and spontaneous release, indicating that minis are not required [10] (Fig. 1b). Like Saitoe et al., they used syntaxin and shibire ts1 mutants to block endogenous, as well as spontaneous, vesicle fusion [12,13], and both teams used similar techniques to assay glutamate-receptor clustering. Technical issues that could explain some of the differences between the studies include the presence of maternal component in syntaxin mutants, and cell death in shibire ts1 mutants that are kept at the restrictive temperature for protracted periods. In addition, the functional consequences of removal of syntaxin or shibire (demonstrating the absence of release) ought to be measured in the same muscles as those used for assaying glutamate-receptor clustering – yet, this was not done by Saitoe et al. [9]. In summary, further studies will have to be carried out to resolve these discrepancies, but the observation by Featherstone et al. that clustering does http://tins.trends.com
occur in the shibirets1 and syntaxin mutants must be viewed as strong positive evidence for clustering in the absence of spontaneous release [10]. Maintenance of receptor clustering
Another issue relates to the maintenance of receptor clustering once initial clustering has occurred. Maintenance of postsynaptic receptor clusters is, according to Saitoe et al., not dependent on glutamate in the synaptic cleft [9]. They injected glutamate together with argiotoxin, an activity-dependent glutamate channel blocker, into developing wild-type embryos, and determined if glutamate-receptor clustering was affected at the end of embryonic development. If glutamate or the glutamate receptors have an effect on receptor cluster size, then blocking the glutamate receptor could affect cluster formation or maintenance. The authors argue that they observed no difference with control animals, suggesting that neither glutamate nor its receptor affect cluster size. Hence, they propose that an
instructive signal for receptor clustering and maintenance, other than glutamate, is released by spontaneous fusion of single vesicles [9] (Fig. 1a). Featherstone et al. come to a different conclusion: they propose that glutamate released into the cleft by a non-vesicular mechanism does affect maintenance of the cluster size [10]. In previous work, Featherstone et al. identified mutations in the gene encoding Drosophila glutamicacid decarboxylase (GAD) [14]. This enzyme is present at the Drosophila NMJ and converts glutamate into GABA. Although GABA is not a neurotransmitter at the Drosophila NMJ, GAD was shown to play an important role in postsynaptic glutamate-receptor field regulation. This finding led to the idea that the levels of presynaptic glutamate might control the level of postsynaptic receptor cluster size. Indeed, when the presynaptic terminal contains more glutamate (e.g. in mutants with blocked glutamate breakdown, like GAD), fewer glutamate receptors cluster postsynaptically at the NMJ. Conversely, when the motoneuron endings contain less glutamate (achieved by accelerating glutamate breakdown), significantly more glutamate receptors cluster at the NMJ [10]. Featherstone et al. found that presynaptic glutamate controls the postsynaptic transmitter-receptor field size independently of synaptic vesicle mediated neurotransmitter release (i.e. independently of both evoked and spontaneous synaptic vesicle fusion). This led the authors to propose that presynaptic glutamate is released into the synaptic cleft, by a non-vesicular mechanism, to act as a negative regulator of postsynaptic glutamate-receptor cluster size [10]. One possibility is that excitatory amino-acid transporters (EAATs) release small amounts of glutamate into the synaptic cleft by reversing their polarity. Two EAATs are present in the Drosophila nervous system [15]; however, no in vivo data are available to suggest the existence of non-vesicular glutamate release at the Drosophila NMJ (Fig. 1b). Again the two sets of data contradict each other. The work by Saitoe et al. implies that glutamate plays no part in regulation of receptor cluster size [9], whereas Featherstone et al. propose a model in which glutamate itself is the effector of field-size regulation in the absence of vesicular release [10]. However, both claims are on loose ground,
Research Update
