Neuron, Vol 36, 675-688, November 2002 Localized Synaptic Potentiation by BDNF Requires Local Protein Synthesis in the Developing Axon Xiao-hui Zhang and Mu-ming Poo Division of Neurobiology, Department of molecular and Cell Biology, University o f California, Berkeley, Berkeley, CA 94720 USA Correspondence: Mu-ming Poo 510-642-2514 (phone) 510-642-2544 (fax) mpoo@uclink.berkeley.edu Summary Brain-derived neurotrophic factor (BDNF) is known to promote neuronal survival, guide axonal pathfinding, and participate in activity-dependent synaptic plastic ity. In Xenopus nerve-muscle cultures, localized contact of a single BDNF-coated bead with the presynaptic axon resulted in potentiation of transmitter secretio n at the developing synapses, but only when the bead was placed within 60 μ ;m from the synapse. The localized potentiation induced by BDNF is accompanied b y a persistent local elevation of [Ca2+]i in the axon and requires constitutive presynaptic protein translation, even for axons severed from the cell body. Thus , presynaptic local TrkB signaling and protein synthesis allow a localized sourc e of BDNF to potentiate transmitter secretion from nearby synapses, a property s uited for spatially restricted synaptic modification by neurotrophins. Introduction Summary Introduction Results Discussion Experimental Procedures References A unique feature of the nervous system is the complex morphological phenotype of neurons and the intricate network of specific nerve connections. Developing axo ns undergo long-range pathfinding to reach their appropriate target cells to for m synaptic connections. After the initial synaptic connections are established, substantial remodeling of these connections may occur, depending on the pattern of synaptic activity. At the cell biological level, it remains largely unknown h ow the neuron carries out its task of transporting newly synthesized constituent s to the tip of the growing axon or presynaptic nerve terminals that are distant from the cell body. It is also unclear how localized synaptic activity triggers a synapse-specific modification that requires supplies of newly synthesized com ponents to be localized to selective synaptic sites. Local protein synthesis at the synaptic site may provide a potential mechanism for achieving synapse specif icity in activity-dependent long-term synaptic modification (Schuman, 1997 ; Ste ward, 1997 ; Martin et al., 2000 ; Steward and Schuman, 2001 ). The existence of translational machinery, e.g., mRNAs and ribosomes, in dendrites has been repor ted (Steward and Levy, 1982 ; Steward and Reeves, 1988 ; Steward, 1997 ), and lo cal protein translation has been demonstrated in dendrites of cultured hippocamp al neurons (Aakalu et al., 2001 ). There is also evidence that local protein syn thesis occurs in developing and mature axons (Davis et al., 1992 ; Olink-Coux an d Hollenbeck, 1996 ; Bassell et al., 1998 ; Eng et al., 1999 ; Koening and Giudi tta, 1999 ; Koening et al., 2000 ; Campbell and Holt, 2001 ; Zheng et al., 2001 ; Brittis et al., 2002 ). Local axonal protein synthesis appears to be required for effective chemotaxis of growth cone in a gradient of guidance molecules (Cam pbell and Holt, 2001 ; Ming et al., 2002 ). After synaptogenesis, local translat ion in the presynaptic axon may continue to be involved in activity-dependent sy naptic plasticity. At Aplysia sensory to motor synapses in cell cultures, for ex ample, presynaptic local protein synthesis is required for branch-specific long- term facilitation induced by serotonin (Martin et al., 1997 ). Even for mature c rayfish neuromuscular junctions, long-term facilitation also requires protein sy nthesis in the presynaptic axon (Beaumont et al., 2001 ). Brief periods of repetitive synaptic activity are known to induce long-term pote ntiation or depression of synaptic transmission (Malenka and Nicoll, 1999 ) as w ell as morphological changes of activated synapses and dendrites (Engert and Bon hoeffer, 1999 ; Maletic-Savatic et al., 1999 ; Yuste and Bonhoeffer, 2001 ). Neu rotrophins have emerged as prime candidate molecules for linking electric activi ty to synaptic changes (McAllister et al., 1999 ; Poo, 2001 ). The synthesis of neurotrophins in many regions of the brain (Zafra et al., 1990 ; Patterson et al ., 1992 ) and in skeletal muscles (Funakoshi et al., 1995 ) is regulated by elec trical activity. Membrane depolarization and synaptic activation (Bl?chl an d Thoenen, 1995 ; Goodman et al., 1996 ; Wang and Poo, 1997 ; Kojima et al., 200 1 ; Hartmann et al., 2001 ) can trigger secretion of neurotrophins, which may in turn modify synaptic structure or function. The latter possibility is supported by the findings that in some preparations, extracellularly applied neurotrophin s can modify basal synaptic transmission (Lohof et al., 1993 ; Kang and Schuman, 1995 ), reduce the synaptic fatigue under high frequency stimulation (Figurov e t al., 1996 ), and alter the growth and branching patterns of axonal and dendrit ic arbors (Diamond et al., 1992 ; Cohen-Cory and Fraser, 1995 ; McAllister et al ., 1995, 1996 ) . An outstanding issue concerning the role of neurotrophins in synaptic plasticity is the synapse specificity of their effects. Is the action of locally secreted neurotrophins spatially restricted? Gene regulation and protein synthesis are li kely to be required for long-term structural and functional modifications of syn apses, but how does the action of neurotrophin remain synapse-specific if global protein synthesis machinery is involved? In hippocampal slices, BDNF and neurot rophin-3 (NT-3) were shown to potentiate synaptic transmission at Schaffer colla teral-CA 1 synapses (Kang and Schuman, 1995 ). Interestingly, blocking protein s ynthesis in the slices in which the synaptic neuropil was isolated from both pre - and postsynaptic cell bodies attenuated the potentiation effects, suggesting t hat protein synthesis in either the dendritic or axonal compartment, or both, is required for the neurotrophin effect (Kang and Schuman, 1996 ). In the present study, we have measured quantitatively the spatial range of neurotrophin effect in synaptic potentiation and examined specifically whether presynaptic protein s ynthesis is involved in the potentiation effect at developing neuromuscular syna pses in culture. Results Summary Introduction Results Discussion Experimental Procedures References Potentiation of Spontaneous Transmitter Secretion by BDNF Beads In 1-day-old Xenopus nerve-muscle cocultures, functional synaptic transmission i s established between spinal neurons and myocytes (Evers et al., 1989 ). Bath ap plication of BDNF and NT-3 results in potentiation of spontaneous and evoked tra nsmitter secretion from the presynaptic neuron (Lohof et al., 1993 ). We first e xamined whether local delivery of BDNF to a presynaptic site can modulate basal synaptic transmission in these cultures. Single beads covalently coated with BDN F (BDNF beads) were manipulated into contact with the presynaptic axon at sites 15–25 μm from the synapse (Figure 1A). The presence of BDNF on those beads was confirmed by immunostaining with specific polyclonal antibodies raised against BDNF (Figure 1B). Miniature excitatory postsynaptic currents (MEPCs) were recorded from innervated myocytes by a whole-cell recording method. Within 8?C10 min following the contact of the BDNF bead, we observed a significant and pe rsistent increase in the frequency of MEPCs (Figures 1D and 1E). Pretreatment of the culture with K252a (200 nM), an inhibitor of receptor tyrosine kinase, abol ished the changes in MEPCs induced by the BDNF bead (Figures 1D and 1E), while K 252b (200 nM), a compound which does not inhibit tyrosine kinase at the concentr ation used, had no effect (n = 6; data not shown). This is consistent with the p revious result (Lohof et al., 1993 ) that the BDNF effect is mediated through Tr kB receptors in the presynaptic axon. The specificity of the effect of BDNF bead s was confirmed by the findings that preheating BDNF beads at 75°C for 45 min a bolished the potentiation effect of the bead (Figure 1E) and that beads coated w ith bovine serum albumin (BSA) or nerve growth factor (NGF), which is known to h ave no effect on these developing synapses (Lohof et al., 1993 ), caused no sign ificant increase in the frequency of MEPCs (Figure 1E). The change in the MEPC f requency appears to represent an enhanced probability of spontaneous quantal sec retion of acetylcholine (ACh) rather than an elevated postsynaptic sensitivity t o ACh. As shown in Figure 1F, the mean amplitude and the amplitude distribution of MEPCs did not show any significant change following the