Streamlined synaptic vesicle cycle in cone photoreceptor terminals

Neuron,Vol.41,755–766,March 4,2004,Copyright 2004by Cell Press

Streamlined Synaptic Vesicle Cycle in Cone Photoreceptor Terminals

thews,1999;Fuchs et al.,2003).Here we ask whether,in addition to exocytosis and endocytosis,other steps in the vesicle cycle are also specialized for continuous Ruth Rea,1Jian Li,2Ajay Dharia,1Edwin S.Levitan,3Peter Sterling,2and Richard H.Kramer 1,*release in cone photoreceptors.

1

Department of Molecular and Cell Biology At conventional phasic synapses that release trans-Division of Neurobiology

mitter during action potentials,the routes of vesicle re-University of California at Berkeley cycling are often direct.This can include rapid reuse,Berkeley,California 94720where endocytosed vesicles are preferentially rere-2

Department of Neuroscience

leased (Ceccarelli et al.,1973;Murthy and Stevens,University of Pennsylvania School of Medicine 1998;Pyle et al.,2000;Richards et al.,2003),and kiss-Philadelphia,Pennsylvania 19104and-run exocytosis,where docked vesicles release mul-3

Department of Pharmacology tiple times while remaining attached to the plasma mem-University of Pittsburgh

brane (Klingauf et al.,1998;Stevens and Williams,2000;Pittsburgh,Pennsylvania 15261

Gandhi and Stevens,2003;Aravanis et al.,2003).How-ever,there are also detours in the vesicle cycle,where retrieved membrane is stored and perhaps processed Summary

before reuse.For example,during endocytosis,recy-cling can occur through bulk membrane retrieval,re-Cone photoreceptors tonically release neurotransmit-sulting in large vacuoles that eventually give rise to new ter in the dark through a continuous cycle of exo-synaptic vesicles (Takei et al.,1996;Richards et al.,cytosis and endocytosis.Here,using the synaptic vesi-2000;Teng and Wilkinson,2000).Moreover,small inter-cle marker FM1-43,we elucidate specialized features nalized vesicles may be recycled by first fusing with of the vesicle cycle.Unlike retinal bipolar cell termi-intracellular endosomes from which new vesicles are nals,where stimulation triggers bulk membrane re-born (Heuser and Reese,1973;Ryan et al.,1996),espe-trieval,cone terminals appear to exclusively endocy-cially during intense stimulation (Richards et al.,2003).tose small vesicles.These retain their integrity until Finally,at conventional synapses only a small minority exocytosis,without pooling their membranes in endo-of vesicles is “readily releasable”(for review,see Zucker somes.Endocytosed vesicles rapidly disperse through and Regehr,2002).Most are relatively distant from re-the terminal and are reused with no apparent delay.lease sites and immobilized by being tethered to the Unlike other synapses where most vesicles are immo-cytoskeleton by proteins such as synapsin (Hilfiker et bilized and held in reserve,only a small fraction (?15%)al.,1999).These “reserve vesicles”are resistant to re-becomes immobilized in cones.Photobleaching ex-lease and only slowly equilibrate with the readily-releas-periments suggest that vesicles move by diffusion and able pool.Hence,a large fraction of synaptic membrane not by molecular motors on the cytoskeleton and that is side tracked from participating in the vesicle cycle.vesicle movement is not rate limiting for release.The The vesicle cycle at ribbon synapses has been studied huge reservoir of vesicles that move rapidly through-primarily in the retinal bipolar cell,which like photore-out cone terminals and the lack of a reserve pool are ceptors contain hundreds of thousands of synaptic vesi-unique features,providing cones with a steady supply cles and employ graded release (Lagnado et al.,1996;for continuous release.von Gersdorff et al.,1996).However,release from the bipolar cell has a strong phasic component.Step depo-Introduction

larization triggers a transient burst of release (6000vesi-cles released over 200ms)followed by a dramatic slow-Vertebrate photoreceptors tonically release neurotrans-ing,suggesting exhaustion of a readily-releasable pool mitter in the dark.To achieve a continuous cycle of that refills only after tens of seconds (von Gersdorff and exocytosis and endocytosis necessary for tonic release,Matthews,1997,1999).Tracer studies show that actively the photoreceptor terminal has functional and structural cycling vesicles remain within 1?m of the plasma mem-specializations.Exocytosis is unusually sensitive to Ca 2?brane (Lagnado et al.,1996;Paillart et al.,2003).Sus-(Rieke and Schwartz,1996),and vesicles are released tained stimulation eventually labels reserve vesicles in at an unusually high rate in the dark,allowing small the center of the bipolar terminal that appear immobile changes in light intensity to generate significant changes and resistant to release (Lagnado et al.,1996),although in neurotransmitter release (Ashmore and Copenhagen,recent studies challenge this view (Holt et al.,2004).In 1983;Rao-Mirotznik et al.,1998).Endocytosis occurs addition to endocytosis of small vesicles,stimulation with an unusually short delay after a rise in Ca 2?,rapidly triggers bulk membrane retrieval,resulting in large vacu-and accurately retrieving membrane added by exo-oles in the bipolar terminal (Holt et al.,2003;Paillart et cytosis (Rieke and Schwartz,1996).The photoreceptor al.,2003).

terminal has synaptic ribbons,the key structural special-How direct is the vesicle cycle in the cone synapse,ization that allows rapid and coordinated exocytosis and which unlike conventional synapses,tonically release endocytosis (for review,see von Gersdorff and Mat-transmitter?Is there bulk membrane retrieval or recy-cling through endosomes?Is there a large pool of immo-bilized reserve vesicles?Or,does the cone synapse

*Correspondence:rhkramer@https://www.360docs.net/doc/8411067227.html,

Neuron

756

Figure1.Specific Uptake of FM1-43into Cone Terminals Is Achieved by Loading the Retina before Dissociation

(A)Single cone photoreceptor,loaded with FM1-43prior to dissociation from the retina.Transmitted light and fluorescence images are overlaid. FM1-43fluorescence is shown in green.T,synaptic terminal.

