JBC DNA damage antibodies

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M609672200 on May 31, 2007

J. Biol. Chem., Vol. 282, Issue 35, 25517-25526, August 31, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/35/25517    most recent
M609672200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sarkisov, D. V.
Right arrow Articles by Wang, S. S.-H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sarkisov, D. V.
Right arrow Articles by Wang, S. S.-H.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Synapse Specificity of Calcium Release Probed by Chemical Two-photon Uncaging of Inositol 1,4,5-Trisphosphate*

Dmitry V. Sarkisov{ddagger}§, Shari E. Gelber, Jeffery W. Walker||, and Samuel S.-H. Wang§**1

From the Departments of {ddagger}Physics and **Molecular Biology and §Program in Neuroscience, Princeton University, Princeton, New Jersey 08544, Division of Maternal and Fetal Medicine, Cornell Weill Medical College, New York, New York 10021, and ||Department of Physiology, University of Wisconsin, Madison, Wisconsin 53706

Received for publication, October 13, 2006 , and in revised form, May 8, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Biological messengers can be "caged" by adding a single photosensitive group that can be photolyzed by a light flash to achieve spatially and temporally precise biochemical control. Here we report that photolysis of a double-caged form of the second messenger inositol 1,4,5-trisphosphate (IP3) triggers focal calcium release in Purkinje cell somata, dendrites, and spines as measured by two-photon microscopy. In calbindin knock-out Purkinje cells, peak calcium increased with flash energy with higher cooperativity for double-caged IP3 than for conventional single-caged IP3, consistent with a chemical two-photon effect. Spine photolysis of double-caged IP3 led to local calcium release. Uncaging of glycerophosphoryl-myo-inositol 4,5-bisphosphate (gPIP2), a poorly metabolizable IP3 analog, led to less well localized release. Thus, IP3 breakdown is necessary for spine-specificity. IP3- and gPIP2-evoked signals declined from peak with similar, slow time courses, indicating that release lasts hundreds of milliseconds and is terminated not by IP3 degradation but by intrinsic receptor dynamics. Based on measurements of spine-dendrite coupling, IP3-evoked calcium signals are expected to be at least 2.4-fold larger in their spine of origin than in nearby spines, allowing IP3 to act as a synapse-specific second messenger. Unexpectedly, single-caged IP3 led to less release in somata and was ineffective in dendrites and spines. Calcium release using caged gPIP2 was inhibited by the addition of single-caged IP3, suggesting that single-caged IP3 is an antagonist of calcium release. Caging at multiple sites may be an effective general approach to reducing residual receptor interaction.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Inositol 1,4,5-trisphosphate binds to receptors on internal organelles, triggering calcium (Ca2+) release from stores and thus elevating cytoplasmic calcium concentration. Calcium elevation can trigger a wide variety of physiological responses in many cell types (1). In neurons, inositol 1,4,5-trisphosphate (IP3)-triggered calcium release can be extremely rapid, beginning considerably less than 100 ms after synaptic activity (2–4). Such a rapid response implies that IP32 is produced within 10s of ms of receptor activation. Furthermore, calcium release mechanisms nearly always have strong inactivation mechanisms that can depend on recent activation. Therefore, the rapid kinetics of IP3 concentration and receptor dynamics are key determinants of the properties of calcium release.

The properties of IP3-dependent calcium release may guide the properties of synaptic plasticity. In cerebellar Purkinje cells, which strongly express IP3 receptors, IP3 receptor-mediated calcium release is necessary for the induction of long-term depression of parallel fiber synapses (5, 6). Calcium release in dendritic spines is thought to act as a coincidence detector between parallel fiber activity, which generates IP3, and climbing fiber activity, which causes calcium entry that boosts release by acting on the IP3 receptor (4, 7). In this model the duration of IP3 receptor activation should shape the timing conditions under which parallel fiber and climbing fiber activity are able to work together to generate maximal calcium release. Probing IP3 receptor dynamics in spines in situ would provide strong constraints to the model.

A critical advance in understanding calcium signaling responses to fast patterns of IP3 production has come as a result of the development of caged compounds (8–10). Caged compounds are signaling molecules rendered biologically inactive by the addition of a light-sensitive group. Upon absorption of a photon in the near ultraviolet, the caged compound undergoes internal photolysis, cleaving the cage group and generating active molecules. When the light comes in the form of a flash, the production of messenger is rapid, and with a focused beam such as from a laser, submicron spatial resolution is possible.

In the case of IP3, the standard caging approach has been to add a 1-(2-nitrophenyl)ethyl (NPE) group at the 4 or 5 position (11). "Single-caged" IP3 produced in this way does not activate IP3 receptors, photolyses in milliseconds, and largely does not interfere with IP3 metabolism by kinases and phosphatases (though in the case of 5-NPE-IP3 see (12)). Single-caged IP3 has been successfully applied to the study of signaling in smooth muscle (13), cell nuclei (14, 15), neurons (3, 5, 6, 16, 17), and astrocytes (18).

As a means of controlling calcium release in neurons, we made double-caged IP3, a form that is caged at both the 4 and 5 positions. This compound is structurally less similar than single-caged IP3 to IP3 and, therefore, should be less likely to interfere with IP3 receptors or degradative enzymes. In addition, double-caged compounds allow an approach, termed chemical two-photon uncaging, that has several advantages. First, the requirement for removing two cage groups gives positive cooperativity, and therefore, in response to focused flashes, spatial resolution in both lateral and axial directions is improved (19). Second, because each molecule bears two cage groups, small amounts of uncaging that occur during routine handling or after solvation in water produce lower levels of residual IP3.

In this study we report the first use of double-caged IP3 in the study of intracellular calcium release. We find that in cerebellar Purkinje neurons, double-caged IP3 is more potent than single-caged IP3 in triggering flash-induced calcium release. In addition, double-caged IP3 can activate release in dendritic locations remote from the cell body, but single-caged IP3 cannot. These phenomena can be explained by a blocking action by single-caged IP3 on calcium release. The improved effectiveness of double-caged IP3 makes it the preferred reagent for rapid control over IP3-dependent calcium release in neurons. Finally, we use double-caged IP3 to probe the synapse specificity of calcium release in single dendritic spines.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Preparation of Caged IP3 Compounds—Synthesis of caged compounds was done under incandescent light. Five mg of IP3 (Mr 551.99, LC Laboratories, Woburn, MA) was dissolved in 500 µl of H2O, and the pH was adjusted to 4.5 with HCl. 1-(Nitrophenyl)-diazoethane was prepared by mixing 0.18 g of the hydrazone of 2-nitrosoacetophenone with 0.7 g of MnO2 in 20 ml of ether, mixing vigorously for 5–10 min, and filtering the mixture. IP3 was pipetted into the 1-(2-nitrophenyl)-diazoethane. The volume was reduced by blowing compressed nitrogen over the mixture, which was then covered with glass wool and mixed vigorously overnight. Unreacted 1-(2-nitrophenyl)-diazoethane was extracted with methylene chloride.