contact with the BDNF bead, as would be expected if postsynaptic ACh sensitivity were changed. Figure 1. Potentiation of Spontaneuous ACh Secretion by BDNF-Coated Beads (A) Phase-contrast image of a 1-day-old Xenopus culture in which a BDNF bead ('' B'') was manipulated into contact with the axon of a spinal neuron (''N'') at a site ~22 μm from an innervated myocyte (''M''). R, recording pipette. Bar = 15 μm. (B) Immunostaining of BDNF beads. (Ba and Bb) Phase-contrast and fluorescence im ages, respectively, of a BDNF bead immunostained with anti-BDNF polyclonal antib ody and rhodamine-conjugated secondary antibody. (Bc and Bd) Same as (Ba) and (B b), except that the primary antibodies were preincubated with the blocking pepti des prior to application to BDNF beads. (C) Immunostaining of BSA-coated beads. Similar to (B), except that anti-BSA mon oclonal antibody was used as the primary antibody (Ca and Cb), and only secondar y antibody was used in the control (Cc and Cd). Bar = 10 μm. (D) Examples of recordings from innervated myocytes after contacts with a single BDNF bead or BSA bead. Continuous traces depict membrane currents of the myocyt e (clamped at -70mV). Downward events are MEPCs (samples shown below at a higher time resolution). K252a: 30 min pretreatment and the drug was present throughou t the recording (at 200 nM). Scales: 1.5 nA, 5 min for slow traces; 0.85 nA, 15 ms for fast traces. (E) Changes in the MEPC frequency with time induced by the bead contact (at T = 0) at sites 25 μm from the synapse. Heat inactivation: 75°C for 45 min. T he MEPC frequency over the 3 min bin was normalized by the mean value before the bead contact for each experiment before averaging (± SEM). (F) Distribution of the MEPC amplitudes before and after a BDNF bead contact. Cu mulative probabilities during a 5 min period before and during 35–40 min after the bead contact were analyzed. Each value represents the mean ??SEM for 13 experiments shown in (E). There was no statistically significant difference between the two data sets (Kolmogorov-Smirnov test, p > 0.05). Insets: sample traces of MEPCs 5 min before and 35?C40 min after the BDNF bead contact. The averages of all MEPCs for the sample tr aces are shown below. Scales: 0.35 nA, 1 min for slow traces; 0.1 nA, 10 ms for fast traces. View larger version: [In this window] [In new window] Potentiation of Evoked Synaptic Transmission by BDNF Beads In addition to the potentiation of spontaneous ACh secretion, contact of single BDNF beads with the presynaptic axon also resulted in an increase in the mean am plitude of impulse-evoked postsynaptic currents (EPCs), which were recorded in i nnervated myocytes in response to stimulation of the presynaptic soma at a low f requency (0.03 Hz). The increase in the mean EPC amplitude correlated with a gra dual increase in the MEPC frequency with time (Figure 2A). Within 20–40 min after the bead contact, the mean EPC amplitude was 221% ??46% (SEM, n = 8 synapses) of the control value obtained before the bead contact. Furthermore, in the presence of K252a (200 nM), there was no significant change in the mean EPC amplitude after the bead contact (98% ??10%, SEM, n = 5; see Figure 4B). The effect of bead contact was presynaptic specific: no change in the mean EPC amplitude was observed at 20?C40 min after the contact of the BDNF bead with the postsynaptic muscle cell (88 % ± 12%, SEM, n = 3). Together with the results on MEPCs, these findings indica te that localized bead-induced activation of TrkB in the presynaptic axon close to the synapse was sufficient to potentiate both spontaneous and evoked transmit ter release at these developing synapses. Figure 2. BDNF Bead Effects Depend on the Distance to the Synapse (A and B) Examples of synaptic currents before and after the contact (at the tim e marked by the arrow) of a single BDNF bead at sites 50 (A) and 150 μm (B ) from the synapse. Trains of test stimuli (7–8 pulses at 0.033 Hz) were applied at 5 min intervals to evoke EPCs at the times marked by triangles. Shown below are averaged traces of EPCs at a higher time resolution. Dotted lines represent the mean EPC amplitude before the bead contact. Scales: 2 nA, 2.5 min for slow traces; 2.5 nA, 15 ms for fast traces. (C and D) Summary of all data for the dependence of synaptic potentiation on the distance from the BDNF bead contact site to the synapse. Each circle represents the mean EPC amplitude (C) and mean MEPC frequency (D) 20–40 min after the bead contact, normalized by the mean control values observed at the same synapse before the bead contact. Solid triangles: mean ??SEM over 20 μm bins. Multiple BDNF beads: 3 or 4 beads were manipulated into contact with axon 90 μm from the synapse. Local puffing of BDNF: repetitive pressure injections of BDNF from a micropipette at sites >100 μm from the synapse. (E) Dependence of synaptic potentiation (as shown by the normalized MEPC frequen cy at 20–40 min) on the distance of the BDNF bead contact site from the soma (same data set as in [D]). View larger version: [In this window] [In new window] Figure 4. Synaptic Potentiation Induced by BDNF Beads Requires Protein Synthesis (A) Experiments similar to that illustrated in Figure 2A, except that the cultur e was pretreated with anisomycin (''Aniso,'' 30 μM) for 45 min or 2 hr. Tr aces shown are recordings of synaptic currents made during a 10 min period befor e and at 20–30 min after the contact (at the time marked by the arrow) with a single BDNF bead or NT-3 bead. Scales: 1 nA, 1.5 min. (B) Summary of data for the mean EPC amplitude (± SEM) at 20–40 min after the bead contact under various treatments, normalized by the mean control values from the same synapse before the bead contact. 0.01 < *p < 0.05; **p < 0.01 (Student's t test). K252a, 200 nM. (C) Changes in the MEPC frequency with time induced by contact of BDNF beads or NT-3 beads in cultures pretreated with anisomycin for 2 hr. Each data point repr esents the mean ± SEM. View larger version: [In this window] [In new window] Spatial Range of BDNF-Induced Potentiation Taking advantage of the precise localization of BDNF beads, we have examined qua ntitatively the spatial range of BDNF action by manipulating single BDNF beads i nto contact with presynaptic axons at various distances from the synapse. We obs erved a clear distance dependence in the action of BDNF beads; the potentiation of both the mean EPC amplitude and mean MEPC frequency was found only for bead c ontact sites within a distance of about 60 μm from the synapse (Figure 2). An example of a recording for a single bead contact at a remote location of 150 μm from the synapse is shown in Figure 2B. The restricted spatial range o f BDNF action cannot be attributed simply to a limited source of BDNF on a singl e bead, since we observed no synaptic potentiation when the source of BDNF was i ncreased at a site remote to the synapses by placing three or more BDNF beads in to contact with the presynaptic axon (distance from synapses 90 μm) or by locally puffing BDNF (5 μg/ml) to the axon (distance from synapses 110 & micro;m) (Figure 2D). The spatial restriction of BDNF action thus reflects not a limited extent of TrkB activation by a single bead, but a spatially restricted range of action of downstream signal cascades triggered by the activated TrkB. F inally, we found no apparent correlation between the degree of synaptic potentia tion and the distance between the bead and the soma (Figure 2E), suggesting that localized BDNF action may not depend on signaling to or from the soma. Thus loc ally secreted BDNF at the synapse can produce highly localized modulatory effect s on synaptic efficacy. Localized Elevation of Ca2+ Synaptic potentiation by BDNF is known to depend on extracellular Ca2+ (Stoop an d Poo, 1995 ). Bath application of BDNF increased the cytoplasmic Ca2+ level ([C a2+]i) persistently in the soma, axon, and synaptic terminals of cultured Xenopu s spinal neurons (Stoop and Poo, 1996 ). Such elevation of [Ca2+]i and subsequen t activation of downstream signaling molecules may be responsible for the BDNF-i nduced synaptic potentiation. To examine the spatial relationship between [Ca2+] i elevation and synaptic potentiation induced by BDNF beads, we have measured ch anges in [Ca2+]i within the presynaptic axon of the spinal neuron. The neurons w ere microinjected with a fluorescent Ca2+-sensitive dye, Oregon Green BAPTA-1 co njugated to dextran (70 kDa). As shown in Figure 3A, we observed a distinct and spatially restricted [Ca2+]i increase (F/Fo = 106%) in the presynaptic axon adja cent to the site of bead contact 30 min after the contact with a single BDNF bea d at distal end of