(B)Single cone loaded with FM1-43following dissociation.IS,inner segment;OS,outer segment.

(C)Retinal slice,retina was loaded with FM1-43prior to slicing.The position of the cone OS,outer plexiform layer(OPL),inner nuclear layer (INL),and inner plexiform layer(IPL)are indicated.

(D)Retinal slice,loaded with FM1-43after slicing.

Scale bars,10?m.

support continuous release by employing a simpler vesi-plexiform later(OPL)and in some cone outer segments,

finely speckled labeling in the inner plexifom layer(IPL), cle cycle without these detours?At other synapses,the

fluorescent styryl dye FM1-43has been used to explore but almost no labeling in cone inner segments or the the vesicle cycle(for review,see Cochilla et al.,1999;

inner nuclear layer(INL).Similar OPL-specific labeling Ryan,2001).After staining the plasma membrane,the was seen when FM1-43was added to intact retina from dye is internalized during endocytosis and can subse-

tiger salamander,Xenopus frogs,goldfish,and rabbit quently be released during exocytosis.However,FM1-(not shown).Both rod and cone terminals were labeled in

these preparations,with little labeling of photoreceptor 43can be taken up by other mechanisms into nonsynap-

tic portions of dissociated neurons(Rouze and Schwartz,inner segments or the INL.Hence,by using intact retina, 1998;Meyers et al.,2003),preventing its effective use

followed by a wash with Advasep-7,we have discovered in photoreceptor synapses.We find that when the dye a generally applicable method for synapse-specific is applied to the intact retina under conditions that stim-

loading of FM1-43in retinal photoreceptors.

ulate continuous release,labeling is selective for synap-When FM1-43was applied after retinal dissociation

(Figure1C)or slicing(Figure1D),the labeling was not tic terminals.Anole lizards have a cone-only retina,so

in this preparation,labeling in the outer plexiform(OPL)specific to terminals.The isolated cone showed labeling is specific to cone https://www.360docs.net/doc/8411067227.html,ing this preparation and

in all regions and was brightest in the inner segment FM1-43we explore the vesicle cycle of cones.(Figure1C).Slices showed labeling in all layers,obscur-

ing clear visualization of cone terminals in the OPL(Fig-

ure1D).This extrasynaptic labeling after dissociation or Results

slicing occurred in the absence of extracellular Ca2?.

Hence,uptake of FM1-43is synapse specific when dye Specific FM1-43Labeling of Cone Terminals

is added to intact retina but not when it is added after Requires Intact Tissue

perturbing the retina by slicing or cell dissociation.Dye We have optimized a procedure to specifically load FM1-

uptake into extrasynaptic regions after perturbation 43into synaptic vesicles of cone terminals in the OPL.

possibly occurs through membrane resealing mecha-The intact dark-adapted lizard retina was incubated in

nisms that involve exocytosis and endocytosis(McNeil normal saline in the dark with45?M dye for15min.To

and Terasaki,2001)or through activation of ion channels prevent exocytosis,which would unload the dye,the

that are permeable to FM1-43(Meyers et al.,2003). retina was then transferred to a Ca2?-free solution con-

taining EGTA.This solution also included Advasep-7,a

cyclodextran that helps remove FM1-43from surface Loading and Unloading Are Ca2?Dependent membranes(Kay et al.,1999).FM1-43labeling was then

To assess whether FM1-43loading is Ca2?dependent, examined in individual dissociated cones(Figure1A)FM1-43was added to intact retina in saline containing and retinal slices(Figure1B).

either1.5mM Ca2?or zero Ca2?plus1mM EGTA(0 In preloaded dissociated cones(Figure1A),intense Ca2?/EGTA).A strong band of labeling appeared within FM1-43fluorescence is apparent in the terminals but

the OPL when FM1-43was added in the presence but not in inner segments.In slices(Figure1B),there was not in the absence of Ca2?(Figure2A).In three experi-strong labeling in rows of cone terminals in the outer

ments with Ca2?-containing saline,there was much

Streamlined Vesicle Cycle in Cone Photoreceptors

757

Figure2.Loading and Unloading of FM1-43

Is Ca2?Dependent

(A)Pseudocolor images of slices taken from

retina preloaded with FM1-43in the presence

or absence of Ca2?.Dashed boxes indicate

OPL and INL regions used for analysis of fluo-

rescence intensity,F.Scale bar,20?m.

(B)OPL fluorescence,F OPL,following Ca2?-

dependent and Ca2?-independent loading of

FM1-43.To compensate for differences in

background loading,the fluorescence of a

region of the IPL was subtracted(F OPL?F IPL).