Caged compounds were purified by high performance liquid chromatography on a Partisil 10 SAX column (initial condition 0.12 M NH4H2PO4/NH4HPO4, pH 5.4, 30% methanol) (12). The sample was injected, and after 5 min, a 15-min linear gradient was begun to a final condition of 0.6 M NH4H2PO4/NH4HPO4, pH 5.4, 30% methanol. The column was monitored by absorption at 260 nm. The peak at 5–7 min was taken as triple-NPE-caged IP3.

Triple-caged IP3 was photolyzed 6 x 30 s with a handheld ultraviolet lamp. Fractions were collected with the following migration times: 5–7 min, triple-caged IP3, 1,4-double-caged IP3, and 1,5-double-caged IP3; 8–10 min, 4,5-double-caged IP3; 14–15 min, 5-caged IP3; 15–16 min, 4-caged IP3; 16–17 min, 1-caged IP3. Fractions from multiple runs were pooled, and the high performance liquid chromatography solvent was then removed using a DEAE-cellulose column at 4 °C. Caged IP3 compounds were eluted with a linear gradient of 0–0.5 M tetraethylammonium bicarbonate (TEAB) more than 100 min. TEAB was removed by rotary evaporation under vacuum. The compound was resuspended in distilled water and stored at -20 °C.

Slice Preparation—Sagittal 300-µm-thick cerebellar brain slices were cut from 17–21-day-old rats or 2–3-month-old calbindin knock-out mice (strain B6.129-Calb1tm1Mpin/J; The Jackson Laboratory, Bar Harbor, ME) in ice-cold artificial cerebrospinal fluid containing 126 mM NaCl, 3 mM KCl, 1 mM NaH2PO4, 20 mM D-glucose, 25 mM NaHCO3, 2 mM CaCl2, and 1 mM MgCl2 and saturated with 95% O2, 5% CO2. Slices were preincubated at 34 °C for 40–60 min and then kept at room temperature throughout the experiment.

Electrophysiology—For recording, slices were transferred to an immersion-type recording chamber perfused at 2–4 ml/min with artificial cerebrospinal fluid solution saturated with 95% O2, 5% CO2. Purkinje cells were visually patched with recording electrodes pulled from 1-mm borosilicate glass to a resistance of 4–7 megaohms. Electrophysiological signals were acquired with an Axopatch 200B amplifier and Clampex 8.0 software (Axon Instruments, Foster City, CA). The patch electrode was filled with a patch solution containing (pH was adjusted to 7.3 with KOH) 133 mM methanesulfonic acid, 7.4 mM KCl, 0.3 mM MgCl2, 3 mM Na2ATP, and 0.3 mM Na3GTP along with 300 µM calcium indicator fluo-5F (Invitrogen) and one of the following caged compounds: 5-caged IP3, a mixture of 4- and 5-caged IP3, 4,5-double-caged IP3, caged glycerophosphoryl-myo-inositol 4,5-bisphosphate (gPIP2), or commercially obtained single-caged IP3 (Invitrogen), which is a mixture of 4 and 5 isomers. In some experiments aurintricarboxylic acid, ammonium salt (ATA; Sigma) was used to block IP3 breakdown. After whole-cell break-in, cells were held in current clamp mode for at least 40 min to allow diffusion of the dye and caged compound to the distal dendrites and spines. Holding currents at -65 mV were -50 to -400 pA, and series resistances were 20–30 megaohms. Series resistance was monitored periodically and compensated by balancing the bridge.

Two-photon Microscopy and Focal Uncaging—Two-photon fluorescence imaging was done using a custom-built microscope controlled by CfNT software (M. Müller, Max Planck Institute for Medical Research, Heidelberg, Germany). 830-nm excitation light from a Mira 900 Ti:sapphire laser (Coherent Inc., Auburn, CA) was focused onto the brain slice by 63x 0.9 NA water immersion objective (Carl Zeiss, Thornwood, NY). For focal uncaging (20) the output of a UV laser (DPSS Lasers, Santa Clara, CA) was attenuated by a polarizing beam splitter, widened by a 5x beam expander (CVI Laser Corp., Albuquerque, NM) to fill the back aperture of the objective, and merged with the excitation beam using a dichroic mirror (Omega Optical, Brattleboro, VT). To compensate for the focal plane shift between uncaging and excitation light, the uncaging beam was slightly converged by adjusting the beam expander. The parfocality of the uncaging spot and the image plane was tested by uncaging in a sample of caged fluorescein dextran solution dried onto a test slide. Flash energies were calculated from energies arriving at the backplane of the objective multiplied by the 70% transmittance of the objective. For focal uncaging in different cell regions, a procedure was usually employed in which uncaging was attempted first in the soma or main dendrites. If measurable responses were observed in these structures, then uncaging attempts were made in fine spiny branchlets and single spines. In some double-caged IP3 experiments, uncaging was done only in spiny branchlets and spines.


Figure 1
View larger version (33K):
[in this window]
[in a new window]

  		FIGURE 1.
Synthesis of double-caged IP3. a, structures of single-caged IP3, caged gPIP2, and double-caged IP3. Brackets indicate alternate isomers. b, diagram of focal uncaging setup diagram indicating the two-photon excitation (red) and emission (green) light paths and the uncaging light path (purple). c, fine, spiny dendrites of Purkinje neurons. d, uncaging spot at the same resolution as panel c. Scale bars, 1 µm.

 
Data Analysis—Full-frame images were acquired at a speed of 5 ms per line (512 pixels/line) or 2 ms per line (128 pixels/line). Fluorescence data were processed using Igor Pro software (Wavemetrics, Lake Oswego, OR). {Delta}F/F0 traces were filtered by convolving with a hanning window of width 5 samples (Figs. 2 and 5) or 11 samples (Fig. 4). Peak fluorescence was defined as the maximum value after 11-sample filtering. Response area was integrated over 200 ms after the UV flash. The initial response slope was defined as the slope of the linear fit between 15 and 85% of peak response. The response latency was defined as ending when the unfiltered trace went more than 1.5 S.D. above preflash baseline for three consecutive samples, taking the middle of the three samples. For purposes of scoring uncaging trials, a successful trial was defined as one that triggered a peak fluorescence change of 50%{Delta}F/F0 or greater. For measuring the dependence of response size on flash energy (Fig. 2), curves for individual locations were normalized by flash energy to minimize the uncertainty of a power fit to pooled data. Except where otherwise indicated, all statistical comparisons are one-tailed tests.