the axon, whereas only a slight change in [Ca2+]i (F/Fo = ~10 %) was observed in the soma or other axonal regions. In another experiment (Figu re 3B), the contact of a single BDNF bead at the proximal end of the axon elicit ed a [Ca2+]i elevation at the site of contact as well as at the nearby soma, but not at the distal end of the axon. In the latter case, the rise in [Ca2+]i at t he soma appeared to be slightly slower than that at the bead contact site. On th e average, significant [Ca2+]i elevation (F/Fo > 10%) was preferentially restric ted within about 40 μm from the site of bead contact (Figure 3C). Figure 3. Localized Ca2+ Elevation Induced by the BDNF Bead Contact (A–C) Bright field and Oregon Green-488 BAPTA fluorescence images of the spinal neuron (''N'') 3 min before and 40 min after BDNF bead (''B'', arrows in [A] and [B]) contact with the axon, or 15 min after focal puffing of soluble BDNF with a micropipette (''P'' in [C]). The intensity of the fluorescence is coded by pseudocolors (brighter colors corresponding to higher values, see color bar). Scale: 20 μm (A), 10 μm (B and C). Graphs on the right show fluorescence intensity changes with time at selected regions in the neurons shown on the left (marked by numbered boxes). Black bar: duration of local puffing of BDNF. (D) Spatial distribution of fluorescence changes along the axon induced by the B DNF bead contact (at the site defined as ''0'' μm). The values from each n euron were normalized to the mean value during the control period before the bea d contact before grand averaging (mean ± SEM, n = 7 neurons). (E) Changes in fluorescence intensity over time after the contact with BDNF bead s or BSA beads (same data set as in [C]). View larger version: [In this window] [In new window] The amplitude and spatial range of bead-induced changes in [Ca2+]i along an axon varied among the seven presynaptic neurons examined. In 3/7 cases, axonal [Ca2+ ]i elevation was found to be restricted to a range of less than 10 μm from the contact site. The rise in [Ca2+]i was gradual over tens of minutes after th e bead contact and persisted at the elevated level for as long as the recording was made (up to 1 hr in 4/7 cases). Consistent with its ineffectiveness in induc ing the synaptic potentiation, contact of the BDNF bead with the muscle Cell, ei ther innervated or isolated, caused no change of [Ca2+]i in the muscle cell (F/F o = 2.5% ± 2.3%, 30 min after bead contact, SEM, n = 4). In further control exp eriments, we found that contact with BSA beads did not change [Ca2+]i in the pre synaptic axon (Figure 3E). In addition, we found that local [Ca2+]i elevation wa s induced at sites both close (<50 μm,="" n="7)" and="" distant="" (="">90 μm, n = 3) from the synapse, suggesting that failure of synaptic potentiation by dist ant contact of the BDNF bead was not due to the absence of TrkB signaling. To further determine whether soluble BDNF, when applied locally to the presynapt ic axons, also induces a localized elevation of [Ca2+]i, we focally puffed solub le BDNF (1 μg/ml) to the presynaptic axons by repetitive pulsatile ejectio ns of BDNF from a micropipette (2 Hz, 20 ms, 3 psi, and 1 μm tip opening) in the presence of a constant perfusion flow that washed the ejected BDNF away f rom the neuron. As shown in Figure 3C, localized increase of [Ca2+]i was observe d within 10 min after the onset of BDNF puffing, and the [Ca2+]i gradually decli ned to the control level with a time course of about 20–25 min after the puffing was terminated. Similar local changes in [Ca2+]i were observed in two other puffing experiments. Effects of Protein Synthesis Inhibitor Using local application of BDNF beads, which avoids direct postsynaptic actions of BDNF, we have examined the role of protein synthesis in BDNF-induced synaptic potentiation. We found that a 45 min pretreatment of anisomycin (30 μM), a protein synthesis inhibitor, did not significantly affect synaptic potentiatio n induced by BDNF beads, which made contacts with the presynaptic axon at 20–40 μm from the synapse, as shown by an increased EPC amplitude and MEPC frequency (Figures 4A and 4B). However, after the 2 hr incubation with anisomycin, BDNF beads became ineffective in inducing synaptic potentiation at contact sites 20?C40 μm from synapses (Figures 4A–4C). The inhibitory effect of 2 hr anisomycin incubation was not due to a deleterious effect of the drug on the transmitter secretion machinery, because after the same treatment of anisomycin, NT-3-coated beads were still effective in potentiating the synapse (Figures 4A?C4C). Assuming that 45 min treatment of anisomycin had completely prevented new protein synthesis, the above results suggest that synaptic potentiation does not require BDNF-induced protein synthesis. On the other hand, continuous protein s ynthesis is required for synaptic potentiation, since prolonged (2 hr) absence o f translation prevented synaptic potentiation. Taken together, our results indic ate that the presynaptic action of localized BDNF depends on constitutive protei n synthesis, whereas that of NT-3 does not. We have also examined whether Ca2+ signaling induced by BDNF in these spinal neu rons also requires protein synthesis. As shown in Figure 5A, we found that the [ Ca2+]i increase in the axon induced by bath application of BDNF (100 ng/ml) can be completely blocked by a 2 hr pretreatment of the culture with anisomycin (30 μM). Those treated axons, although unresponsive to bath-applied BDNF, stil l responded with [Ca2+]i elevation when a high K+ (60 mM) solution was subsequen tly applied. Pretreatment of the culture with anisomycin for 45 min partially re duced the BDNF-induced [Ca2+]i elevation in the axon (Figures 5B and 5C). Thus, Ca2+ signaling mediated by BDNF/TrkB also requires continuous protein synthesis in these neurons. Figure 5. Inhibitory Effects of Anisomycin on BDNF-Induced Ca2+ Elevation (A and B) Fluorescence images of the axon of spinal neurons 3 min before and 40 min after bath application of BDNF (100 ng/ml) in the cultures pretreated with a nisomycin (30 μM) for 2 hr (A) or 45 min (B). In (A), 60 mM K+ was added 4 5 min after BDNF application. Scale bars: 20 μm (A), 10 μm (B). (C) Summary of changes in [Ca2+]i induced by bath application of BDNF in normal cultures and in cultures pretreated with anisomycin for 45 min or 2 hr. 0.01 < * p < 0.05; **p < 0.01. View larger version: [In this window] [In new window] Pre- versus Postsynaptic Protein Synthesis To further determine whether pre- or postsynaptic protein synthesis is required for the synaptic potentiation induced by the BDNF bead, we selectively injected gelonin—a membrane-impermeant ribosome-inactivating protein (Stirpe et al., 198 0 )—into the pre- or postsynaptic cell, together with fluorescently labeled dex tran as a marker. When BDNF beads were manipulated into contact with the presyna ptic axon near the synapse 2 hr after presynaptic injection of gelonin and dextr an, we observed no potentiation of spontaneous ACh release ( Figures 6Aa and 6B) . In contrast, postsynaptic injection of gelonin and dextran did not affect the potentiation of spontaneous ACh release induced by the BDNF bead ( Figures 6Ab a nd 6B). Injection of fluorescent dextran alone into either the pre- or postsynap tic cell also had no effect on the BDNF-induced potentiation (Figure 6B). Thus, presynaptic protein synthesis appears be required for BDNF-induced synaptic pote ntiation. Figure 6. Synaptic Potentiation by BDNF Beads Requires Presynaptic Protein Synthesis (A) Effects of microinjection of gelonin and FITC-dextan into the presynaptic ne uron (Aa) or the postsynaptic myocyte (Ab) on BDNF bead-induced synaptic potenti ation. Phase-contrast and FITC fluorescence images (bar = 10 μm) are shown together with example recordings of MEPCs before and after the bead contact (ma rked by the arrow, scales: 1 nA, 5 min). (B) Summary of data for the mean MEPC frequency (± SEM) at 30–40 min after the BDNF bead contact, normalized by the control values prior to the contact. **p < 0.01. View larger version: [In this window] [In new window] Protein Synthesis Is Required for Synaptic Potentiation in Transected Axons Previous studies on these Xenopus cultures have shown that synaptic potentiation by bath-applied BDNF does not require signaling to and from the presynaptic som a, whereas that induced by ciliary neurotrophic factor does (Stoop and Poo, 1995 ). This is consistent with our present finding that there is no clear correlati on between the degree of synaptic potentiation and the distance of the bead from the soma (Figure 2E). These findings suggest that local signaling and protein s ynthesis in the axon