(C)A single slice from retina preloaded with

FM1-43was imaged every30s in high K?,

Ca2?-free solution and1.5mM Ca2?was ap-

plied as indicated.Fluorescence measure-

ments across the OPL are plotted against

time.

(D)Average loss of fluorescence from retina

preloaded with FM1-43after different times

in high K?/Ca2?solutions(n?3–5slices).

(E)Kinetics of FM1-43loading in Ca2?/high

K?saline.Fluorescence measurements were

taken from the OPL of retina that were loaded

for indicated times.Continuous line is a single

exponential fit with??332s.

(F)Kinetics of FM1-43unloading in Ca2?/high

K?saline.Cone terminals in preloaded retina

were visualized in retina slices,and loss of

fluorescence was monitored at10s intervals.

Continuous line is a single exponential fit with

??345s.

greater labeling intensity in this region of the OPL com-slices of retina containing FM1-43-loaded terminals pared to a region of the same size in the INL,whereas

were exposed to Ca2?/high K?saline,and the decline in Ca2?-free saline,there was no difference in OPL ver-in fluorescence was measured at10s intervals(Figure

2F).These experiments showed that unloading also oc-sus INL labeling(Figure2B).Hence,FM1-43loading into

cone terminals is Ca2?dependent.curs over a single exponential time course(??345s; To assess whether un loading of FM1-43is also Ca2?

n?6).Hence,loading and unloading occur over a nearly dependent,retinas were loaded and sliced and then identical time course,consistent with the idea that endo-exposed to saline containing high K?(50mM)to depolar-

cytosis and exocytosis are closely linked in photorecep-ize cones and open voltage-gated Ca2?channels.In tor terminals(Rieke and Schwartz,1996).

an individual slice(Figure2C),FM1-43release could

repeatedly be started by exposure to Ca2?/high K?sa-Loaded FM1-43Is Restricted to Synaptic Vesicles line and stopped by exposure to0Ca2?/EGTA.Analysis

Electron micrographs of thin(90nm)sections showed of FM1-43release from many slices is shown in Figure the cone terminals to be densely packed with small 2D.All slices were incubated for a total period of15min

synaptic vesicles throughout the terminal(Figure3A). and exposed for various periods of time to Ca2?/high The terminals were roughly hemispherical with an aver-K?saline.Dye unloads progressively with increasing

age diameter of6?m,so we estimate from the number time in Ca2?,with more than half unloading within5min.of vesicles in a single section that a terminal contains Hence,unloading of FM1-43is Ca2?dependent.

about250,000synaptic https://www.360docs.net/doc/8411067227.html,rger membranous To learn more about the kinetics of endocytosis and structures(endosomes or cisternae)were absent.Syn-exocytosis,we carefully examined the time course of

aptic ribbons were also seen at characteristic triad junc-FM1-43loading and unloading(Figures2E and2F).To tions with horizontal and bipolar cell dendrites,and there measure the kinetics of loading,intact retinas were ex-

were also a few mitochondria.

posed to dye in Ca2?/high K?saline for a variable time FM1-43inserts into plasma membrane and becomes (2–20min)before washing with0Ca2?/EGTA,slicing,

internalized upon endocytosis of vesicles(Betz et al., and analyzing fluorescence in cone terminals.Data from1992).However,FM1-43can also enter some cells 40retinas(n?3–6per time point)showed that dye

through ion channels,including the vanilloid receptor accumulation could be fit with a single exponential time(TRPV1)and a purinergic receptor(P2X2)(Meyers et al., course(??332s)that reached steady state within about

2003).Hence,under certain conditions,FM1-43loading 10min(Figure2E).To measure the rate of unloading,in retinal bipolar cells(Rouze and Schwartz,1998)and

Neuron

758

Figure3.AM1-43Is Localized to Synaptic

Vesicles

(A)Electron micrograph of a lizard cone termi-

nal filed with thousands of synaptic vesicles

dispersed throughout.Also shown are synap-

tic ribbons(r),postsynaptic dendrites from

horizontal and bipolar cells(hc and bc),and

mitochondria(mit).

(B)Photobleaching of AM1-43in the pres-

ence of DAB.The blue rectangle indicates the

region centered on the OPL that was exposed

to photobleaching laser light.Fluorescence

and transmitted light images are overlaid to

show AM1-43fluorescence(green)and DAB

reaction product(black)following photocon-

version.Scale bar,10?m.

(C)Semithin section of a slice photocon-

verted as in(B).The blue rectangle indicates

the photobleached region.DAB reaction

product in cone terminals appears brown.

The tissue was counterstained with Toluidine

blue.Scale bar,25?m.