Calculation of IP3 Photolysis—The efficiency of IP3 generation was calculated using the inverse relationship between latency and IP3 concentration (21). The availability of multiple measurements using single-caged IP3 was used to calculate the uncertainty in overall fits and in IP3 concentration. With 100 µM single-caged IP3 in the pipette and uncaging in the soma, calcium response latencies were measured in response to varying pulse energies. IP3 concentration would be expected to increase with pulse energy P as (IP3) = C(1 - e-P/g), where C is the concentration of single-caged IP3, and g is the energy to photolyze 1 - 1/e of the cage groups. g was varied to determine a best fit for [IP3] to single-caged IP3 latency data. The resulting uncaging efficiency factor was g = 0.59 ± 0.07 µJ over a range of pulse energies of 0.3–1.0 µJ. This calibration was then applied to double-caged IP3 experiments to calculate the IP3 concentration as [IP3] = C(1 - e-P/g)2.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
The caged compounds used in this study are shown in Fig. 1a, in which modifications to the phosphate groups at positions 1, 4, and 5 are indicated as R1, R2, and R3, respectively. The 4,5-biphosphate motif has been found to be necessary for IP3 receptor activation (12, 22), and modification at the 1-position has been shown to moderately affect the potency of receptor activation and strongly inhibit metabolism by 3-kinases and 5-phosphatases (23). IP3 caged with NPE groups at either the 4 or 5 position (single-caged IP3) was synthesized or obtained commercially as a mixture of the two isomers. NPE-caged glycerophosphoryl-myo-inositol 4,5-bisphosphate (caged-gPIP2) (23) was obtained commercially. IP3 caged with NPE groups at both the both the 4 and 5 positions ("double-caged IP3") was synthesized by preparation of triple-caged IP3 followed by partial photolysis. Photolysis and measurement of calcium release was performed on a system (Fig. 1b) in which an ultraviolet laser beam was directed into the focal plane of a two-photon microscope and a brain slice chamber (Fig. 1c), thus achieving submicron resolution of both uncaging and fluorescence measurement (average half-maximal width of fluorescent spot after uncaging = 0.59 µm; Fig. 1d).

To test the ability of double-caged IP3 to trigger calcium release, we recorded from Purkinje neurons, which express IP3 receptors at extremely high density (24), using whole-cell patch recording. Patch electrodes contained 100 µM double-caged IP3 with 300 µM fluo-5F added to monitor changes in cytoplasmic calcium. Under these conditions all parts of the cell, including dendrites and spines, were visible (Fig. 1c), indicating perfusion with patch solution. In observations of the cell body by two-photon fluorescence microscopy (Fig. 2a), light pulses (0.13–1.0 µJ) generated reproducible sudden elevations in calcium that began 10–100 ms after the pulse (15 locations in 10 cells). Maximum responses were 600 ± 250 (mean ± S.D.) % {Delta}F/F0 above baseline fluorescence (n = 7, 1 cell body and 6 dendrites). Thus photolysis of double-caged IP3 can trigger robust calcium release signals in Purkinje neurons.

In previous studies using conventional single-caged IP3, the kinetics of photolysis-driven calcium release were shown to depend strongly on IP3 concentration (21). We found likewise that photolyzing increasing amounts of double-caged IP3 triggered progressively faster-rising and larger calcium transients (Fig. 2b), with half-maximal responses reached using flash energies of 0.36 ± 0.05 µJ (mean ± S.D., for half-maximal rate of rise of response), 0.27 ± 0.02 µJ (mean ± S.D., for half-maximal peak of response), and 0.24 ± 0.03 µJ (mean ± S.D., for half-maximal total area of response integrated over time; n = 2 locations in 1 primary dendrite and 1 cell body).


Figure 2
View larger version (32K):
[in this window]
[in a new window]

  		FIGURE 2.
Somatic responses to focal photolysis of single-caged and double-caged IP3. a, calcium signals in response to focal somatic uncaging pulses (arrow) of double-caged IP3 at varying energies (0.76, 0.57, 0.44, 0.32, 0.25, and 0.13 µJ). b, dependence of maximum slope of response (open squares), total area of response (solid squares). and slope of response (open circles) on flash energy for double-caged IP3. c, log-log plot of total area of response as a function of flash energy in wild-type rat Purkinje neurons. d and e, log-log relationship of total area of response (d) and peak response (e) as a function of flash energy in calbindin knock-out mice using single-caged (open symbols) and double-caged (closed symbols) IP3.

 
Because two photolysis events at a double-caged IP3 molecule are required to generate one molecule of IP3, the dependence of the amount of calcium released on flash energy would be expected to show more cooperativity (19) for double-caged IP3 than for single-caged IP3. However, in wild-type rats, for a range of flash energies up to saturating levels, only a weak difference was observed (Fig. 2c), with a log-log slope 2.53 ± 0.19 for double-caged IP3 (44 responses at 4 locations in 3 cells) versus 2.42 ± 0.20 for single-caged IP3 (18 responses, 2 locations in 2 cells; comparison with double-caged IP3 p = 0.36, one-tailed test). Measurements of cytoplasmic calcium in wild-type tissue may not scale linearly with the amount of calcium released because of the presence of the calcium buffer proteins calbindin and parvalbumin, which are plentiful in Purkinje cells (25, 26). In particular, calbindin has fast binding kinetics and can suppress submicromolar fast-rising calcium transients but not larger ones, thus leading to positive cooperativity at a step downstream of calcium mobilization (27). This effect would tend to be independent of and, thus, mask any differences between the observed flash energy dependence of single-caged and double-caged responses. Such an effect would not be expected of parvalbumin, whose binding on-kinetics are entirely slow (28). We therefore made comparisons between single-caged and double-caged IP3 in knock-out mice lacking calbindin (Figs. 2, d and e).

In calbindin knock-out mice differences between double-caged IP3 and single-caged IP3 were seen in the dependence of both the total area of response (Fig. 2d; log-log slope 3.08 ± 0.19 in double-caged IP3 versus 1.40 ± 0.11 in single-caged IP3; ratio of slopes = 2.20 ± 0.22, greater than 1, p < 0.01, two-tailed test) and the peak of fluorescence signal (Fig. 2e; log-log slope 1.22 ± 0.14 in single-caged IP3, 34 responses at 5 locations in 1 cell, normalized to saturated peak responses of 340 ± 70% {Delta}F/F0; 2.75 ± 0.23 in double-caged IP3, 47 responses at 5 locations in 1 cell, normalized to saturated peak responses of 350 ± 40% {Delta}F/F0; ratio of slopes = 2.26 ± 0.31, greater than 1, p < 0.01, two-tailed test) on flash energy. The ratios of slopes are indistinguishable from 2 (p = 0.32 for peak response and p = 0.36 for area), consistent with the expectation that physiological effects produced by chemical two-photon uncaging should have a dependence on flash energy that is the square of the dependence seen with conventional uncaging (19).