may be sufficient for the BDNF action. To directly test the role of local protein synthesis, we examined the effect of BDNF beads on synapt ic efficacy after transection of the axon near the soma with a sharp micropipett e (Figure 7A). Immediately after the transection, spontaneous ACh secretion exhi bited a transient increase, presumably due to a surge of Ca2+ influx at the inju red site, and then gradually subsided to a normal level (Figure 7A). In these ex periments, only synapses with long axons (100 μm) were used, and the ''cut loose'' axons were allowed to recover for 2 hr before a BDNF bead was manipulat ed into contact with the axon near the synapse. We found that the frequency of s pontaneous ACh release gradually decreased within the 2 hr recovery period after transection. However, for bead contacts at 20–40 μm from the cut-loose synapse, we observed a significant increase in MEPC frequency similar to that found in intact neurons. Moreover, when transected axons were treated for 2 hr after axon transection with either anisomycin (30 μM) or cycloheximide (50 μg/ml), two structurally different protein synthesis inhibitors, we found no potentiation effect of the BDNF bead on the cut-loose synapse (Figure 7B). In contrast, the same treatment with the transcription inhibitor actinomycin D (10 μg/ml) did not affect the potentiation induced by the BDNF bead (n = 4; data not shown). Taken together, these results strongly support the notion that presynaptic protein synthesis, but not RNA transcription, within the developing axon is required for synaptic potentiation induced by localized exposure to BDNF. Figure 7. Synaptic Potentiation by BDNF Beads Requires Local Protein Synthesis. (A) A phase-contrast image of a cut-loose synapse after the axon was transected (at marked arrows) from the soma of a spinal neuron (''N''). A BDNF bead (''B'') was manipulated into contact with the axon at a site close to the innervated my ocyte (''M''). Scale bar: 20 μm. An example recording of MEPCs from a syna pse in which the presynaptic axon was transected at time marked by the vertical bar. Scale: 1 nA, 4 min. (B) Summary of data for changes in the MEPC frequency with time after contact wi th the BDNF bead (T = 0) in the absence and the presence of protein synthesis in hibitors anisomycin (30 μM) or cycloheximide (50 μg/ml), which was a pplied for 2 hr after the transection. Control: no BDNF bead contact and no inhi bitor treatment. (C) Nucleic acid labeling of a spinal neuron with Hoecchast 33342 and SYTO 64 dy es. White arrowhead: Hoechst staining of the nucleus. Scale bar: 15 μm. (D) Immunostaining of cultured spinal neurons (''N'') and muscle cells (''M'') w ith an anti-ribosomal P (RPP) antigen antibody. The F-actin labeling by Alex-568 -phallotoxin helped to visualize the fixed cells. Control: no primary antibody w as applied. Scale bar: 15 (upper) or 10 μm (lower). View larger version: [In this window] [In new window] The Presence of Translation Machinery in the Axon To investigate the distribution of RNAs and ribosomes in these neurons, we used a membrane-permeable dye, SYTO 64, that fluoresces on binding to nucleic acids a nd an anti-ribosomal P antigen antibody that recognizes three protein components (P0, P1, and P2) of the 60S subunit of the ribosome. As shown in Figure 7C, the axon of the spinal neuron showed SYTO 64 staining. That this SYTO 64 staining o f the axon represents RNA rather than DNA staining was further confirmed by the staining of the nucleus by Hoechst 33342, a membrane-permeable dye that specific ally binds to DNA (Figure 7C). The axons also showed strong positive staining wi th the anti-ribosomal P antigen antibody (Figure 7D). Labeling of F-actin with A lex568-phallotoxin helped to visualize the fixed cells in the latter experiments . Taken together, these staining results confirmed the presence of key component s of the translation machinery in the developing axon of these cultured spinal n eurons. Discussion Summary Introduction Results Discussion Experimental Procedures References In this study, we have examined the role of axonal protein synthesis in BDNF-ind uced synaptic potentiation of Xenopus neuromuscular synapses in culture, using b eads covalently coated with BDNF. We find that (1) synaptic potentiation induced by local application of BDNF to the presynaptic axon is spatially restricted to nearby synapses located within about 60 μm from the site of contact; (2) Ca2+ signals elicited by local BDNF application also spread over a similar spati al range along the axon; and (3) this localized potentiation effect of BDNF requ ires continuous protein synthesis within the presynaptic axon. Together, our dat a suggest that the spatially restricted action of BDNF is achieved by local TrkB signaling and protein synthesis within the axonal cytoplasm. It has been proposed that during activity-dependent synapse refinement, nerve te rminals compete for limiting amounts of neurotrophic factors secreted by the tar get cell (Purves and Lichtman, 1985 ). There is evidence that neurotrophins are required for the formation of ocular dominance columns and act as activity-depen dent signals to modulate the pattern of synaptic connectivity (Maffei et al., 19 92 ; Cabelli et al., 1995, 1997 ) . Application of exogenous neurotrophins can a lso induce changes in the morphology of axons and dendrites (Diamond et al., 199 2 ; Cohen-Cory and Fraser, 1995 ; McAllister et al., 1995, 1996 ) and promote t ransmitter secretion from the presynaptic nerve terminals (Lohof et al., 1993 ; Kang and Schuman, 1995 ). To serve for their roles in the formation and refineme nt of synaptic connections, the action of neurotrophins must be under precise sp atial and temporal regulation. The present results demonstrate directly that the synaptic action of BDNF can be localized. There appears to be an upper limit of about 60 μm for the spread of BDNF/TrkB-activated cytoplasmic effectors w ithin the axon. These results are also consistent with an earlier finding that s howed a similar range of restricted spread of synaptic potentiation caused by NT -4 secreted from the postsynaptic myocyte (Wang et al., 1998 ). Thus, the restri cted range of the synaptic action allows neurotrophins to serve as localized neu romodulators, specific to synapses adjacent to the secretion site. In the CA1 re gion of the hippocampus, activity-induced long-term potentiation (LTP) requires BDNF (Korte et al., 1995 ; see review by Poo, 2001 ). Interestingly, when LTP is locally induced at Schaffer collateral-pyramidal cell synapses, there is a spat ially restricted spread of potentiation to nearby synapses within a range of abo ut 70 μm (Engert and Bonhoeffer, 1997 ). Localized action of activity-indu ced BDNF secretion may account for the spatially restricted spread of LTP at cen tral synapses. Given the limited extracellular diffusion of secreted neurotrophins (Bl?chl and Thoenen, 1995 ; Wang et al., 1998 ), spatial specificity in BDNF-induced sy naptic potentiation may be achieved by differential localization of TrkB recepto rs or spatially restricted propagation of intracellular signaling molecules trig gered by the activated TrkB. The TrkB receptors are found at distinct neuronal c ompartments in various brain regions (Cabelli et al., 1996 ; Tongiorgi et al., 1 997 ; Drake et al., 1999 ; Aoki et al., 2000 ). The pattern of expression and su bcellular localization of TrkB in visual cortical neurons undergo developmental changes: an early appearance of TrkB on axon fibers and later expression on cell bodies and dendrites (Cabelli et al., 1996 ). Following the transfection of neu rons in hippocampal cultures or intact cortical slices with epitope-tagged trkB and trkC cDNAs, both receptors exhibited homogeneous distribution along the axon (Kryl et al., 1999 ). The distribution of endogenous TrkB (S. Andersen, persona l communication) and TrkC (Chang and Popov, 1999 ) in developing Xenopus axons a lso appeared to be relatively uniform. This is consistent with our finding that BDNF beads can induce local Ca2+ elevation at regions distant from the synaptic sites. Thus, localized synaptic actions of BDNF observed in the present study we re unlikely to result from a differential distribution of TrkB receptors near th e synapse. To further examine the spatial range of intracellular signals triggered by TrkB activation, we had focused our attention on Ca2+, which is known to mediate cyto plasmic signaling cascades initiated by BDNF. Synaptic potentiation induced by b ath-applied BDNF is accompanied by a gradual increase of [Ca2+]i in the soma, al ong the axon, and in the axonal terminals of cultured Xenopus spinal neurons (St oop and Poo, 1996 ). A similar [Ca2+]i increase in the soma has also been observ ed in cultured hippocampal neurons (Berninger et al., 1993 ; Li et al., 1998 ). In the present study, we found that the spread of axonal Ca2+ signals elicited b y the contact of the BDNF bead is restricted to a similar range as that found fo r bead-induced synaptic potentiation along the axon (Figure 3D), suggesting that Ca2+ elevation in axon terminals may be responsible for triggering synaptic pot entiation. As shown in Figure 3C, spatially restricted [Ca2+]i elevation can als o be elicited by focal application of soluble BDNF. Thus the spatial restriction of [Ca2+]i elevation induced by BDNF beads was not due to the fact that BDNF co valently bound to the bead cannot be internalized. It reflects a spatially confi ned cytoplasmic signaling triggered by localized TrkB activation. The nature of this local signal remains to be determined. It has been shown that the release of Ca2+ from internal endoplasmic reticulum v ia inositol 1,4,5-trisphosphate (InsP3) receptors and store-operated capacitativ e Ca2+ entry are critical for BDNF-induced [Ca2+]i elevation in cultured neurons (Li et al., 1998 ; Kleiman et al., 2000 ). Blocking either of these two Ca2+ so urces abolishes BDNF-induced synaptic potentiation. The BDNF-elevated Ca2+ can p romote spontaneous transmitter release directly by elevating the basal rate of s ynaptic vesicle exocytosis (Miledi, 1973 ) or indirectly by modulating the effic acy of vesicular fusion machinery via activation of the downstream effector(s) o f Ca2+. Indeed, synaptic vesicle-associated proteins—synapsin, synaptophysin, a nd synaptobrevin—have been implicated as downstream targets of the BDNF/TrkB si gnaling pathway (Pozzo-Miller et al., 1999 ). Application of BDNF to cortical sy naptosomes also elicits a mitogen-activated protein (MAP) kinase-dependent phosp horylation of synapsin-I, leading to an increase in the availability of synaptic vesicles for release (Jovanovic et al., 2000 ). Previous studies have shown that the neurotrophin-Trk complex can be rapidly int ernalized and transported retrogradely or anterogradely over long distances with in the neuron and across the synapse, providing a mechanism for long-range signa ling in neural circuits (von Bartheld et al., 1996a , 1996b, 2001 ; Bhattacharyy a et al., 1997 ; Grimes et al., 1997 ) . This long-range signaling is critical f or promoting neuronal survival and maintaining global neuronal properties of the presynaptic neuron (Lewin and Barde, 1996 ; Huang and Reichardt, 2001 ; Sofroni ew et al., 2001 ). Using beads covalently coated with NGF, MacInnis and Campenot (2002 ) found that the survival signal initiated by NGF binding to TrkA at the axon terminals was transported retrogradely to the soma without the internalizat ion of NGF, suggesting that long-range TrkA signaling is mediated by internalize d TrkA itself or downstream effectors, such as activated phosphatidylinositol 3- kinase (PI-3 kinase). In contrast, we have shown that BDNF, in the absence of in ternalization, can exert a localized action on synaptic transmission, presumably by a spatially restricted spread of downstream effector(s) of TrkB. Moreover, B DNF-induced synaptic potentiation does not require signaling to or from the neur onal soma (Stoop and Poo, 1995 ). In agreement with the latter finding, we found that the contact of single BDNF beads remained effective in inducing the synapt ic potentiation even after the axon was severed from the soma (Figure 6). These findings underscore the notion of two distinct modes of neurotrophin action: a l ocal action resulting from spatially restricted spread of cytoplamic downstream effectors triggered by Trk activation, and a global action that involves long-ra nge trafficking of internalized neurotrophin-Trk complexes, Trk alone, or downst ream effectors of Trk signaling pathway. Although BDNF and NT-3 were found to exert similar potentiating effects at sever al synapses (Lohof et al., 1993 ; Kang and Schuman, 1995 ), signal transduction mechanisms mediated by their receptors, TrkB and TrkC, respectively, differ sign ificantly (Segal and Greenberg, 1996 ; Patapoutian and Reichardt, 2001 ). For ex ample, BDNF/TrkB-induced synaptic potentiation depends on extracellular Ca2+ and is regulated by the level of cytosolic cAMP, whereas synaptic potentiation by N T-3/TrkC depends on Ca2+ release from internal stores and is independent of Ca2+ influx and the cytosolic cAMP level (Boulanger and Poo, 1999 , He et al., 2000 ; Yang et al., 2001 ). A similar difference in signal transduction was also foun d for the growth cone turning induced by gradients of BDNF and NT-3 (Song et al. , 1998 ; Song and Poo, 1999 ). Furthermore, the NT-3-induced synaptic potentiati on effect is independent of MAP kinase activation, but requires activation of bo th PI-3 kinase and phospholipase C- (Yang et al., 2001 ), whereas activation of MAP kinase is required for synaptic modulation by BDNF (Gottschalk et al., 1999 ; Jovanovic et al., 2000 ; Ying et al., 2002 ). Interestingly, unlike the locali zed synaptic effect of BDNF reported here, synaptic potentiation of NT-3 was fou nd to spread over long distances (300–400 μm) along the presynaptic axon from the site of local perfusion of NT-3 (Chang and Popov, 1999 ). Taken together, the differences in signal transduction between BDNF and NT-3 are likely to account for their differential dependence of synaptic potentiation on axonal protein synthesis observed in the present study and for the different spatial ranges of synaptic action of these two neurotrophins. Our results on the effects of translation inhibition on BDNF bead actions clearl y suggest that local constitutive protein synthesis within the developing axon i s required for the synaptic potentiation induced by BDNF beads (Figures 6 and 7) . A previous study has shown that treatment of hippocampus slices with protein translation inhibitors markedly attenuated the BDNF-induced potentiation on syna ptic transmission at the Schaffer collateral synapses on CA1 pyramidal cells, ev en for slices in which the synaptic neuropil was isolated from both pre- and pos tsynaptic cell bodies, suggesting that this synaptic action of BDNF requires loc al protein synthesis in either axonal or dendritic compartments, or both (Kang a nd Schuman, 1996 ). Recent evidence has further indicated that BDNF can initiate local protein synthesis in the dendrites of cultured hippocampal neurons (Aakal u et al., 2001 ). In the present study, we found that BDNF beads were effective in potentiating the synapse in the presence of translation inhibitor anisomycin, unless anisomycin treatment starts 2 hr before the bead contact. Assuming 45 mi n incubation with anisomycin is effective in blocking protein synthesis, BDNF-in itiated protein synthesis is not required for synaptic potentiation. Instead, co ntinuous synthesis of a critical presynaptic component seems to be required for the potentiation effect. Since prolonged inhibition of protein synthesis also ab olished BDNF-induced Ca2+ signaling, this critical component appears to act at a step upstream from Ca2+ signaling. It has been shown recently that the protein synthesis inhibitor anisomycin can reversibly block synapse formation and functi onal maturation of Alysia sensorimotor synapses in the absence of cell bodies (S chacher and Wu, 2002 ), suggesting that synaptic proteins may be rapidly synthes ized and degraded at or near synaptic sites. Our present finding on the effect o f 2 hr treatment of anisomycin is consistent with the requirement of constitutiv e protein synthesis in BDNF-induced synaptic potentiation. Constitutive local pr otein synthesis in the developing axon was also found to be critical for the ada ptation of growth cone turning responses (Ming et al., 2002 ), a process essenti al for long-range growth cone chemotaxis in a gradient of diffusible guidance cu es. Local protein synthesis thus plays a role not only in axonal guidance, but a lso in neurotrophin-dependent modulation of developing synapses. Experimental Procedures Summary Introduction Results Discussion Experimental Procedures References Cell Culture Preparation Xenopus nerve-muscle cultures were prepared as described previously (Tabti et al ., 1998 ). Briefly, the neural tube and associated myotomal tissue of 1-day-old Xenopus embryos (stage 20–22) were dissociated in Ca2+/Mg2+-free saline containing 115 mM NaCl, 2.6 mM KCl, 0.4 mM EDTA, and 10 mM HEPES (pH 7.6) for 20?C25 min. The cells were plated on clean glass coverslips and were cultured for 2 2–24 hr at room temperature (20°C–22°C) before use. The culture medium