(D and E)Electron micrographs of an AM1-

43photoconverted(D)and control(E)cone

terminal.Dense DAB reaction product is lo-

calized to synaptic vesicles in(D).Scale bar,

250nm.

cochlear hair cells(Meyers et al.,2003)does not require Movement of Endocytosed Vesicles

extracellular Ca2?and is not readily unloaded by promot-

Where do vesicles go after endocytosis in cone termi-ing Ca2?entry.Since some TRP channels are activated nals?To answer this question,cone terminals in intact by Ca2?(Launay et al.,2002),it seemed possible that

retina were loaded with FM1-43in the dark.The retina such channels could contribute to Ca2?-dependent up-was then washed with Advasep-7,to remove dye from take of FM1-43in cone terminals.To test whether load-

surface membranes,plus0Ca2?/EGTA,to prevent un-ing in cone terminals results exclusively from endocyto-loading.Then,in the continued presence of0Ca2?/

EGTA,we applied FM4-64to selectively stain surface sis,we used AM1-43,a structural analog of FM1-43that

can be crosslinked to proteins by fixation with glutaral-membranes.FM4-64is chemically similar to FM1-43,but dehyde.Intense illumination of AM1-43,in the presence

it has a different emission spectrum,allowing labeling by of diaminobenzidine(DAB),results in“photoconversion”the two dyes to be distinguished.Thus,green FM1-43 of the DAB to form an electron-dense reaction product

fluorescence appeared in the interior of cone terminals that is visible with electron microscopy(EM)(Henkel(Figure4A)with red FM4-64on the surface membranes

of cone terminals and other cells in the vicinity(Figure4B). et al.,1996a;Harata et al.,2001).Figure3B shows a

fluorescence image of a slice with AM1-43-loaded termi-Superimposing the two images(Figure4C)shows nals that had been photobleached by prolonged illumi-

FM1-43entirely filling the cone terminal,suggesting that nation with strong488nm light.Transmitted light images endocytosed vesicles are not restricted to specific re-from this slice show the brown reaction product local-

gions and that no additional structures external to the ized in cone terminals(Figure3C).EM images from pho-cone terminal are labeled with FM1-43.A density profile toconverted(Figure3C)and control regions(Figure3D)

taken through a terminal(Figure4D)shows that Ca2?-of the slice show that a large proportion(?80%)of dependent labeling with FM1-43and Ca2?-independent synaptic vesicles in the photoconverted region were

labeling with FM4-64show no overlap except within1 filled with electron-dense material,whereas in the con-?m of the membrane,near the limit of our spatial resolu-trol region,?1%of the vesicles were electron dense.

tion.Hence endocytosed vesicles disperse throughout Similar results were obtained in three experiments.cone terminals,and these are the only structures in the About50%of photoconverted vesicles had electron-

OPL that incorporate appreciable amounts of FM1-43. dense“coats,”suggesting a role for clathrin in endocy-How rapidly do endocytosed vesicles disperse?To tosis(Brodin et al.,2000;Slepnev and De Camilli,2000).

address this,we loaded with AM1-43in the dark,wash-Notably,photoconversion did not raise the background ing with Advasep-7and0Ca2?/EGTA,and then fixed density of cytoplasm,suggesting that AM1-43was re-

with glutaraldehyde.Fluorescence distributed nearly stricted to the lumen of vesicles.Hence,dye loading of evenly through the terminal,either with a2min or15 cone terminals occurs specifically through endocytosis

min loading period(Figures5A and5B).The normalized of vesicles.density profiles in Figure5C show that,even though

Streamlined Vesicle Cycle in Cone Photoreceptors

759

Figure4.FM1-43Fills the Entire Cone Terminal and No Other Structures in the OPL

(A)Confocal image of two cone terminals in a slice previously loaded with green FM1-43with Ca2?and then placed in red FM4-64without Ca2?.A500–550nm emission filter was used to pass only the FM1-43fluorescence.Scale bar,5?m.

(B)The same region imaged through a650LP filter to pass only FM4-64fluorescence.Gain settings were previously adjusted to remove bleedthrough of FM1-43fluorescence(about20%of the total signal).

(C)Overlay of images in(A)and(B),showing localization of the two dyes.Green labeling is in synaptic vesicles inside the terminals and red labeling is on surface membranes.

(D)Average fluorescence measured across one terminal,showing the intensity profile of both dyes.The region used for analysis is indicated in(C)(white dotted rectangle).

less total fluorescence is present,terminals loaded for the FM1-43such that vesicles within the region were ?2min show the same fluorescence profile as terminals

no longer fluorescent(Figure6A).Over time,the fluores-loaded for15min.cence within the bleached region recovered and fluores-After dispersal,do labeled vesicles remain mobile?To

cence in the remainder of the terminal decreased as address this question,loaded terminals were exposed to vesicles redistributed within the terminal.In control ex-Ca2?/high K?saline previously.Fluorescence remained

periments,application of high K?/Ca2?saline was effec-nearly uniform across the terminal,both after a short2tive in unloading the FM1-43that remained after local min unloading period(Figure5D)and a long15min

photobleaching,indicating that the procedure had not unloading period(Figure5E).Again,a density profile functionally damaged the terminal.Fluorescence inten-through the terminal shows that the distribution of fluo-

sity profiles across the terminal including the bleached rescence remains remarkably similar both at the begin-region show how the fluorescence has redistributed at ning and the end of the unloading period(Figure5F).

different times after the bleach(Figure6B).The mobility Hence,during unloading,labeled vesicles redistribute of vesicles in the terminal can be assessed by comparing rapidly and are drained from the entire terminal,leaving

the recovery of fluorescence in the bleached region to no local areas devoid of vesicles.the fluorescence across the whole terminal at each time To quantify vesicle mobility in cone terminals,we used

point.The fluorescence recovery can be fitted to a biex-Fluorescence Recovery After Photobleaching(FRAP)ponential function with time constants of?2s and9s (Figure6)in FM1-43-loaded terminals where vesicle cy-