IP3 receptors are expressed abundantly throughout Purkinje cells, including dendrites and dendritic spines (24). Because calcium release is necessary for inducing synaptic long-term depression at synaptic inputs impinging on spiny dendrites (3–5), we tested the ability of single-caged IP3 and double-caged IP3 to trigger calcium release in dendrites and single spines (Fig. 3 and Table 1). We performed uncaging in 12 cells filled with 100 µM single-caged IP3, 7 cells filled with 300 µM single-caged IP3, and 78 cells filled with 100 µM double-caged IP3. A successful uncaging response was defined as one in which the peak fluorescence exceeded 50%{Delta}F/F0. We successfully evoked calcium responses in at least one location (soma, dendrite, or spine) in 92% of cells (11 of 12) filled with 100 µM single-caged IP3, 86% of cells (6 of 7) filled with 300 µM single-caged IP3, and 99% of cells (77 of 78) filled with 100 µM double-caged IP3.


View this table:
[in this window]
[in a new window]

  		TABLE 1
Success rate of flash-evoked responses from focal photolysis of caged inositol-1,4,5-trisphosphate agonists A successful trial was defined as one that triggered a peak fluorescence change of 50% {Delta}F/F0 or greater. Boldface values indicate ratios significantly lower than the result for double-caged IP3 in the same structures (p < 0.05, Fisher exact test).

 


Figure 3
View larger version (44K):
[in this window]
[in a new window]

  		FIGURE 3.
Uncaging responses in dendritic spines. Typical responses to uncaging in spiny dendrites of 100 µM double-caged (dc) IP3 (a), 100 µM single-caged IP3 (b), 300 µM single-caged (sc) IP3 (c), and 100 µM caged (c) gPIP2 (d). In the images the circles indicate uncaging locations, and the lines indicate the orientation of the line scan. Scale bars, 1 µm.

 
Uncaging at locations away from the cell body revealed a strong contrast between single-caged IP3 and double-caged IP3. In dendrites (including all except for spiny branchlets), 100 µM double-caged IP3 was still effective at inducing calcium release (84%, 79 successes of 94 locations at which uncaging was done in 45 cells). However, 100 µM single-caged IP3 was considerably less successful (42%, 20 of 48 locations, 12 cells, different from double-caged IP3, p < 0.01, Fisher exact test). The response rate was once again high in recordings in which the concentration of single-caged IP3 was 300 µM (72%, 13 of 18 locations, 7 cells). Thus, in the main dendrites a deficit in calcium release by single-caged IP3 can be overcome by increasing the concentration of caged compound used.

The difficulty of evoking calcium responses with single-caged IP3 reached extremes in the terminal spiny branchlets that give rise to the postsynaptic spines to parallel fibers (Fig. 3). In spiny branchlets, 100 µM double-caged IP3 was still effective at triggering calcium signals (83%, 135 successes of 162 locations, 53 cells), but 100 µM single-caged IP3 failed completely (0%, 0 of 6 locations, 5 cells). Even increasing the concentration of single-caged IP3 to 300 µM did not give a level of response (33%, 3 of 9 locations, 2 cells) that matched the effectiveness of double-caged IP3. Finally, in spines, using double-caged IP3 we still had a high success rate, 74% (89 of 120 spines attempted, 37 cells), whereas 100 or 300 µM single caged IP3 again failed (0%, 0 of 6 spines attempted, 4 cells). Thus, in fine branchlets of the dendritic arbor, single-caged IP3 was considerably less effective at triggering calcium release than double-caged IP3, and in dendritic spines it failed completely.

Calcium release by single-caged IP3 could be blocked by a contaminant such as IP3 itself, which can block release via use-dependent inactivation (29). However, in dendrites 100 µM single-caged IP3 was not effective whether it came from Invitrogen (2 successes of 13 locations) or was synthesized by us (mixed isomers, 1 success of 4 locations; 5-caged IP3, 12 successes of 22 locations). This is consistent with a previous finding that low concentrations (1 µM) of IP3 do not block Purkinje cell responses to single-caged IP3.3 These findings suggest that the lack of calcium release in single-caged IP3 experiments derives from the caged compound itself and not an impurity.

The failure of photolysis of single-caged IP3 to trigger calcium release in the dendritic arbor raises the possibility that single-caged IP3 itself might interfere with calcium release. Based on structure-function analysis (30), the agonist binding pocket of the IP3 receptor interacts with all three phosphate groups of IP3. Screening of derivatives of IP3 indicates that alterations of the 1-phosphate affects binding affinity to the receptor (31), whereas alterations of the 4,5-diphosphate function affect both affinity (32) and agonist activity (22, 33). Thus single-caged IP3, which has either exposed 1- and 4-phosphates or exposed 1- and 5-phosphates, might act as a competitive antagonist to the IP3 receptor. This hypothesis predicts that caged gPIP2, which bears only one unmodified phosphate, would allow calcium release.

We found that in Purkinje cells loaded with 300 µM caged gPIP2, uncaging in dendrites and spines led to robust calcium release (Fig. 3d and Table 1). Photolysis of caged gPIP2 could evoke large calcium responses in the cell body, main dendrite, distal spiny dendrites, and spines (Fig. 4a). To further test whether single-caged IP3 could antagonize calcium release, we measured its ability to interfere with the action of caged gPIP2. We compared responses in cells filled with 300 µM caged gPIP2 with cells filled with 300 µM gPIP2 plus 300 µM single-caged IP3. In the additional presence of 300 µM single-caged IP3, responses were considerably smaller in small structures, with calcium release occurring only in large dendrites close to the cell body (Fig. 4b), reminiscent of the single-caged IP3 experiments. At the farthest points of the dendritic arbor tested, calcium release was entirely absent.


Figure 4
View larger version (84K):
[in this window]
[in a new window]

  		FIGURE 4.
Blocking effects of single-caged IP3 on calcium release. a, an example of a Purkinje neuron filled with 300 µM fluo-5F and 300 µM caged gPIP2. The white traces show calcium signals generated at the indicated locations by a pulse of uncaging light. b, responses in a Purkinje neuron filled with 300 µM fluo-5F, 300 µM caged gPIP2, and 300 µM single-caged IP3 (synthesized by us). Scale bars in a and b, 20 µm for images, 100 ms, and 200% {Delta}F/F0 for calcium signals. c, data pooled from all caged gPIP2 experiments with no additional cage (open symbols) and 100 µM single-caged IP3 obtained commercially (asterisks) and 300 µM commercially obtained (filled squares) or synthesized (filled triangles and circles) single-caged IP3 added. d, response amplitudes sorted according to structure: soma (0.08 µJ), primary dendrite (0.2 µJ), spiny dendrite (0.2 µJ), and dendrite spines (0.2 µJ). The line segments indicate means.