consi sted of (v/v) 50% Leibovitz's medium (L-15, GIBCO Life Technologies), 1% fetal b ovine serum (Hyclone, Longan, UT), and 49% Ringer's solution containing 115 mM N aCl, 2.5 mM KCl, 2 mM CaCl2, and 10 mM HEPES (pH 7.6). Preparation of Coated Beads The carboxyl-EDAC (1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride) method was used to covalently couple the protein—recombinant human BDNF (MW 22 kDa, Promega, Madison, WI), purified murine NGF (MW 26 kDa, Promega), or recomb inant human NT-3 (MW 27.2 kDa, PeproTech, Rocky Hill, NJ) or BSA (Sigma)—to pol ystyrene carboxylated beads (9.8 ± 0.8 μm in diameter, Ploysciences, Marr ington, PA). Carboxylated beads in 0.1 ml solution (2.5% w/v, ~4.55 × 106 beads ) were rinsed three times with 0.1 M carbonate buffer (pH 9.6) and 0.02 M phosph ate buffer (pH 4.5) sequentially. The beads were then treated with EDAC (1% w/v) freshly prepared in the phosphate buffer for 3 hr using an end-to-end rotary sh aker. After being washed three times with 0.2 M borate buffer (pH 8.5), the bead s were resuspended in the borate buffer containing the protein to be coupled (at the final concentration of 16.7 μg/ml) and allowed to mix on a rotary sha ker overnight (~12–14 hr) at 4°C. The beads were pelleted, while the supernatant was kept for determining the amount of proteins that had not bound to the beads, by spectrophotometric measurement (Bio-Rad, Hercules, CA). Methanolamine (0.1 M) and BSA (1% w/v) were used sequentially to treat the beads for 30 min, serving to block unreacted sites. Finally, the beads coated with the protein were stored at 4?鉉 in 0.02 M phosphate buffer containing (in w/v) 0.88% NaCl, 1% BSA, 0.1% Na2N3, and 5% glycerol (pH 7.4). Based on the number of beads per ml and the amount of proteins bound, BDNF bound to each bead was estimated to be 0.85 ??0.05 pg (SD, n = 5). Assuming BDNF binds uniformly onto the bead surface, the surface density of BDNF was ~2.7 fg/μm2. The BDNF- or NT-3-coated beads remain biologically active in the storage buffer for up to one month. The reaction used to couple the protein to the carboxylated beads is expected to have linked -NH2 groups of the proteins with -COOH groups on the bead surface so that the BDNF molecule may have adopted various orientations on the beads. A previous study has shown that neurotrophin molecules coupled to the beads by the present method do not detach within 3 hr after exposure to the saline solution (Gallo et al., 1997 ). BDNF on the coated bead was confirmed by immunostaining with a rabbit polyclonal IgG antibody against BDNF (Santa Cruz Biotechnology, Santa Cruz, CA) and a rhodamine-labeled secondary antibody. A monoclonal IgG antibody against BSA (Sigma) was used to detect BSA on the BSA-coated beads. Electrophysiology Synaptic currents were recorded from innervated myocytes by the whole-cell patch recording method, using a patch clamp amplifier (Axonpatch 200B, Axon Instrumen ts, Foster City, CA). Myocytes were voltage clamped at -70mV. Recordings were pe rformed at room temperature in a bath solution containing 140 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1 mM MgCl2, and 10 mM HEPES (pH 7.4). The solution inside the recor ding pipette (1–2 M) contained 145 mM KCl, 1 mM NaCl, 1 mM MgCl2, 1 mM adenosine 5'-triphosphate, and 10 mM HEPES (pH 7.2). The perforated whole-cell recording was used to stimulate the spinal neuron by a step depolarization of membrane potential from the clamped voltage of -60mV to 0mV (duration ~1.5?C2 ms). The intrapipette solution contained 140 mM potassium gluconate, 1 mM NaC l, 1 mM MgCl2, 10 mM HEPES (pH 7.2), and amphotericin B (200 μg/ml, CalBio chem, San Diego, CA). The test stimulus at a low frequency (7–8 pulses at 0.033 Hz, spaced by 5 min intervals) was used to minimize test-induced synaptic depression at these developing neuromuscular synapses. The membrane currents recorded were filtered at 5 kHz, digitized, and acquired with a microcomputer for further analysis using the WCP program (kindly provided by J. Dempster, University of Starthclyde, UK). In a typical experiment, 10?C20 min of stable control recording was obtained before the bead contact. The be ad was manipulated into the contact with the axon with a heat-polished suction p ipette (tip opening of 3 μm). The distance from the bead contact site to t he synapse was defined by the axon trajectory between the centers of the bead an d the innervated myocyte, assuming synapses are located at the center of the myo cyte. In experiments studying protein synthesis, actinomycin D, anisomycin, or c ycloheximide (all from Calbiochem) was added in the culture medium for 45 min or 2 hr before the bead contact or bath application of BDNF and was continuously p resent in the recording solution. All three drugs were dissolved in dimethyl sul foxide (DMSO) and diluted by the bath solution to the desired concentration, wit h the final DMSO concentration of 0.1% (v/v). In experiments involving axon tran section, the drugs were applied to the culture medium 30 min after transection, allowing resealing of the injury site. Local puffing of BDNF was delivered through a micropipette (tip opening ~1–2 μm) placed at a distance of 15?C20 μm from the axon. The micropipette was filled with the bath solution c ontaining BDNF (5 μg/ml). Repetitive pressure injection (3 psi in amplitud e, 20 ms in duration, 2 Hz) was applied to the micropipette with an electrically gated valve (Picospritzer, General Valves, Fairfield, NJ). The concentration of BDNF at the exposed axon was estimated to have reached a steady state of ~1.3 & micro;g/ml within 15 s (Lohof et al., 1992 ), whereas it was less than 5 ng/ml a t the distant synaptic sites, a concentration insufficient to induce synaptic po tentiation in the present system (Boulanger and Poo, 1999 ). In experiments stud ying Ca2+ elevation elicited by focal application of BDNF, the micropipette fill ed with 1 μg/ml BDNF was placed at a distance of 5 μm from the axon, and the bath perfusion flow (~2 ml/min) was present continuously. Injections of Drugs and Fluorescent Dye Gelonin (30 μM, Sigma), together with 10 kDa dextran-fluorescein (350 &mic ro;M, Molecular Probes, Eugene, OR), was injected into presynatic spinal neurons or postsynaptic myocytes at the cell body with sharp microelectrodes (12–16 M), using an Eppendorf Pressure Injection system (a Transjector 5246 in conjunction with a Micromanipulator 5171, Hamburg, Germany). A constant backpressure of 0.4?C0.6 psi was applied to the electrode. A brief pressure pulse (0.2–0.3 s per cell and 0.1?C0.3 psi above the backpressure) was used for injection. Coinjection of dextran- fluorescein was used to mark the injected cells. Calcium-sensitive dye Oregon Gr een-488 BAPTA-1 (conjugated to 70 kDa dextran, 330 μM, Molecular Probes) w as injected into the presynaptic spinal neuron using the same technique. The dru g and dye were diluted to the desired concentrations in the injection buffer con taining 140 mM KCl, 1 mM NaCl, 0.1 mM EGTA, and 10 mM HEPES (pH 7.2). Ca2+ Imaging Xenopus spinal neurons were microinjected with Ca2+-sensitive dye, Oregon Green- 488 BAPTA-1 conjugated to 70 kDa dextran. Calcium imaging was done using a Leica confocal imaging system (TCS SP) equipped with an argon gas ion laser and a Lei ca inverted microscope (DM IRBE) fitted with a Leica 63× objective (HCX PL Apo; NA, 1.32). The Oregon Green BAPTA-1-dextran was excited at 488 nm, and its fluo rescence signal was collected at 500–540 nm through the emission filter. Images were acquired every 15 s at 512 ??512 pixels. To monitor laser power output, transmission images also were collected simultaneously. Digital images were analyzed using ImageJ (http://rsb.info.nih.gov/ij). To quantify changes in the Ca2+ level, the mean fluorescence intensity within the axon at a length of 60 μm across the bead contact sites was measured over the fixed polygon area that covered the axons throughout the measurement period. To measure the spatial profile of Ca2+ signals along the axon, the polygons at a length of 10 μm were used. Ratios of fluorescence intensity change (F = F - Fo) and basal fluorescence intensity (Fo) were used to represent the Ca2+ signals. Staining for Nucleic Acids and Ribosomal Proteins The SYTO 64 dyes and Hoechst 33342 (Molecular Probes) were used to stain nucleic acids. The culture was incubated with 0.5 μM SYTO 64 for 20 min at room t emperature. After an extensive wash, the culture was incubated with the solution containing 1 μg/ml Hoechst 3342 for another 10 min. The SYTO 64 fluoresce nce images were acquired by the confocal microscopy (Leica TCS SP), and the Hoec hst 33342 fluorescence images were captured by a CCD