(Figure6C).However,recovery is incomplete,such that cling had been stopped with0Ca2?/https://www.360docs.net/doc/8411067227.html,ing a con-even after recovery reaches steady state(25s)the ratio focal scanning microscope,a small area(2?m2)in the

of fluorescence in the bleached to unbleached regions center of a cone terminal was scanned with an Argon is87%?2%(n?10).This suggests that about87% laser(488nm)at full power for about1.5s.This bleached

of vesicles are sufficiently mobile to exchange between

Neuron

760

Figure5.Vesicles Disperse Rapidly during FM1-43Loading and Remain Evenly Dispersed during Unloading

(A and B)Vesicles disperse evenly and rapidly during loading.FM1-43fluorescence in terminals following a2min(A)or15min(B)load with AM1-43.Slices were fixed with glutaraldehyde immediately after loading to prevent unloading during image acquisition.

(C)Very similar fluorescence profiles across the terminals loaded for2and15min.Average fluorescence,F,was normalized to peak average fluorescence,F peak.Analysis regions are indicated in(A)and(B)(red rectangles).

(D and E)Vesicles remain evenly dispersed during unloading.Retina was preloaded with FM1-43for15min and then unloaded in high K?solution for2(D)or15min(E).

(F)Very similar fluorescence profiles across the terminals unloaded for2and15min.

Scale bars,1?m.

the unbleached and bleached region(the mobile frac-applying a Ca2?ionophore.Adding5?M ionomycin in tion).This stands in stark contrast to the neuromuscular

conjunction with normal(1.5mM)external Ca2?in-junction,where FRAP measurements have shown that creased internal Ca2?from?50nM to?0.5?M within little recovery occurs even1hr after FM1-43photo-

2min(Rea et al.,unpublished data),but had no effect bleaching,indicating that vesicles are immobile(Henkel on the mobile fraction of vesicles(n?27)(Figure6D). et al.,1996b).

Neither adding1?M forskolin to activate adenylate cy-The position of the bleached region within the terminal clase and trigger cAMP production(n?26)nor adding

50?M8-Br-cAMP(n?13)affected vesicle mobility. did not substantially affect the rate or extent of FRAP.

This applied even to areas at the base of the terminal Hence,vesicle mobility does not appear to be regulated that should contain synaptic ribbons.However,differ-

by intracellular Ca2?or cAMP.Treatment with1?M ences in the mobility of free vesicles and those residing okadaic acid(OA),a protein phosphatase inhibitor,de-on ribbons may be difficult to resolve considering that

creased the percentage of mobile vesicles to74%?our EM evidence shows that less than1%of vesicles11%(n?8),slightly less than control(p?0.1).Okadaic reside on ribbons.

acid slows the spread of FM1-43fluorescence in bipolar Further analysis of the FRAP results provides an esti-cell terminals but has no effect on the kinetics of release

(Guatimosim et al.,2002).Together with these results, mate of the diffusion coefficient(D)for mobile synaptic

vesicles.To determine the rate of vesicle movement,our findings suggest that protein phosphorylation may we fitted the FRAP time course to the solution of the

play some role in regulating vesicle mobility,but it is diffusion equation based on a fit of the photobleach unlikely to have a large impact on vesicles cycling at profile with a Fourier expansion(Berk et al.,1993).This

ribbon synapses.

solution takes the form of an infinite series of exponen-

tials,but the first two terms are sufficient to describe Endocytosed Vesicles Retain Their Integrity

the FRAP time course.Hence,FRAP data were fitted

Throughout the Vesicle Cycle

with a double exponential function.By using the time To determine whether endocytosed vesicles retain their constant for the slowest component(9s),which was

identity or whether they fuse with other vesicles or endo-best resolved with our data acquisition,it was possible somes in cone terminals,we compared simultaneous

versus sequential loading of FM1-43and FM4-64.The to estimate D for synaptic vesicles within photoreceptor

terminals.This yielded a value of1.1?10?9cm2/s.Be-absorbance spectrum of FM4-64overlaps with the emis-cause D?(distance)2/6(time)for diffusion in three di-

sion spectrum of FM1-43,so when they are mixed mensions,a synaptic vesicle will on average move1?m together,FM4-64can quench FM1-43fluorescence in1.5s.

(Rouze and Schwartz,1998).To test whether vesicular Is the mobility of vesicles modulated by intracellular membranes intermingle during the vesicle cycle,we messengers that regulate the release rate?We tested

used the four protocols illustrated in Figure7.First,cone whether elevated intracellular Ca2?influences FRAP by terminals were loaded for15min in the dark with45?M

Streamlined Vesicle Cycle in Cone Photoreceptors

761

Figure6.FRAP Experiments Reveal Rapid Vesicle Mobility

(A)Series of images taken from a single terminal in a slice preloaded with FM1-43,showing fluorescence before and after photobleaching of a2?m2region(white square).Confocal section thickness was1.7?m.Scale bar,1?m.

(B)Profiles of average fluorescence,F,across the terminal.The white dotted rectangles in(A)indicate the profile region,and the shaded rectangle indicates the bleached region.

(C)Time course of FRAP for terminals in Ca2?-free saline.Photobleaching at time?0s.Average fluorescence in the bleached region was divided by average fluorescence over the whole terminal for each image.These ratios were then normalized to the ratio before the bleach (time??2s).Continuous line shows fit to a double exponential function with time constants of1.4and9s.