not supported by solid data. Saitoe et al. showed that extracellular glutamate does not influence receptor cluster size when the glutamate receptor is blocked using argiotoxin, an open channel blocker. However, the inhibition of glutamate channel activity by argiotoxin is not absolute [16] and furthermore, the authors also co-injected the argiotoxin with glutamate to open and block the glutamate receptors. This could well alleviate the effect of the toxin – addition of glutamate could rescue the partial block in glutamate-receptor activity by the toxin. Hence, a dramatic effect on glutamatereceptor clustering is not expected per se. By contrast, Featherstone et al. provide no direct evidence that glutamate secreted by the presynaptic cell is acting to regulate receptor downregulation on the postsynaptic side [10]. Other mechanisms, possibly regulated by presynaptic glutamate, could also maintain and regulate postsynaptic glutamate-receptor clusters. Notably, in the mouse a glutamate-receptor clustering agent, NARP, was identified, and there is a Drosophila homologue of NARP (GenBank accession number Y17570; Flybase: CG3100) [11]. This homologue could also be involved in clustering of glutamate receptors at the fly NMJ. In addition, recent advances in the quest to understand the development of the cytomatrix at the active zone have led to identification of a population of 80 nm diameter dense core vesicles. These vesicles carry active-zone components, but not synaptic-vesicle-specific proteins [17]. The emerging model predicts that these vesicles fuse with the presynaptic membrane on stimulation and deposit fresh active zone material (e.g. syntaxin,
TRENDS in Neurosciences Vol.25 No.8 August 2002
SNAP-25). Hence, it is possible that fusion events of a different nature to minis are required for pre- and postsynaptic density development. However, as these vesicles themselves transport part of the synaptic vesicle fusion machinery to the active zone, it is currently not known if fusion of these specialized vesicles requires the classical synaptic vesicle fusion machinery. In conclusion, the studies by Saitoe et al. [9] and Featherstone et al. [10] aim to identify the mechanisms leading to the initiation of receptor clustering and regulation of cluster size. The studies present contradicting data and propose conflicting models. Future work, both in vertebrates and in Drosophila, will be needed to shed light on the mechanisms of initiation and maintenance of postsynaptic glutamate-receptor clustering. References 1 Turrigiano, G.G. (2000) AMPA receptors unbound: membrane cycling and synaptic plasticity. Neuron 26, 5–8 2 Lissin, D.V. et al. (1998) Activity differentially regulates the surface expression of synaptic AMPA and NMDA glutamate receptors. Proc. Natl. Acad. Sci. U. S. A. 95, 7097–7102 3 O Brien, R.J. et al. (1998) Molecular mechanisms of glutamate receptor clustering at excitatory synapses. Curr. Opin. Neurobiol. 8, 364–369 4 Broadie, K. and Bate, M. (1993) Activitydependent development of the neuromuscular synapse during Drosophila embryogenesis. Neuron 11, 607–619 5 Malenka, R.C. and Nicoll, R.A. (1999) Long-term potentiation – a decade of progress? Science 285, 1870–1874 6 O’Brien, R.J. et al. (1999) Synaptic clustering of AMPA receptors by the extracellular immediateearly gene product Narp. Neuron 23, 309–323 7 Prokop, A. (1999) Integrating bits and pieces: synapse structure and formation in Drosophila embryos. Cell Tissue Res. 297, 169–186 8 Broadie, K. and Bate, M. (1993) Innervation directs receptor synthesis and localization in
385
9
10
11 12
13
14
15
16
17
18
Drosophila embryo synaptogenesis. Nature 361, 350–353 Saitoe, M. et al. (2001) Absence of junctional glutamate receptor clusters in Drosophila mutants lacking spontaneous transmitter release. Science 293, 514–517 Featherstone, D.E. et al. (2002) Developmental regulation of glutamate receptor field size by nonvesicular glutamate release. Nat. Neurosci. 