 
For flashes of light that saturated calcium responsiveness, the response size varied as a function of distance from the soma (Fig. 4c). Although photolysis of the gPIP2/single-caged IP3 mixture evoked calcium signals only up to 70 µm away from the cell body, uncaging of 300 µM caged gPIP2 alone successfully evoked calcium release in all structures regardless of distance from the soma. Calcium responses to uncaging over the cell body, main dendrite, spiny dendrites, and spines are plotted in Fig. 4d. On average, responses in cells filled with caged gPIP2/single-caged IP3 mixture were smaller than responses in cells filled with caged gPIP2 alone in every part of the soma and dendrites. Even the addition of a smaller amount of single-caged IP3, 100 µM, did not lead to larger calcium signals (Fig. 4c, asterisks) despite the fact that IP3 is a 10-fold more potent agonist than gPIP2 (23). Taken together, these observations demonstrate that high micromolar levels of single-caged IP3 impair calcium release by other compounds.

We next wanted to quantify the antagonist activity of single-caged IP3 by using the latency to first calcium rise, a distinctive kinetic signature of IP3-dependent calcium release. To determine the amount of IP3 produced we used the inverse relationship between the latency to first calcium rise and IP3 concentration (Fig. 5, a and b) (21). This allowed us to use observations using single-caged IP3 to calibrate the uncaging efficiency of our system and then use the calibration to calculate the amount of IP3 produced from uncaging double-caged IP3. From previous measurements (21) the relationship between latency and IP3 is parametrized by the latency-[IP3] product, 875 ± 85 ms-µM. In our experiments latencies were well fitted to this relationship using an uncaging efficiency factor of g = 0.59 ± 0.07 µJ. This calibration was then applied to double-caged IP3 experiments (Fig. 5c) to calculate the IP3 concentration. Latencies using double-caged IP3 were 9.5 ± 1.7 ms shorter than expected (mean ± S.E., n = 9 flash energies) ms (Fig. 5, d and e). When expressed as the ratio of expected latency to observed latency, a similar discrepancy was seen across a range of flash energies and IP3 concentrations (Fig. 5f), with an average latency ratio of 1.41 ± 0.08 (mean ± S.E., n = 9 flash energies).

We next used double-caged IP3 and caged gPIP2 to investigate the properties of calcium release in spiny branchlets and single spines. When IP3 or gPIP2 is uncaged locally, the resulting calcium transient should be determined by the time course of local IP3 receptor activation and the diffusion and removal of released calcium. For comparison we measured the time course of climbing fiber-evoked signals (Fig. 6a) in which calcium enters during a pan-dendritic action potential lasting only ~10 ms (34). Climbing fiber-evoked spine signals peaked within 8–12 ms and fell to half of the peak value in less than 100 ms. IP3-evoked spine signals rose more slowly and had a more prolonged decline (Fig. 6a, bottom). In this example the time course of the IP3-evoked release signal was similar to a convolution of the climbing fiber response with a flux kernel 180 ms long (Fig. 6a, bottom).

To quantify the approximate duration of flash-evoked calcium release, we measured the time after a stimulus for the fluorescence to rise and then fall to half of its peak value (Fig. 6b). For climbing fiber-evoked action potentials this interval was 77 ± 5 ms (mean ± S.E., n = 65) in dendrites and 56 ± 2 ms (mean ± S.E., n = 80) in spines. For IP3 (signals smaller than 400% {Delta}F/F0) the time from flash to half-fall was significantly longer than action potential-evoked signals both in dendrites (244 ± 10 ms, mean ± S.E., n = 54, p < 0.001) and in spines (305 ± 16 ms, mean ± S.E., n = 27, p < 0.001). The time to half-fall of gPIP2-evoked transients (smaller than 400% {Delta}F/F0) was also longer (in dendrites, 191 ± 11 ms, mean ± S.E., n = 46, p < 0.001; in spines, 255 ± 22 ms, mean ± S.E., n = 10, p < 0.001). These times are longer not only than the time course of action potential-mediated signals but also longer than typical diffusion times from a micron-scale source, which would be expected to be on the order of 10s of ms (35). Thus, both agonists lead to flux through calcium release channels that can last for several hundred milliseconds.


Figure 5
View larger version (36K):
[in this window]
[in a new window]

  		FIGURE 5.
Delays to calcium rise observed using single-caged and double-caged IP3. a and b, sample traces showing decreasing latency with higher uncaging energies with 100 µM double-caged IP3 (a) and single-caged IP3 (b). c, graph of estimated [IP3] as a function of flash energy for single-caged (sc) and double-caged (dc) IP3. The dashed line indicates the estimated concentration of single-caged IP3 produced as an intermediate product from uncaging double-caged IP3. d, delay to calcium rise measured using single-caged IP3 (open symbols) and double-caged IP3 (filled symbols). The shaded gray area indicates the ±1 S.D. confidence band for the fit (solid curve) of single-caged data to directly calibrated measurements by Khodakhah (25) (asterisks). e, inverse-transformed delay data. f, ratio of the expected delay for a given concentration of IP3 (from single-caged data) to the actual delay observed using double-caged IP3.

 
The time course of receptor activation could potentially be shaped in part by degradation of IP3. If this were true, then gPIP2, which is degraded more slowly than IP3, would be expected to trigger longer-lasting calcium signals than IP3. The time from flash to half-fall of gPIP2-evoked signals was not longer than that of IP3-evoked signals in either dendrites (p = 0.87) or spines (p = 0.75). We further tested the role of IP3 degradation pathways by including in the patch electrode ATA (0.5–1 µM), a blocker of IP3 3-kinase, a major breakdown pathway for IP3 (36). Under this condition, after IP3 uncaging the flash to half-fall time was not prolonged (210 ± 30 ms in cell bodies and dendrite, n = 5 locations in 3 cells versus 176 ± 16 ms in n = 3 cell bodies without ATA; also compare with Fig. 6b). Thus, blocking the 3-kinase did not increase the duration of release. Taken together, our measurements indicate that the duration of calcium release is not limited by the rate of IP3 degradation but by IP3 receptor dynamics such as receptor unbinding or inactivation.

The falling kinetics of calcium transients were shorter in dendrites than in spines. In the shafts of spiny dendrites, the time to fall to half-maximal values from peak was IP3 (151 ± 9 ms, mean ± S.E., n = 54) and gPIP2 (137 ± 8 ms, mean ± S.E., n = 46). In spines the time to fall from peak to half-maximum was longer for both agonists (Fig. 6b; IP3, t1/2 = 202 ± 15 ms, mean ± S.E., n = 27; p = 0.13 compared with IP3-evoked spine signals; gPIP2, t1/2 = 197 ± 16 ms, mean ± S.E., n = 10; p = 0.08 compared with gPIP2-evoked spine signals). The fact that spine release transients were longer may be explained by the fact that in dendrites, unconfined diffusion of calcium from a focal source into adjacent volumes would tend to shorten signals.