(D)The size of the mobile fraction of vesicles is unaffected by intracellular messengers.The mobile fraction was estimated from the percentage of FRAP measured at25s after bleach in slices incubated for30min with0Ca2?saline alone or saline also containing50?M8-Br cAMP,1?M forskolin,5?M ionomycin(with1.5mM Ca2?for15min),or1?M OA.Data are from three to five experiments in each case.

FM1-43alone(protocol1).This served as a reference To control for the possibility that the extra FM4-64added for the next experiment,in which the same concentration

in protocol2reduced FM1-43fluorescence not by of FM1-43was loaded for15min,but this time in the quenching but by saturating the uptake of the dye,we presence of90?M FM4-64(protocol2).Since FM1-43

examined terminals loaded with135?M FM1-43alone and FM4-64should label the same vesicles,protocol2for15min.The fluorescence signal in these terminals should establish the maximal amount of quenching

was1.4?0.02times larger than terminals loaded with when FM1-43and FM4-64are maximally intermixed in45?M FM1-43(n?6),a significant increase(p?0.01). the same internalized membranes.Third,we loaded45

Hence,dye uptake was not saturating in these experi-?M FM1-43for15min and then added90?M FM4-64ments.

for5min(protocol3).If the vesicles loaded with FM1-

Figure7(bottom)compares the results of protocols 43coalesce with vesicles containing FM4-64,there3and4.Loading sequentially with FM1-43for15min should be some quenching.Finally,FM1-43was loaded

and then FM4-64for5min results in a FM1-43produced for15min,followed by5min with neither dye(protocol fluorescence that was not significantly different(86%?4).This controlled for the unloading of FM1-43that

15%)from that obtained by loading with FM1-43for would occur during the5min of loading FM4-64in proto-15min followed by5min unloading(n?3).Hence, col3.

sequential loading does not allow quenching of FM1-43 Figure7(top)compares the results of protocols1and by FM4-64,suggesting that vesicles loaded separately 2.Simultaneous loading with FM1-43and FM4-64for

with one dye or the other do not intermix.

15min results in a69%?4%reduction in the FM1-43It is possible that,rather than fusing with an internal fluorescence(n?4),suggesting substantial quenching.

endosome after fission from the plasma membrane,

Neuron

762

Figure7.Vesicles Retain Their Integrity

Throughout the Vesicle Cycle

Retinas were loaded with four different proto-

cols.Protocol1(independent loading):FM1-

43was loaded for15min in the presence of

Ca2?and then transferred to Ca2?-free solu-

tion for imaging.Protocol2(simultaneous

loading):FM1-43and FM4-64were mixed to-

gether and then loaded for15min in the pres-

ence of Ca2?.Orange color indicates mixing

of the two dyes in the vesicle membrane.Pro-

tocol3(sequential loading):retina were

loaded with FM1-43for15min,washed in

Ca2?-free saline,and then loaded further with

FM4-64for5min.Diagram indicates the pos-

sibility that FM1-43and FM4-64may mix,if

vesicles individually containing one dye fuse

with endosomes,eventually producing vesi-

cles containing both dyes(orange).Note that

some FM1-43vesicles unload during the sec-

ond loading step.Protocol4:FM1-43was

loaded for15min and then unloaded in Ca2?

for5min to estimate the extent of FM1-43

unloading in the second step of protocol3.

Bar charts show the average FM1-43fluores-

cence,F,measured from terminals under

each condition,normalized to the fluores-

cence of terminals loaded by Protocol1,F1.

large vesicles are endocytosed and then subdivide to lizard cones,even when they have been strongly depo-form smaller synaptic vesicles.Evidence for endocyto-

larized with Ca2?/high K?saline.

sis of ferritin into large structures in retinal bipolar cells

has been obtained from EM observations(Paillart et Discussion

al.,2003)and using a tetrarhodamine-labeled40kDa

dextran(TAMRA-40K-DEX)which is also preferentially Activity-Dependent Loading of FM1-43

into Cone Terminals

taken up into large vacuoles(Holt et al.,2003).EM stud-

ies on chick photoreceptors also show that extracellular FM1-43is an important tool for studying presynaptic horseradish peroxidase tracer appears in large vacuoles

function and membrane trafficking in neurons and other in some photoreceptor terminals in a dark-dependent cell types(for review,see Cochilla et al.,1999;Ryan, manner(Cooper and McLaughlin,1983).To test whether

2001).It has been particularly useful for exploring synap-large vacuoles form in lizard cone terminals,we incu-tic function in ribbon synapses of retinal bipolar cells bated slices in Ca2?/high K?saline with TAMRA-40K-

(Lagnado et al.,1996;Rouze and Schwartz,1998;Mack DEX for15min(n?2)or30min(n?2).Subsequent et al.,2000;Zenisek et al.,2002;Guatimosim et al.,2002). examination of cone terminals with confocal microscopy

However,Ca2?-independent extrasynaptic labeling has showed TAMRA-40K-DEX internalization in26%(15precluded its use in dissociated photoreceptors.We

have shown that simply by applying the dye while the min)and18%(30min)of terminals as compared

to?95%when terminals were loaded with FM1-43.Ter-retina is still intact and then washing with Advasep-7, minals that did take up TAMRA-40K-DEX showed dif-

photoreceptor terminals are loaded selectively without fuse rather than punctate labeling,suggesting uptake labeling neighboring structures such as horizontal and into relatively small structures.Examination of EM im-

bipolar cell terminals.Loading and unloading are Ca2?ages also failed to show large vacuoles in any of20dependent,and electron microscopy shows that the dye cone terminals examined.Therefore,it is unlikely that

is located specifically in synaptic vesicles.Thus,even bulk membrane retrieval is a part of the vesicle cycle in though FM1-43can enter cells through routes other than

Streamlined Vesicle Cycle in Cone Photoreceptors

763

endocytosis(Rouze and Schwartz,1998;Meyers et al.,

2003),dye uptake into cone terminals in intact retina

occurs primarily through endocytosis of surface mem-

brane.