camera (PhotoMetrics, Hunti ngton Beach, CA). Cultures for immunocytochemistry were plated on the poly-L-lys ine-coated coverslips. The cultures were fixed with 4% paraformaldehyde (EM Scie nce, Gibbstown, NJ) for 15 min at room temperature and washed three times with p hosphate buffer saline (PBS). Fixed cells were incubated with a human autoantibo dy against ribosomal P protein 1:150 (ImmunoVision, Springdale, AR) at 4°C over night after being treated with PBS containing 0.25% Triton X-100, 5% normal goat serum, and 1% BSA for 20 min. The Alex488 goat anti-human IgG secondary antibod y (Molecular Probes) was used to visualize the specific immunostaining, and Alex 568-phallotoxin (Molecular Probes) was also used to label F-actin for visualizin g the fixed cells. The fixed cells then were incubated with Alex568 phallotoxin (13.2 μM, Molecular Probes) for 20 min to label the F-actin. The fluoresce nt images were captured by confocal microscopy (Leica TCS SP). Acknowledgments We thank S. Andersen, S. Wang, and J. Henley for helpful discussions, and S. And ersen for initial help in the bead preparation. This work was supported by a gra nt from the National Institutes of Health (NS 37831). X-h.Z. was supported by a long-term fellowship from Human Frontier Science Program. References Summary Introduction Results Discussion Experimental Procedures References Aakalu G., Smith W.B., Nguyen N., Jiang C. and Schuman E.M. (2001) Dynamic visua lization of local protein synthesis in hippocampal neurons. Neuron, 30:489-502. [Medline][Summary][Full Text] Aoki C., Wu K., Elste A., Len G., Lin S., McAuliffe G. and Black I.B. (2000) Loc alization of brain-derived neurotrophic factor and TrkB receptors to posynaptic densities of ***** rat cerebral cortex. J. Neurosci. Res., 59:454-463. [Medline] Bassell G.J., Zhang H., Byrd A.L., Femino A.M., Singer R.H., Taneja K.L., Lifshi tz L.M., Herman I.M. and Losik K.S. (1998) Sorting of -actin mRNA and protein to neurite and growth cones in culture. J. Neurosci., 18:251-265. [Medline] Beaumont V., Zhong N., Fletcher R., Froemke R.C. and Zucker R.S. (2001) Phosphor ylation and local presynaptic rotein synthesis in calcium- and calcineurin-depen dent induction of crayfish long-term facilitation. Neuron, 32:489-501. [Medline] [Summary][Full Text] Berninger B., Garcia D.E., Inagaki N., Hahnel D. and Lindholm D. (1993) BDNF and NT-3 induced intracellular Ca2+ elevation in hippocampal neurons. Neuroreport, 4:1303-1306. [Medline] Bhattacharyya A., Watson F.L., Bradlee T.A., Pomeroy S.L., Stiles C.D. and Segal R.A. (1997) Trk receptors function as rapid retrograde signal carriers in the a dult nervous system. J. Neurosci., 17:7007-7016. [Medline] Bl?chl A. and Thoenen H. (1995) Characterization of nerve growth factor (NG F) release from hippocampal neurons: evidence for a constitutive and an unconven tional sodium-dependent regulated pathway. Eur. J. Neurosci., 7:1220-1228. [Medl ine] Brittis P.A., Lu Q. and Flanagan J.G. (2002) Axonal protein synthesis provides a mechanism for localized regulation at an intermediate target. Cell, 110:223-235 . [Medline][Summary][Full Text] Boulanger L. and Poo M.M. (1999) Gating of BDNF-induced synaptic potentiation by cAMP. Science, 284:1982-1984. [Medline] Cabelli R.J., Hohn A. and Shatz C.J. (1995) Inhibition of ocular dominance colum n formation by infusion of NT-4/5 or BDNF. Science, 267:1662-1666. [Medline] Cabelli R.J., Allendoefer K.L., Radeke M.J., Welcher A.A., Feinstein S.C. and Sh atz C.J. (1996) Changing patterns of expression and subcellular localization of TrkB in the developing visual system. J. Neurosci., 16:7965-7980. [Medline] Cabelli R.J., Shelton D.J., Segal R.A. and Shatz C.J. (1997) Blockade of endogen ous ligands of TrkB inhibits formation of ocular dominance columns. Neuron, 19:6 3-67. [Medline][Summary][Full Text] Campbell D.S. and Holt C.E. (2001) Chemotropic responses of retinal growth cones mediated by rapid local protein synthesis and degradation. Neuron, 32:1013-1026 . [Medline][Summary][Full Text] Chang S. and Popov S.V. (1999) Long-range signaling within growing neurites medi ated by neurotrophin-3. Proc. Natl. Acad. Sci. USA, 96:4095-4100. [Medline] Cohen-Cory S. and Fraser S.E. (1995) Effects of brain-derived neurotrophic facto r on optic axon branching and remodelling in vivo Nature, 378:192-196. [Medline] Davis L., Dou P., DeWit M. and Kater S.B. (1992) Protein synthesis within neuron al growth cones. J. Neurosci., 12:4867-4877. [Medline] Diamond J., Holmes M. and Coughlin M. (1992) Endogenous NGF and nerve impulses r egulate the collateral sprouting of sensory axons in the skin of the ***** rat. J. Neurosci., 12:454-466. [Medline] Drake C.T., Milner T.A. and Patterson S.L. (1999) Ultrastructural localization o f full-length TrkB immunoreactivity in rat hippocampus suggests multiple roles i n modulating activity-dependent synaptic plasticity. J. Neurosci., 19:8009-8026. [Medline] Eng H., Lund K. and Campenot R.B. (1999) Synthesis of beta-tubulin, actin and ot her proteins in axons of sympathetic neurons in compartmentalized culture. J. Ne urosci., 19:1-9. [Medline] Engert F. and Bonhoeffer T. (1997) Synapse specificity of long-term potentiation breaks down at short distances. Nature, 388:279-284. [Medline] Engert F. and Bonhoeffer T. (1999) Dendritic spine changes associated with hippo campus long-term synaptic plasticity. Nature, 399:66-70. [Medline] Evers J., Laser M., Sun Y.A., Xie Z.P. and Poo M.M. (1989) Studies of nerve-musc le interactions in Xenopus cell culture: analysis of early synaptic currents. J. Neurosci., 9:1523-1539. [Medline] Figurov A., Pozzo-Miller L., Olafsson P., Wang T. and Lu B. (1996) Regulation of synaptic responses to high frequency stimulation and LTP by neurotrophins in th e hippocampus. Nature, 381:706-709. [Medline] Funakoshi H., Belluardo N., Arenas E., Yamamoto Y., Casabona A., Persson H. and Ibanez C.F. (1995) Muscle-derived neurotrophin-4 as an activity-dependent trophi c signal for ***** motor neurons. Science, 268:1495-1499. [Medline] Gallo G., Lefcort F.B. and Letourneau P.C. (1997) The TrkA receptor mediates gro wth cone turning toward a localized source of nerve growth factor. J. Neurosci., 17:5445-5454. [Medline] Goodman L.J., Valverde J., Lim F., Geschwind M.D., Federoff H.J., Geller A.I. an d Hefti F. (1996) Regulated release and polarized localization of brain-derived neurotrophic factor in hippocampal neurons. Mol. Cell. Neurosci., 7:222-238. [Me dline] Gottschalk W.A., Jiang H., Tartaglia N., Feng L., Figurov A. and Lu B. (1999) Si gnaling mechanisms mediating BDNF modulation of synaptic plasticity in the hippo campus. Learn. Mem., 6:243-256. Grimes M.L., Beattie E. and Mobley W.C. (1997) A signaling organelle containing the nerve growth factor-activated receptor tyrosine kinase, TrkA. Proc. Natl. Ac ad. Sci. USA, 94:9909-9914. [Medline] Hartmann M., Heumann R. and Lessmann V. (2001) Synaptic secretion of BDNF after high-frequency stimulation of glutamatergic synapses. EMBO J., 20:5887-5897. [Me dline] He X., Yang F., Xie Z. and Lu B. (2000) Intracellular Ca2+ and Ca2+/calmodulin-d ependent kinase II mediate acute potentiation of neurotransmitter release by neu rotrophin-3. J. Cell Biol., 149:783-792. [Medline] Huang E.J. and Reichardt L.F. (2001) Neurotrophins: roles in neuronal developmen t and function. Annu. Rev. Neurosci., 24:677-736. [Medline] Jovanovic J.N., Czernik A.J., Fienberg A.A., Greengard P. and Sihra T.S. (2000) Synapsins as mediators of BDNF-enhanced neurotransmitter release. Nat. Neurosci. , 3:323-329. Kang H. and Schuman E.M. (1995) Long-lasting neurotrophin-induced enhancement of synaptic transmission in the ***** hippocampus. Science, 267:1658-1662. [Medlin e] Kang H. and Schuman E.M. (1996) A requirement for local protein synthesis in neu rotrophin-induced hippocampal synaptic plasticity. Science, 273:1402-1406. [Medl ine] Kleiman R.J., Tian N., Krizaj D., Hwang T.N., Copenhagen D.R. and Reichardt L.F. (2000) BDNF-induced potentiation of spontaneous twitching in innervated myocyte s requires calcium release from intracellular stores. J. Neurophysiol., 84:472-4 83. [Medline] Koening E. and Giuditta A. (1999) Protein-synthesizing machinery in the axon com partment. Neuroscience, 89:5-15. [Medline] Koening E., Martin R., Titmus M. and Sotelo-Silveria J.R. (2000) Cryptic periphe ral ribosomal domains distributed intermittently along mammalian myelinated axon s. J. Neurosci., 20:8390-8400. [Medline] Kojima M., Takei N., Numakawa T., Ishikawa Y., Suzuki S., Matsumoto T., Katoh-Se mba R., Nawa H. and Hatanaka H. (2001) Biological characterization and optical i maging of brain-derived neurotrophic factor-green fluorescent