5, 141–146 Verstreken, P. and Bellen, H.J. (2001) The meaning of a mini. Science 293, 443–444 Pennetta, G. et al. (1999) Dissecting the molecular mechanisms of neurotransmitter release in Drosophila. In Neurotransmitter Release (Bellen, H.J., ed.), pp. 304–351, Oxford University Press Zhang, B. and Ramaswami, M. (1999) Synaptic vesicle endocytosis and recycling. In Neurotransmitter Release (Bellen, H.J., ed.), pp. 389–431, Oxford University Press Featherstone, D.E. et al. (2000) Presynaptic glutamic acid decarboxylase is required for induction of the postsynaptic receptor field at a glutamatergic synapse. Neuron 27, 71–84 Besson, M.T. et al. (1999) Identification and structural characterization of two genes encoding glutamate transporter homologues differently expressed in the nervous system of Drosophila melanogaster. FEBS Lett. 443, 97–104 DiAntonio, A. et al. (1999) Glutamate receptor expression regulates quantal size and quantal content at the Drosophila neuromuscular junction. J. Neurosci. 19, 3023–3032 Zhai, R.G. et al. (2001) Assembling the presynaptic active zone: a characterization of an active zone precursor vesicle. Neuron 29, 131–143 Sheng, M. and Lee, S.H. (2001) AMPA receptor trafficking and the control of synaptic transmission. Cell 105, 825–828
Patrik Verstreken* Hugo J. Bellen‡ Program in Developmental Biology, Dept of Molecular and Human Genetics, Division of Neuroscience, Howard Hughes Medical Institute, Baylor College of Medicine, Houston, TX, USA. *e-mail: [email protected] ‡e-mail: [email protected]
Response: Meaningless minis? Minoru Saitoe, Thomas L. Schwarz, Joy A. Umbach, Cameron B. Gundersen and Yoshi Kidokoro First, let us restate our vesicular hypothesis. We are proposing that the agent that triggers receptor clustering at the postsynaptic membrane is released by the same mechanism as minis [1]. If the agent is contained in the large dense-cored vesicles, it is released with the same mechanism as clear vesicles. By contrast, http://tins.trends.com
the non-vesicular hypothesis [2] is proposed to explain different sizes of receptor clusters in various mutants, and postulates that non-vesicular release of glutamate downregulates the cluster size. It should be noted that receptor clusters have to form before this mechanism can work.
The major differences between our work and that of Featherstone et al. are that the latter demonstrated receptor clusters by immunofluorescence staining in Drosophila embryos with the null allele syx∆229 in ventral muscle number 6 [2], whereas we did not detect such immunofluorescence in dorsal muscles 1 and 9 of embryos with the
0166-2236/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0166-2236(02)02225-7
9 Tokumaru, H. et al. (2001) SNARE complex oligomerization by synaphin/complexin is essential for synaptic vesicle exocytosis. Cell 104, 421–432 10 Itakura, M. et al. (1999) Transfection analysis of functional roles of complexin I and II in the exocytosis of two different types of secretory vesicles. Biochem. Biophys. Res. Commun. 265, 691–696 11 Chen, X. et al. (2002) Three-dimensional structure of the complexin/SNARE complex. Neuron 33, 397–409
TRENDS in Neurosciences Vol.25 No.8 August 2002
12 Pabst, S. et al. (2000) Selective interaction of complexin with the neuronal SNARE complex. J. Biol. Chem. 275, 19808–19818 13 Pabst, S. et al. (2002) Rapid and selective binding to the synaptic SNARE complex suggests a modulatory role of complexins in neuroexocytosis. J. Biol. Chem. 277, 7838–7848 14 Xu, T. et al. (1999) Inhibition of SNARE complex assembly differentially affects kinetic components of exocytosis. Cell 99, 713–722
383
Karla E. Marz Phyllis I. Hanson* Dept of Cell Biology and Physiology, Washington University School of Medicine, 660 South Euclid Avenue, Box 8228, St Louis, MO 63110, USA. *e-mail: [email protected]
Meaningless minis? Mechanisms of neurotransmitterreceptor clustering Patrik Verstreken and Hugo J. Bellen Initiation and maintenance of the postsynaptic neurotransmitter-receptor field are important steps during synapse formation and maturation, as they play a determinative role in regulating synaptic strength. However, the mechanisms directing neurotransmitter-receptor clustering and maintenance are poorly understood. Recently, two models explaining glutamate-receptor clustering at the Drosophila neuromuscular junction have been proposed. One model postulates that release of an agent via single vesicle fusion events (minis) is required for the initiation of postsynaptic glutamatereceptor clustering, and that glutamate is not responsible for initiation or maintenance of the postsynaptic receptor field. The other model rules out a role for minis in initiation of clustering, and suggests a role for non-vesicular release of glutamate in receptor-field maintenance. Here, we compare and discuss the data underlying both models.