To characterize flash-evoked calcium release further we measured the peak calcium change as a function of the initial slope of the transient (Fig. 6c). For a given initial slope, IP3- and gPIP2-evoked release reached higher peak values than climbing fiber-evoked transients, as expected for more prolonged release. In CF-evoked transients the peak and slope were nearly linearly related (Fig. 6c; power-law slope 0.77 ± 0.05 in dendrites, 0.89 ± 0.06 in spines), indicating that the duration of calcium entry is not strongly dependent on the amount of flux. In contrast, for uncaging responses the dependence of peak on slope was markedly sublinear (power-law slope for IP3 of 0.40 ± 0.04 in dendrites, 0.38 ± 0.06 in spines; power-law slope for gPIP2 of 0.47 ± 0.04 in dendrites, 0.50 ± 0.06 in spines). Sublinear slopes suggest that the duration of release varies inversely with the initial rate. The shortened duration of calcium release is not caused by depletion of stores; after an initial flash to uncage gPIP2 or IP3, a second flash delivered as soon as 5 ms later can still evoke large amounts of additional release.4 A shortened duration of release is, therefore, likely to be explained by use-dependent or calcium-dependent inactivation of release (37, 38).


Figure 6
View larger version (44K):
[in this window]
[in a new window]

  		FIGURE 6.
Spine and dendrite dynamics of IP3, gPIP2, and climbing fiber-evoked calcium transients. a, sample voltage trace of CF-evoked complex spike (top trace) and corresponding fluorescence change (second trace). CF-evoked calcium transient peaks and falls in less time than the IP3-mediated calcium signal (bottom trace). Convolution of the CF-evoked transient with a ramped kernel function 180 ms long is similar to the IP3-induced calcium signal. b, left, pooled data on the time from stimulus to falling half-maximal values for IP3, gPIP2, and CF-evoked calcium transients (right). Pooled data on the time from peak to half-maximal values. Only signals peaking at less than 400% {Delta}F/F0 were used. c, dependence of calcium transient peak on the slope of the initial rise. d, comparison of peaks of the IP3-induced signals in adjacent structures for IP3 and gPIP2 uncaging. The dark lines indicate the arithmetic means of all experiments. Results are plotted for {Delta}F/F0 peaks of less than 400% in dendrites (dend) and 600% in spines.

 
To quantify the degree of localization of IP3 action, we compared uncaging-evoked calcium responses in spines and dendrites to changes in adjacent structures. When uncaging of IP3 was done in spines, peak fluorescence changes in the spine were larger than in the adjacent dendritic shaft in 28 of 33 cases (greater than chance, p < 0.001). The average ratio of peak spine response to peak shaft response was 1.65 ± 0.26 (mean ± S.E.; Fig. 6d). When gPIP2 was uncaged to generate responses with similar peak amplitude ({Delta}F/F0 < 6), the spine response was larger than the shaft response in 21 of 30 cases (greater than chance, p < 0.05), with an average ratio of 1.20 ± 0.20; Figs. 3d and 6d).

In the converse case, when IP3 uncaging was done in dendritic shafts, peak changes in the shaft were larger than in the immediately adjacent spine in 18 of 19 cases (greater than chance, p < 0.001) and an average ratio of 1.45 ± 0.17. In contrast, gPIP2 uncaging in shafts (peak {Delta}F/F0 < 4) led to a larger response in the shaft than in the adjoining spine in 8 of 16 experiments (not significantly different from chance, p = 0.6) with an average ratio of 1.08 ± 0.18. The results indicate that degradation contributes substantially to restricting the action of IP3 whether it is produced in a spine or in a shaft.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
We find that IP3 can act as a synapse-specific second messenger. In dendritic spines, calcium release triggered by pulses of IP3 leads to prolonged calcium release that is largely restricted to the spine of origin. The duration of release is explainable if IP3 receptor channels remain open for up to hundreds of milliseconds after presentation of agonist.

Our results suggest that the principal determinant of the time course of IP3 receptor opening is internal receptor dynamics. Release duration is inversely related to the initial rate of release, suggesting that receptors may close by inactivating. The fact that a similar inverse relation is seen using a poorly hydrolyzable agonist, gPIP2, indicates that IP3 degradation enzymes do not additionally limit release duration.

The biological function of IP3 degradation enzymes may be to localize signals to single spines. IP3 3-kinase and 5-phosphatase are expressed strongly in Purkinje cell spines (39–41). We propose that these enzymes degrade enough IP3 to prevent its action from spreading out of a spine. Because calcium release is necessary for the induction of cerebellar long-term depression (4–6), confinement of IP3 to the spine in which it is originally generated provides a means of limiting synaptic plasticity to active synapses and not nearby inactive ones.

An upper limit to how much calcium release can spread from one spine to another can be obtained by multiplying the ratio of peak changes in target spines and adjacent shafts (1.65) with the ratio for the converse case (1.45) to obtain a product of ~2.4. This product indicates that when IP3 is produced, calcium signals would be at least 2.4-fold smaller in nearby spines than in the spine where IP3 is produced. The actual factor would be even greater for two reasons. First, IP3 diffusing from one spine to another would be diluted over the distance between spine necks, a factor not considered here. Second, in some of our experiments uncaging was likely to be spatially dispersed due to light scattering by tissue, implying that the measured ratios of signals are upper bounds to the true ratios that would be observed if IP3 were produced entirely within a single spine.

How much of the IP3 made in a spine is metabolized before it gets out? This quantity can be estimated by converting spine-shaft ratios into estimates of IP3 agonist concentration. Assuming a 2.4-fold cooperativity of agonist to trigger calcium release, a ratio of 1.65 between peak calcium at a spine release site over a nearby shaft corresponds to a 19% drop in IP3, and for shaft uncaging the observed ratio of 1.45 corresponds to a leak from shaft into spines of 14%. In both cases the difference is likely to be a lower limit because both unfocused uncaging and calcium diffusion between spine and shaft would tend to equalize observed calcium gradients relative to the expected IP3 gradient. Therefore, our findings are consistent with an interpetation that at least one-fifth of spine-produced IP3 is prevented from passing through spine necks due to degradation.

Taken together, these estimates suggest the following model for the action of IP3 in Purkinje neurons. When IP3 is produced in a spine, degradation contributes substantially to restricting its action. Once the remaining IP3 escapes to the shaft, diffusion and further degradation act to end calcium release. Because the accurate interpretation of IP3-evoked calcium measurements depends on understanding how calcium itself moves, testing and refinement of this model will require measurement of the calcium economy of Purkinje neuron spines.

Uncaging of IP3 to activate spine calcium release was made possible by the use of 4,5-bis-NPE-IP3 (double-caged IP3). We find that modifying IP3 with light-sensitive cage groups at multiple locations generates a caged compound that can be used to trigger calcium release over a wide range of concentrations. Double-caged IP3 can be used in brain slices to achieve submicron resolution in dendritic spines of cerebellar Purkinje neurons. In particular, IP3 modified at multiple sites is useful at high micromolar concentrations, a range in which the single-caged version of the compound interferes with IP3 receptor activity.