The synapse-specific loading of FM1-43has enabled

us to explore the vesicle cycle in cones.Moreover,activ-

ity-dependent loading and unloading opens the oppor-

tunity to explore how light and other stimuli regulate

transmitter release.Most of what we know about how

release is regulated comes indirectly from recordings

of postsynaptic responses from retinal bipolar and hori-

zontal cells.However,these responses are determined

not only by the rate of photoreceptor exocytosis but also

by diffusion in the synaptic cleft,the rate of glutamate

reuptake(Gaal et al.,1998),the properties of postsynap-

Figure8.The Steps in the Vesicle Cycle of Cones

tic glutamate receptors,and the effects of electrical

(1)Fusion of synaptic vesicles with the plasma membrane;(2a) and/or chemical feedback from HCs onto cone terminals retrieval by endocytosis of small vesicles;(3)dispersal of vesicles (Kamermans and Spekreijse,1999).It was established throughout the cone terminal;(4a)binding to a synaptic ribbon; long ago that cone terminals accumulate markers,such

(5)movement along the ribbon,priming,and docking to plasma

membrane;(1)fusion of synaptic vesicles with the plasma mem-as horseradish peroxidase,in small vesicles in the dark

brane.Our evidence suggests that step2b(bulk membrane re-(Ripps et al.,1976;Schacher et al.,1976;Schaeffer and

trieval),step4b(immobilization of vesicles by tethering to the cy-Raviola,1978).However,these experiments required

toskeleton),and step4c(fusion with an intracellular endosome)are long incubation times(hours)and labeling could not be not usual parts of the vesicle cycle.

observed in real time,but only after permeabilizing the

tissue.FM1-43allows direct measurement of how light

affects cone exocytosis in real time(Rea et al.,unpub-The vesicle cycle of cones does not seem to involve lished data).Moreover,measurements can be made not bulk membrane retrieval or an endosomal intermediate only on isolated cells but also on arrays of cone terminals step.When endocytosed vesicles are retrieved from the in retinal slices or retinal flat mounts(Rea et al.,unpub-plasma membrane,they must recoup important lipid lished data),allowing release to be evaluated as it occurs and protein constituents needed for the next round of in the intact system.Even though the OPL contains exocytosis.By avoiding the mixing of membrane con-horizontal,bipolar,and Mu¨ller glial cells along with pho-stituents that could occur in vacuoles formed by bulk toreceptor terminals,the only structures exhibiting con-membrane retrieval or by fusing vesicles with an endo-siderable membrane cycling in the dark,and therefore somal intermediate,cone terminals avoid the extra step FM1-43loading,are rod and cone terminals.Since anole of sorting and parsing out of these constituents as new lizards possess cones but no rods,we have discovered vesicles are born.Perhaps conventional synapses with a way to understand synaptic information transfer in a a large reserve pool can afford to side track at least defined synaptic layer of the CNS from a homogenous some of their vesicles through an endosomal intermedi-population of cells.ate(Heuser and Reese,1973;Ryan et al.,1996;Richards

et al.,2000).However,even at these synapses there is The Vesicle Cycle in Cones

evidence that many cycling vesicles retain their integrity Figure8shows our proposed sequence of events in the and are reused(Ceccarelli et al.,1973;Murthy and Ste-

vens,1998)and that endosomes may only be important vesicle cycle.After fusion of synaptic vesicles with the

membrane and release of the neurotransmitter gluta-with intense stimulation(Richards et al.,2003).Most mate(1),membrane is retrieved by endocytosis of small

conventional synapses are remote from neuronal so-vesicles(2a).Unlike other synapses,endocytosis does mata.These synapses might use an intermediate endo-not appear to involve bulk membrane retrieval(2b).En-

somal step in the vesicle cycle as a quality control mech-docytosed vesicles diffuse from their site of origin anism for extensively reused synaptic vesicle proteins. throughout the terminal(3)and eventually encounter

Perhaps the relatively close proximity of cone terminals and bind to a synaptic ribbon(4a).Steps4b and4c to inner segments makes such a step unnecessary. illustrate events that are common in many synapses but