For neurons to communicate effectively with one another it is crucial that the presynaptic density, the site at which neurotransmitters are released, is aligned with the postsynaptic receptor field. In addition, the abundance of the receptors in the postsynaptic field is known to limit the postsynaptic response. Hence, alignment and size of the postsynaptic receptor field, as well as those of the presynaptic density, are important determinants in regulating synaptic strength [1]. Numerous vertebrate neurons communicate with each other using glutamate as an excitatory neurotransmitter. Glutamate is also utilized as a neurotransmitter at the http://tins.trends.com
fruit-fly neuromuscular junction (NMJ). The postsynaptic clustering of receptors is believed to be mediated by the presynaptic neuron, because glutamate-receptor clustering at the postsynaptic site depends on presynaptic activity [2–4]. The presynaptic signal seems to induce a postsynaptic cascade of events involving many proteins, including kinases and vesicle trafficking proteins in vertebrate cells, that eventually leads to the regulation of glutamate-receptor insertion into the membrane and clustering at the synapse [5]. The nature of the presynaptic signal resulting in glutamate-receptor clustering is currently unknown in Drosophila, but a candidate, neuronal activity regulated pentraxin (NARP), has been identified in the mouse [6]. As in mammalian neurons, the trigger to cluster glutamate receptors at the NMJ of Drosophila embryos is believed to originate presynaptically (because the clustering is an activity-dependent process, requiring a contact between the motor nerve ending and the muscle). Glutamate receptors are known to be present in muscles before synapse formation, but do not cluster at the postsynaptic density before nerve contact or in the absence of electrical activity [4,7,8]. Hence, nerve–muscle contact and electrical activity are essential for initial clustering. However, many crucial questions remain unanswered, or are at least the subject of intense debate. Which substances induce and maintain clustering? Are different substances involved in induction and maintenance? Are minis required for clustering? Do different types of vesicles control the processes of membrane addition to the
synapse, neurotransmitter release and release of clustering agents? Do these forms of vesicular release all depend on the same protein machinery? Are there non-vesicular forms of release? In the next sections we compare data presented in two studies, one by Saitoe et al. [9] and one by Featherstone et al. [10]. We outline the proposed models that try to answer some of these questions. Initiation of receptor clustering
The availability of Drosophila mutants with blocked neurotransmitter release has allowed investigation of the role of synaptic transmitter release in receptor clustering. Some mutations that block evoked release [e.g. synaptobrevin, cysteine string protein (CSP)] do not block the spontaneous fusion events that are measured as minis, whereas other mutations block both evoked and spontaneous release (e.g. shibire or syntaxin) [11,12,13]. Saitoe et al. [9] showed that glutamate receptors clustered normally when only evoked activity was blocked in synaptobrevin or CSP mutants. However, when both evoked and spontaneous neurotransmitter release were blocked, using syntaxin or shibire mutants, the authors reported no clustering of postsynaptic glutamate receptors. These data indicate that spontaneous neurotransmitter release, usually considered merely the leakage of transmitter from single vesicle fusions, plays a paramount role in inducing postsynaptic receptor clustering [9,11] (Fig. 1a). Although conceptually appealing (minis, thus, having a meaning), these data were recently challenged by Broadie and colleagues (University of Utah,
0166-2236/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0166-2236(02)02197-5
Research Update
384
(a)
TRENDS in Neurosciences Vol.25 No.8 August 2002
(b)
Initiation of receptor clustering; Saitoe et al. (2001) Synaptic vesicle Glutamate receptor subunit Endosome
Regulation of receptor cluster size; Featherstone et al. (2002) Excitatory amino acid transporter Pre-synaptic glutamate
2 1 1 2