We find that single-caged IP3 blocks calcium release in Purkinje cells. The structural similarity of single-caged IP3 to heparin, a potent blocker of IP3-dependent calcium release with micromolar affinity (42–44), is suggestive because heparin consists of linked six-carbon rings each bearing two anionic groups. Thus, blockade of calcium release by single-caged IP3 is likely to take place by direct antagonism at the IP3 receptor. Although we do not know the binding affinity of single-caged IP3 for the IP3 receptor, an estimate can be made from the 1.4-fold difference in latency using 100 µM single-caged IP3. Assuming a single-site competition model, this latency shift would be accounted for by an IC50 of 140 µM. This value would be consistent with the impairment of calcium release in neurons filled with 100–300 µM single-caged IP3.

The antagonist activity of single-caged IP3 is of particular relevance for experiments done at high concentrations, as is done in neurons. Sensitivity of the inositol 1,4,5-trisphosphate receptor to IP3 varies widely under different conditions. In vitro studies report sensitivity that varies considerably with the cell type (45–47), with affinities in the nanomolar range in hepatocytes (48), pancreatic acinar cells (49, 50), and smooth muscle (51) but an EC50 of 25 µM in neurons in controlled cytoplasmic environments (52) and intact cell bodies (21, 50, 53). In experiments on non-neuronal cells, up to 10 µM single-caged IP3 has been used, a concentration that does not activate or block calcium release (12). However, in the dendrites of cerebellar Purkinje cells, calcium release has been activated, using up to 600 µM single-caged IP3 (3, 54), in contrast with the concentrations of less than 100 µM typically used in Purkinje cell somata (5, 21). Our observation that high concentrations of single-caged IP3 interfere with release would account for the discrepancy between past somatic and dendritic experimental designs. In the future, for controlling calcium release in dendrites, double-caged IP3 is the preferred reagent.

The use of double-caged IP3 will still generate a certain amount of single-caged IP3 as a residual by-product (Fig. 5c). However, since relatively little double-caged compound is necessary to evoke release (typically 100 µM), the amount of single-caged IP3 produced in a flash would never exceed 50 µM, well below our estimated IC50 of 140 µM. Thus double-caging is a sufficient strategy for IP3, and an even more secure approach such as triple-caging is unnecessary.

The difference in the effectiveness of single-caged IP3 and double-caged IP3 (or caged gPIP2) was more apparent in dendrites than in cell bodies. However, it should be noted that single-caged IP3 suppressed gPIP2-evoked calcium release at all locations (Fig. 4, c and d). Considering that calcium release is a strongly nonlinear function of IP3 concentration (Fig. 2) (53), our observations are broadly consistent with a blocking effect of single-caged IP3 at all cellular locations.

An additional reason for a difference between soma and dendrites is incomplete diffusion equilibration from soma into the dendrites, leading to lower levels of caged compound (and, therefore, generating lower levels of IP3) in the dendrites. In addition to general free diffusion, movement of single-caged IP3 through the dendrites and to the spines might be impeded by binding interactions with IP3 receptors. The levels of IP3 produced might be further limited by the fact that dendrites from neurons recorded near the slice surface run at greater depths, where uncaging light is more likely to be scattered or absorbed. These factors would tend to reduce the effectiveness of conventional single-caged IP3 in remote regions. In somata, single-caged IP3 remains a useful tool when used at concentrations of 100 µM or less.

Our findings suggest a general principle in the design and use of caged compounds. Whenever a single-caged agonist may interfere with receptor action, a compound caged at multiple sites is less similar to the structure of the native agonist than the single-caged compounds and would, therefore, be less likely to interact with the receptor. The same principle would apply to residual agonist activity, as has been observed in the case of caged ATP and potassium channels (10). An additional advantage comes from ease of handling; because of the requirement for multiple uncaging steps, compounds caged at more than one site are less likely to accumulate free agonist during handling under standard room illumination. Thus, multiply caged compounds may have general value in the convenient manipulation of receptor-activated physiological processes.


   FOOTNOTES
 
* This work was supported by a Whitehall Foundation grant, a W. M. Keck Foundation Young Investigator Award, National Science Foundation Grant 0347719, and National Institutes of Health Grant NS045193 (to S. S.-H. W.) and a Burroughs-Wellcome Interfaces of Science fellowship (to D. V. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 To whom correspondence should be addressed: Dept. of Molecular Biology, Lewis Thomas Laboratory, Washington Rd., Princeton, NJ 08544. Tel.: 609-258-0388; Fax: 609-258-1028; E-mail: sswang{at}princeton.edu.

2 The abbreviations used are: gPIP2, glycerophosphoryl-myo-inositol 4,5-bisphosphate; NPE, 1-(2-nitrophenyl)ethyl; IP3, inositol 1,4,5-trisphosphate; CF, climbing fiber. Back