protein suggest an activity-dependent local release of brain-derived neurotrophic factor in neurit es of cultured hippocampal neurons. J. Neurosci. Res., 64:1-10. [Medline] Korte M., Carroll P., Wolf E., Brem G., Thoenen H. and Bonhoeffer T. (1995) Hipp ocampal long-term potentiation is impaired in mice lacking brain-derived neurotr ophic factor. Proc. Natl. Acad. Sci. USA, 92:8856-8860. [Medline] Kryl D., Yacoubian T., Haapasalo T., Castren E., Lo D. and Barker P.A. (1999) Su bcellular localization of full-length and truncated Trk receptor isoforms in pol arized neurons and epithelial cells. J. Neurosci., 19:5823-5833. [Medline] Lewin G.R. and Barde Y.A. (1996) Physiology of neurotrophins. Annu. Rev. Neurosc i., 19:289-317. [Medline] Li Y.X., Zhang Y., Lester H.A., Schuman E.M. and Davidson N. (1998) Enhancement of neurotransmitter release induced by brain-derived neurotrophic factor in cult ured hippocampal neurons. J. Neurosci., 18:10231-10240. [Medline] Lohof A.M., Quillan M., Dan Y. and Poo M.M. (1992) Asymmetric modulation of cyto solic cAMP activity induces growth cone turning. J. Neurosci., 12:1253-1261. [Me dline] Lohof A.M., Ip N.Y. and Poo M.M. (1993) Potentiation of developing neuromuscular synapses by the neurotrophins NT-3 and BDNF. Nature, 363:350-353. [Medline] MacInnis B.L. and Campenot R.B. (2002) Retrograde support of neuronal survival w ithout retrograde transport of nerve growth factor. Science, 295:1536-1539. [Med line] Maffei L., Berardi N., Domenici L., Pairsi V. and Pizzorusso T. (1992) Nerve gro wth factor (NGF) prevents the shift in ocular dominance distribution of visual c ortical neurons in monocularly deprived rats. J. Neurosci., 12:4651-4662. [Medli ne] Malenka R.C. and Nicoll R.A. (1999) Long-term potentiation–a decade of progress? Science, 285:1870-1874. [Medline] Maletic-Savatic M., Malinow R. and Svoboda K. (1999) Rapid dendritic morphogenes is in CA1 hippocampal dendrites induced by synaptic activity. Science, 283:1923- 1927. [Medline] Martin K.C., Casadio A., Zhu H., E Y., Rose J.C., Chen M., Bailey C.H. and Kande l E.R. (1997) Synapse-specific, long-term facilitation of aplysia sensory to mot or synapses: a function for local protein synthesis in memory storage. Cell, 91: 927-938. [Medline][Summary][Full Text] Martin K.C., Barad M. and Kandel E.R. (2000) Local protein synthesis and its rol e in synapse-specific plasticity. Curr. Opin. Neurobiol., 10:587-592. [Medline] McAllister A.K., Lo D.C. and Katz L.C. (1995) Neurotrophins regulate dendritic g rowth in developing visual cortex. Neuron, 15:791-803. [Medline][Summary] McAllister A.K., Katz L.C. and Lo D.C. (1996) Neurotrophin regulation of cortica l dendritic growth requires activity. Neuron, 17:1057-1064. [Medline][Summary][F ull Text] McAllister A.K., Katz L.C. and Lo L.C. (1999) Neurotrophins and synaptic plastic ity. Annu. Rev. Neurosci., 22:295-318. [Medline] Miledi R. (1973) Transmitter release induced by injection of calcium ions into n erve terminals. Proc. R. Soc. Lond. B Biol., 183:421-425. Ming G.L., Wong S.T., Henley J., Yuan S.B., Song H.J., Spitzer N.C. and Poo M.M. (2002) Adaptation in the chemotactic guidance of nerve growth cones. Nature, 41 7:411-418. [Medline] Olink-Coux M. and Hollenbeck P.J. (1996) Localization and active transport of mR NA in axons of sympathetic neurons in culture. J. Neurosci., 16:1346-1358. [Medl ine] Patapoutian A. and Reichardt L. (2001) Trk receptors: mediators of neurotrophin action. Curr. Opin. Neurobiol., 11:272-280. [Medline] Patterson S.L., Grover L.M., Schwartzkroin P.A. and Bothwell M. (1992) Neurotrop hin expression in rat hippocampal slices: a stimulus paradigm inducing LTP in CA 1 evokes increases in BDNF and NT-3 mRNAs. Neuron, 9:1081-1088. [Medline][Summar y] Poo M.M. (2001) Neurotrophins as synaptic modulators. Nat. Rev. Neurosci., 2:24- 32. Pozzo-Miller L., Gottschalk W., Zhang L., McDermott K., Du J., Gopalakrishnan R. , Oho C., Sheng Z. and Lu B. (1999) Impairments in high-frequency transmission, synaptic vesicle docking, and synaptic protein distribution in the hippocampus o f BDNF knockout mice. J. Neurosci., 19:4972-4983. [Medline] Purves D. and Lichtman J.W. Principles of Neural Development. (1985) Sunderland: Sinauer Associates Schacher S. and Wu F. (2002) Synapse formation in the absence of cell bodies req uires protein synthesis. J. Neurosci., 22:1831-1839. [Medline] Schuman E.M. (1997) Synapse specificity and long-term information storage. Neuro n, 18:339-342. [Medline][Full Text] Segal R.A. and Greenberg M.E. (1996) Intracellular signaling pathways activated by neurotrophic factors. Annu. Rev. Neurosci., 19:463-490. [Medline] Sofroniew M.V., Howe C.L. and Mobley W.C. (2001) Nerve growth factor signaling, neuroprotection, and neural repair. Annu. Rev. Neurosci., 24:1217-1281. [Medline ] Song H.J. and Poo M.M. (1999) Signal transduction underlying growth cone guidanc e by diffusible factors. Curr. Opin. Neurobiol., 9:355-363. [Medline] Song H.J., Ming G.L., He Z., Lehmann M., McKerracher L., Tessier-Lavigen M. and Poo M.M. (1998) Coversion of neuronal growth cone responses from replusion to at traction by nucleotides. Science, 281:1515-1518. [Medline] Steward O. (1997) mRNA localization in neurons: a multipurpose mechanism? Neuron , 18:9-12. [Medline][Full Text] Steward O. and Levy W.B. (1982) Preferential localization of polyribosomes under the base of dendritic spines in granule cells of the dentate gyrus. J. Neurosci ., 2:284-291. [Medline] Steward O. and Reeves T.M. (1988) Protein synthetic machinery beneath postsynapt ic sites on CNS neurons: association between polyribosomes and other organelles at the synaptic site. J. Neurosci., 8:176-184. [Medline] Steward O. and Schuman E.M. (2001) Protein synthesis at synaptic sites on dendri tes. Annu. Rev. Neurosci., 24:299-325. [Medline] Stirpe F., Olsnes S. and Pihl A. (1980) Gelonin, a new inhibitor of protein synt hesis, nontoxic to intact cells. Isolation, characterization, and preparation of cytotoxic complexes with concanavalin A. J. Biol. Chem., 255:6947-6953. [Medlin e] Stoop R. and Poo M.M. (1995) Potentiation of transmitter release by ciliary neur otrophic factor requires somatic signaling. Science, 267:695-699. [Medline] Stoop R. and Poo M.M. (1996) Synaptic modulation by neurotrophic factors: differ ential and synergistic effects of brain-derived neurotrophic factor and ciliary neurotrophic factor. J. Neurosci., 16:3256-3264. [Medline] Tabti N., Alder J. and Poo M.M. (1998) Culturing spinal neurons and muscle cells from Xenopus embryos. In: Banker G. and Goslin K. (Eds.) Culturing Nerve Cells. Cambridge, MA: MIT Press Tongiorgi E., Righi M. and Cattaneo A. (1997) Activity-dependent dendritic targe ting of BDNF and TrkB mRNAs in hippocampal neurons. J. Neurosci., 17:9492-9505. [Medline] von Bartheld C.S., Byers M.R., Williams R. and Bothwell M. (1996) Anterograde tr ansport of neurotrophins and axodendritic transfer in the developing visual syst em. Nature, 379:830-833. [Medline] von Bartheld C.S., Williams R., Lefcort F., Clary D.O., Reichardt L.F. and Bothw ell M. (1996) Retrograde transport of neurotrophins from the eye to the brain in chick embryos: roles of the p75NTR and TrkB receptors. J. Neurosci., 16:2995-30 08. [Medline] von Bartheld C.S., Wang X. and Butowt R. (2001) Anterograde axonal transport, tr anscytosis, and recycling of neurotrophic factors: the concept of trophic curren cies in neural networks. Mol. Neurobiol., 24:1-28. [Medline] Wang X.H. and Poo M.M. (1997) Potentiation of developing synapses by postsynapti c release of neurotrophin-4. Neuron, 19:825-835. [Medline][Summary][Full Text] Wang X.H., Berninger B. and Poo M.M. (1998) Localized synaptic actions of neurot rophin-4. J. Neurosci., 18:4985-4992. [Medline] Yang F., He X.P., Feng L.Y., Mizuno K., Liu X.W., Russell J., Xiong W.C. and Lu B. (2001) PI-3 kinase and IP3 are both necessary and sufficient to mediate NT-3- induced synaptic protentiation. Nat. Neurosci., 4:19-28. Ying S.W., Futter M., Rosenblum K., Webber M.J., Hunt S.P., Bliss T.V. and Bramh am C.R. (2002) Brain-derived neurotrophic factor induces long-term potentiation in intact ***** hippocampus: requirement for ERK activation coupled to CREB and upregulation of Arc synthesis. J. Neurosci., 22:1532-1540. [Medline] Yuste R. and Bonhoeffer T. (2001) Morphological changes in dendritic spines asso ciated with long-term synaptic plasticity. Annu. Rev. Neurosci., 24:1073-1089. [ Medline] Zafra F., Hengerer B., Leibrock J., Thoenen H. and Lindholm D. (1990) Activity d ependent regulation of BDNF and NGF mRNAs in the rat hippocampus is mediated by non-NMDA glutamate receptors. EMBO J., 9:3545-3550. [Medline] Zheng J.Q., Kelly T.K., C