Initiation: Minis are required, evoked release is not required Regulation of cluster size: Glutamate most likely has no role (at least) it does not act through glutamate receptors
Initiation: Neither minis nor evoked release is required Regulation of cluster size: Non vesicular glutamate release is a negative regulator of receptor cluster size TRENDS in Neurosciences
Fig. 1. Boutons at the neuromuscular junction of Drosophila. (a) Saitoe et al. have proposed a model in which spontaneous vesicle fusion (1) induces postsynaptic receptor clustering (2) [9]. According to the model, the fusion of a single vesicle not only results in liberation of neurotransmitter into the synaptic cleft, but possibly also leads to release of a factor inducing clustering of glutamate receptors. This factor has yet to be identified. Maintenance of receptor cluster size is, according to Saitoe et al., not dependent on glutamate in the cleft. (b) In contrast to Saitoe et al., Featherstone et al. have proposed a model in which glutamate-receptor cluster size is controlled by the abundance of presynaptic cytosolic glutamate [10]. In their model, Featherstone et al. propose that glutamate from a non-vesicular source is a signaling molecule that negatively regulates receptor clustering. (1) Glutamate could be released into the synaptic cleft by reversal of the polarity of excitatory amino-acid transporters, but to date no in vivo evidence supports this transport mechanism. (2) The presynaptic signal could be a regulator of the glutamatereceptor cycle. Glutamate receptors are continuously internalized and transported to the endosome. They are re-inserted into the plasma membrane (arrows) and they cluster at the active zone [18]. Regulation of cycling rates and regulation of receptor breakdown by trafficking to the lysosome (not shown) would allow for tight control of receptor cluster size at the postsynaptic density.
UT, USA) [10]. Featherstone et al. showed that glutamate-receptor clustering did occur in the absence of evoked and spontaneous release, indicating that minis are not required [10] (Fig. 1b). Like Saitoe et al., they used syntaxin and shibire ts1 mutants to block endogenous, as well as spontaneous, vesicle fusion [12,13], and both teams used similar techniques to assay glutamate-receptor clustering. Technical issues that could explain some of the differences between the studies include the presence of maternal component in syntaxin mutants, and cell death in shibire ts1 mutants that are kept at the restrictive temperature for protracted periods. In addition, the functional consequences of removal of syntaxin or shibire (demonstrating the absence of release) ought to be measured in the same muscles as those used for assaying glutamate-receptor clustering – yet, this was not done by Saitoe et al. [9]. In summary, further studies will have to be carried out to resolve these discrepancies, but the observation by Featherstone et al. that clustering does http://tins.trends.com
occur in the shibirets1 and syntaxin mutants must be viewed as strong positive evidence for clustering in the absence of spontaneous release [10]. Maintenance of receptor clustering
Another issue relates to the maintenance of receptor clustering once initial clustering has occurred. Maintenance of postsynaptic receptor clusters is, according to Saitoe et al., not dependent on glutamate in the synaptic cleft [9]. They injected glutamate together with argiotoxin, an activity-dependent glutamate channel blocker, into developing wild-type embryos, and determined if glutamate-receptor clustering was affected at the end of embryonic development. If glutamate or the glutamate receptors have an effect on receptor cluster size, then blocking the glutamate receptor could affect cluster formation or maintenance. The authors argue that they observed no difference with control animals, suggesting that neither glutamate nor its receptor affect cluster size. Hence, they propose that an