3 K. Khodakhah, personal communication. Back

4 D. Sarkisov and S. S.-H. Wang, manuscript in preparation. Back


   ACKNOWLEDGMENTS
 
We thank M. Cho and S. Kyin for technical assistance, K. Khodakhah for helpful discussions, and M. Schell for advice on IP3 3-kinase inhibitors.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Carafoli, E., and Klee, C. B. (1999) Calcium as a Cellular Regulator, Oxford University Press, New York
  2. Takechi, H., Eilers, J., and Konnerth, A. (1998) Nature 396, 757-760[CrossRef][Medline] [Order article via Infotrieve]
  3. Finch, E. A., and Augustine, G. J. (1998) Nature 396, 753-756[CrossRef][Medline] [Order article via Infotrieve]
  4. Wang, S. S.-H., Denk, W., and Häusser, M. (2000) Nat. Neurosci. 3, 1266-1273[CrossRef][Medline] [Order article via Infotrieve]
  5. Khodakhah, K., and Armstrong, C. M. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 14009-14014[Abstract/Free Full Text]
  6. Inoue, T., Kato, K., Kohda, K., and Mikoshiba, K. (1998) J. Neurosci. 18, 5366-5373[Abstract/Free Full Text]
  7. Ito, M. (2002) Nat. Rev. Neurosci. 3, 896-902[Medline] [Order article via Infotrieve]
  8. Lester, H. A., and Nerbonne, J. M. (1982) Annu. Rev. Biophys. Bioeng. 11, 151-175[CrossRef][Medline] [Order article via Infotrieve]
  9. Adams, S. R., and Tsien, R. Y. (1993) Annu. Rev. Physiol. 55, 755-784[CrossRef][Medline] [Order article via Infotrieve]
  10. Nerbonne, J. M. (1996) Curr. Opin. Neurobiol. 6, 379-386[CrossRef][Medline] [Order article via Infotrieve]
  11. Walker, J. W., Somlyo, A. V., Goldman, Y. E., Somlyo, A. P., and Trentham, D. R. (1987) Nature 327, 249-252[CrossRef][Medline] [Order article via Infotrieve]
  12. Walker, J. W., Feeney, J., and Trentham, D. R. (1989) Biochemistry 28, 3272-3280[CrossRef][Medline] [Order article via Infotrieve]
  13. McCarron, J. G., MacMillan, D., Bradley, K. N., Chalmers, S., and Muir, T. C. (2004) J. Biol. Chem. 279, 8417-8427[Abstract/Free Full Text]
  14. Leite, M. F., Thrower, E. C., Echevarria, W., Koulen, P., Hirata, K., Bennett, A. M., Ehrlich, B. E., and Nathanson, M. H. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 2975-2980[Abstract/Free Full Text]
  15. Echevarria, W., Leite, M. F., Guerra, M. T., Zipfel, W. R., and Nathanson, M. H. (2003) Nat. Cell Biol. 5, 440-446[CrossRef][Medline] [Order article via Infotrieve]
  16. Wang, S. S.-H., and Augustine, G. J. (1995) Neuron 15, 755-760[CrossRef][Medline] [Order article via Infotrieve]
  17. Fierro, L., DiPolo, R., and Llano, I. I. (1998) J. Physiol. (Lond.) 510, 499-512[Abstract/Free Full Text]
  18. Fiacco, T. A., and McCarthy, K. D. (2004) J. Neurosci. 24, 722-732[Abstract/Free Full Text]
  19. Pettit, D. L., Wang, S. S.-H., Gee, K. R., and Augustine, G. J. (1997) Neuron 19, 465-471[CrossRef][Medline] [Order article via Infotrieve]
  20. Sarkisov, D. V., and Wang, S. S.-H. (2006) Nat. Protoc. 1, 828-832[CrossRef][Medline] [Order article via Infotrieve]
  21. Khodakhah, K., and Ogden, D. (1995) J. Physiol. (Lond.) 487, 343-358[Medline] [Order article via Infotrieve]
  22. Wilcox, R. A., Primrose, W. U., Nahorski, S. R., and Challiss, R. A. (1998) Trends Pharmacol. Sci. 19, 467-475[CrossRef][Medline] [Order article via Infotrieve]
  23. Bird, G. S., Obie, J. F., and Putney, J. W., Jr. (1992) J. Biol. Chem. 267, 17722-17725[Abstract/Free Full Text]
  24. Satoh, T., Ross, C. A., Villa, A., Supattapone, S., Pozzan, T., Snyder, S. H., and Meldolesi, J. (1990) J. Cell Biol. 111, 615-624[Abstract/Free Full Text]
  25. Fierro, L., and Llano, I. (1996) J. Physiol. (Lond.) 496, 617-625[Medline] [Order article via Infotrieve]
  26. Schmidt, H., Stiefel, K. M., Racay, P., Schwaller, B., and Eilers, J. (2003) J. Physiol. (Lond.) 551, 13-32[Abstract/Free Full Text]
  27. Maeda, H., Ellis-Davies, G. C., Ito, K., Miyashita, Y., and Kasai, H. (1999) Neuron 24, 989-1002[CrossRef][Medline] [Order article via Infotrieve]
  28. Lee, S. H., Schwaller, B., and Neher, E. (2000) J. Physiol. (Lond.) 525, 419-432[Abstract/Free Full Text]
  29. Hajnoczky, G., and Thomas, A. P. (1994) Nature 370, 474-477[CrossRef][Medline] [Order article via Infotrieve]
  30. Bosanac, I., Michikawa, T., Mikoshiba, K., and Ikura, M. (2004) Biochim. Biophys. Acta 1742, 89-102[Medline] [Order article via Infotrieve]
  31. Nakanishi, W., Kikuchi, K., Inoue, T., Hirose, K., Iino, M., and Nagano, T. (2002) Bioorg. Med. Chem. Lett. 12, 911-913[CrossRef][Medline] [Order article via Infotrieve]
  32. Wilcox, R. A., Fauq, A., Kozikowski, A. P., and Nahorski, S. R. (1997) FEBS Lett. 402, 241-245[CrossRef][Medline] [Order article via Infotrieve]
  33. Womack, M. D., Walker, J. W., and Khodakhah, K. (2000) J. Gen. Physiol. 115, 339-346[Abstract/Free Full Text]
  34. Llinas, R., and Sugimori, M. (1980) J. Physiol. (Lond.) 305, 197-213[Abstract/Free Full Text]
  35. Svoboda, K., Tank, D. W., and Denk, W. (1996) Science 272, 716-719[Abstract]
  36. Mayr, G.W., Windhorst, S., and Hillemeier, K. (2005) J. Biol. Chem. 280, 13229-13240[Abstract/Free Full Text]
  37. Michikawa, T., Hirota, J., Kawano, S., Hiraoka, M., Yamada, M., Furuichi, T., and Mikoshiba, K. (1999) Neuron 23, 799-808[CrossRef][Medline] [Order article via Infotrieve]
  38. Parker, I., Yao, Y., and Ilyin, V. (1996) Biophys. J. 70, 222-237[Medline] [Order article via Infotrieve]
  39. Go, M., Uchida, T., Takazawa, K., Endo, T., Erneux, C., Mailleux, P., and Onaya, T. (1993) Neurosci. Lett. 158, 135-138[CrossRef][Medline] [Order article via Infotrieve]
  40. Yamada, M., Kakita, A., Mizuguchi, M., Rhee, S. G., Kim, S. U., and Ikuta, F. (1993) Brain Res. 606, 335-340[CrossRef][Medline] [Order article via Infotrieve]
  41. De Smedt, F., Boom, A., Pesesse, X., Schiffmann, S. N., and Erneux, C. (1996) J. Biol. Chem. 271, 10419-10424[Abstract/Free Full Text]
  42. Nilsson, T., Zwiller, J., Boynton, A. L., and Berggren, P. O. (1988) FEBS Lett. 229, 211-214[CrossRef][Medline] [Order article via Infotrieve]
  43. Cullen, P. J., Comerford, J. G., and Dawson, A. P. (1988) FEBS Lett. 228, 57-59[CrossRef][Medline] [Order article via Infotrieve]
  44. Ghosh, T. K., Eis, P. S., Mullaney, J. M., Ebert, C. L., and Gill, D. L. (1988) J. Biol. Chem. 263, 11075-11079[Abstract/Free Full Text]
  45. Ferris, C. D., Huganir, R. L., Supattapone, S., and Snyder, S. H. (1989) Nature 342, 87-89[CrossRef][Medline] [Order article via Infotrieve]
  46. Chadwick, C. C., Saito, A., and Fleischer, S. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 2132-2136[Abstract/Free Full Text]
  47. Meyer, T., Wensel, T., and Stryer, L. (1990) Biochemistry 29, 32-37[CrossRef][Medline] [Order article via Infotrieve]
  48. Hajnoczky, G., Hager, R., and Thomas, A. P. (1999) J. Biol. Chem. 274, 14157-14162[Abstract/Free Full