instructive signal for receptor clustering and maintenance, other than glutamate, is released by spontaneous fusion of single vesicles [9] (Fig. 1a). Featherstone et al. come to a different conclusion: they propose that glutamate released into the cleft by a non-vesicular mechanism does affect maintenance of the cluster size [10]. In previous work, Featherstone et al. identified mutations in the gene encoding Drosophila glutamicacid decarboxylase (GAD) [14]. This enzyme is present at the Drosophila NMJ and converts glutamate into GABA. Although GABA is not a neurotransmitter at the Drosophila NMJ, GAD was shown to play an important role in postsynaptic glutamate-receptor field regulation. This finding led to the idea that the levels of presynaptic glutamate might control the level of postsynaptic receptor cluster size. Indeed, when the presynaptic terminal contains more glutamate (e.g. in mutants with blocked glutamate breakdown, like GAD), fewer glutamate receptors cluster postsynaptically at the NMJ. Conversely, when the motoneuron endings contain less glutamate (achieved by accelerating glutamate breakdown), significantly more glutamate receptors cluster at the NMJ [10]. Featherstone et al. found that presynaptic glutamate controls the postsynaptic transmitter-receptor field size independently of synaptic vesicle mediated neurotransmitter release (i.e. independently of both evoked and spontaneous synaptic vesicle fusion). This led the authors to propose that presynaptic glutamate is released into the synaptic cleft, by a non-vesicular mechanism, to act as a negative regulator of postsynaptic glutamate-receptor cluster size [10]. One possibility is that excitatory amino-acid transporters (EAATs) release small amounts of glutamate into the synaptic cleft by reversing their polarity. Two EAATs are present in the Drosophila nervous system [15]; however, no in vivo data are available to suggest the existence of non-vesicular glutamate release at the Drosophila NMJ (Fig. 1b). Again the two sets of data contradict each other. The work by Saitoe et al. implies that glutamate plays no part in regulation of receptor cluster size [9], whereas Featherstone et al. propose a model in which glutamate itself is the effector of field-size regulation in the absence of vesicular release [10]. However, both claims are on loose ground,
Research Update
not supported by solid data. Saitoe et al. showed that extracellular glutamate does not influence receptor cluster size when the glutamate receptor is blocked using argiotoxin, an open channel blocker. However, the inhibition of glutamate channel activity by argiotoxin is not absolute [16] and furthermore, the authors also co-injected the argiotoxin with glutamate to open and block the glutamate receptors. This could well alleviate the effect of the toxin – addition of glutamate could rescue the partial block in glutamate-receptor activity by the toxin. Hence, a dramatic effect on glutamatereceptor clustering is not expected per se. By contrast, Featherstone et al. provide no direct evidence that glutamate secreted by the presynaptic cell is acting to regulate receptor downregulation on the postsynaptic side [10]. Other mechanisms, possibly regulated by presynaptic glutamate, could also maintain and regulate postsynaptic glutamate-receptor clusters. Notably, in the mouse a glutamate-receptor clustering agent, NARP, was identified, and there is a Drosophila homologue of NARP (GenBank accession number Y17570; Flybase: CG3100) [11]. This homologue could also be involved in clustering of glutamate receptors at the fly NMJ. In addition, recent advances in the quest to understand the development of the cytomatrix at the active zone have led to identification of a population of 80 nm diameter dense core vesicles. These vesicles carry active-zone components, but not synaptic-vesicle-specific proteins [17]. The emerging model predicts that these vesicles fuse with the presynaptic membrane on stimulation and deposit fresh active zone material (e.g. syntaxin,
TRENDS in Neurosciences Vol.25 No.8 August 2002
SNAP-25). Hence, it is possible that fusion events of a different nature to minis are required for pre- and postsynaptic density development. However, as these vesicles themselves transport part of the synaptic vesicle fusion machinery to the active zone, it is currently not known if fusion of these specialized vesicles requires the classical synaptic vesicle fusion machinery. In conclusion, the studies by Saitoe et al. [9] and Featherstone et al. [10] aim to identify the mechanisms leading to the initiation of receptor clustering and regulation of cluster size. The studies present contradicting data and propose conflicting models. Future work, both in vertebrates and in Drosophila, will be needed to shed light on the mechanisms of initiation and maintenance of postsynaptic glutamate-receptor clustering. References 1 Turrigiano, G.G. (2000) AMPA receptors unbound: membrane cycling and synaptic plasticity. Neuron 26, 5–8 2 Lissin, D.V. et al. (1998) Activity differentially regulates the surface expression of synaptic AMPA and NMDA glutamate receptors. Proc. Natl. Acad. Sci. U. S. A. 95, 7097–7102 3 O Brien, R.J. et al. (1998) Molecular mechanisms of glutamate receptor clustering at excitatory synapses. Curr. Opin. Neurobiol. 8, 364–369 4 Broadie, K. and Bate, M. (1993) Activitydependent development of the neuromuscular synapse during Drosophila embryogenesis. Neuron 11, 607–619 5 Malenka, R.C. and Nicoll, R.A. (1999) Long-term potentiation – a decade of progress? Science 285, 1870–1874 6 O’Brien, R.J. et al. (1999) Synaptic clustering of AMPA receptors by the extracellular immediateearly gene product Narp. Neuron 23, 309–323 7 Prokop, A. (1999) Integrating bits and pieces: synapse structure and formation in Drosophila embryos. Cell Tissue Res. 297, 169–186 8 Broadie, K. and Bate, M. (1993) Innervation directs receptor synthesis and localization in
385
9
10
11 12
13
14
15
16
17
18
Drosophila embryo synaptogenesis. Nature 361, 350–353 Saitoe, M. et al. (2001) Absence of junctional glutamate receptor clusters in Drosophila mutants lacking spontaneous transmitter release. Science 293, 514–517 Featherstone, D.E. et al. (2002) Developmental regulation of glutamate receptor field size by nonvesicular glutamate release. Nat. Neurosci. 5, 141–146 Verstreken, P. and Bellen, H.J. (2001) The meaning of a mini. Science 293, 443–444 Pennetta, G. et al. (1999) Dissecting the molecular mechanisms of neurotransmitter release in Drosophila. In Neurotransmitter Release (Bellen, H.J., ed.), pp. 304–351, Oxford University Press Zhang, B. and Ramaswami, M. (1999) Synaptic vesicle endocytosis and recycling. In Neurotransmitter Release (Bellen, H.J., ed.), pp. 389–431, Oxford University Press Featherstone, D.E. et al. (2000) Presynaptic glutamic acid decarboxylase is required for induction of the postsynaptic receptor field at a glutamatergic synapse. Neuron 27, 71–84 Besson, M.T. et al. (1999) Identification and structural characterization of two genes encoding glutamate transporter homologues differently expressed in the nervous system of Drosophila melanogaster. FEBS Lett. 443, 97–104 DiAntonio, A. et al. (1999) Glutamate receptor expression regulates quantal size and quantal content at the Drosophila neuromuscular junction. J. Neurosci. 19, 3023–3032 Zhai, R.G. et al. (2001) Assembling the presynaptic active zone: a characterization of an active zone precursor vesicle. Neuron 29, 131–143 Sheng, M. and Lee, S.H. (2001) AMPA receptor trafficking and the control of synaptic transmission. Cell 105, 825–828
Patrik Verstreken* Hugo J. Bellen‡ Program in Developmental Biology, Dept of Molecular and Human Genetics, Division of Neuroscience, Howard Hughes Medical Institute, Baylor College of Medicine, Houston, TX, USA. *e-mail: [email protected] ‡e-mail: [email protected]
Response: Meaningless minis? Minoru Saitoe, Thomas L. Schwarz, Joy A. Umbach, Cameron B. Gundersen and Yoshi Kidokoro First, let us restate our vesicular hypothesis. We are proposing that the agent that triggers receptor clustering at the postsynaptic membrane is released by the same mechanism as minis [1]. If the agent is contained in the large dense-cored vesicles, it is released with the same mechanism as clear vesicles. By contrast, http://tins.trends.com
the non-vesicular hypothesis [2] is proposed to explain different sizes of receptor clusters in various mutants, and postulates that non-vesicular release of glutamate downregulates the cluster size. It should be noted that receptor clusters have to form before this mechanism can work.
The major differences between our work and that of Featherstone et al. are that the latter demonstrated receptor clusters by immunofluorescence staining in Drosophila embryos with the null allele syx∆229 in ventral muscle number 6 [2], whereas we did not detect such immunofluorescence in dorsal muscles 1 and 9 of embryos with the
0166-2236/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0166-2236(02)02225-7