Our diagram of the vesicle cycle does not include are,at most,of minor significance to the vesicle cycle much shorter cycles that might occur,including multiple of cones.Only a small proportion of vesicles become

rounds of kiss-and-run exocytosis,the shortest vesicle tethered to the cytoskeleton(4b),and there is no evi-cycle possible.Recent studies on hippocampal syn-dence for vesicles fusing and intermingling their mem-

apses in culture suggest that many fusion events are of branes with other vesicles in endosomes(4c).In con-the kiss-and-run variety(Aravanis et al.,2003;Gandhi trast,the high mobility of synaptic vesicles is important

and Stevens,2003),characterized by repeated transient for increasing the likelihood that vesicles will encounter and incomplete fusion of vesicles with the plasma mem-and bind to a synaptic ribbon(4a).Vesicles then move

brane.FM1-43is,at best,incompletely released during along the synaptic ribbon,are primed for release,and these events,because fusion pore openings are too dock to the plasma membrane(5).Elevated intracellular

brief to allow rapid release of the dye,which dissociates Ca2?triggers fusion of docked synaptic vesicles(1),slowly from membranes.However,FM2-10is more hy-completing the cycle.

drophilic,dissociates more rapidly,and therefore is fully

Neuron

764

released even during kiss-and-run exocytosis(Klingauf have no effect on mobility.Okadaic acid,a protein phos-

et al.,1998).We find no difference in the unloading rate

phatase inhibitor,slightly decreases vesicle mobility in

of FM1-43and FM2-10during application of high K?/cones,suggesting that there might be some other phos-

Ca2?saline(Rea et al.,unpublished data).Therefore,the

phoprotein that can regulate interactions between syn-

vesicle cycling revealed by FM1-43may account for aptic vesicles and other structures in the cytoplasm.

nearly all of the transmitter release from the cone termi-

However,it appears that the mobility of vesicles and

nal,without involving kiss-and-run exocytosis.presumably their delivery to synaptic ribbons is largely

unregulated in cones.The average time required for a

synaptic vesicle to traverse a6?m diameter cone termi-Synaptic Vesicle Mobility

nal is seven times faster than the time constant of re-FM1-43-containing vesicles disperse evenly through

lease(48s versus345s),indicating that vesicle mobility cone terminals within2min after uptake.Even many

is not rate limiting for release.Therefore,subsequent minutes after dye uptake,vesicles remain mobile,sug-

steps in the vesicle cycle,for example,movement along gesting that most vesicles do not become tethered to

the ribbon,docking to the plasma membrane,or fusion the cytoskeleton.The percentage of vesicles that are

itself,must be rate limiting for continuous release.More-mobile and the diffusion rate of mobile vesicles are both

over,small changes in the size of the mobile fraction of unaffected by elevating Ca2?.This is in contrast to neu-

vesicles or in the rate of their movement should have romuscular and hippocampal synapses,where the mo-

little effect on release from cones.

bility of vesicles is highly restricted both before and

Most synaptic terminals,including those of retinal bi-during stimulation(Henkel et al.,1996b;Kraszewski et

polar cells(Lagnado et al.,1996;von Gersdorff et al., al.,1996).We have calculated the diffusion coefficient

1996),contain a large reserve pool of synaptic vesicles of synaptic vesicles in cones,which we measure as

that are relatively immobile and cannot easily be re-having an average diameter of about50nm(unpublished

leased.Strong and prolonged stimulation depletes the data).Assuming that the cytoplasmic viscosity and den-

readily-releasable pool,resulting in depression of the sity of the cytoskeletal matrix in cone terminals are simi-

rate of vesicle fusion and therefore synaptic transmis-lar to those of fibroblasts,the measured diffusion coeffi-

sion.With intense stimulation at phasic synapses,vesi-cient of synaptic vesicles(0.11?m2/s)is similar to the

cles are mobilized from the reserve pool to help replen-rate that an inert particle of similar diameter(50nm)

ish the readily-releasable pool(Hilfiker et al.,1999; would diffuse passively(0.15?m2/s)(Luby-Phelps et al.,

Zucker and Regehr,2002).In contrast,our results sug-1987).This suggests that synaptic vesicles in cones

gest that there is almost no reserve pool in cones.About diffuse without binding to structures that could hinder

87%of labeled vesicles remain mobile1hr after dye their movement.Consistent with this,synaptic vesicles

loading,the majority of vesicles in loaded terminals ap-in cones appear to move randomly with no particular

pear electron dense after photoconversion,and most directionality,for example,toward release sites at the

of the loaded dye can be released upon depolarization base of the terminal.

with high K?.

In bipolar cells,indirect measurements have sug-

The lack of a reserve pool and the huge number of gested that peripheral vesicles are mobile,but central

highly mobile vesicles insures that synaptic ribbons in vesicles are not(Lagnado et al.,1996).However,a recent

cones remain fully charged with synaptic vesicles.This study using more direct measures of mobility suggests

limits the extent of synaptic depression,even in the that most vesicles throughout the bipolar terminal are

face of a high tonic release rate in the dark.Intense mobile and that this is also unaffected by raising intracel-

stimulation of cochlear hair cells,which also have tonic lular Ca2?(Holt et al.,2004).These features of vesicle

synapses,depletes the number of docked vesicles and mobility are qualitatively similar to those in cone termi-

vesicles in the cytoplasm,but to a lesser extent the nals.Yet,vesicles move ten times more quickly in cone

number associated with ribbons(Lenzi et al.,2002).This terminals than in bipolar terminals.Thus,mobility seems

suggests that ribbons have a relatively high affinity for to correlate with the capacity for tonic release,with

vesicle binding.We propose that cytoplasmic vesicles conventional phasic terminals having immobile vesicles,

are kept highly mobile to help keep ribbons saturated tonic cone terminals having maximally mobile vesicles,

with vesicles.This should minimize fluctuations in the and phasic-tonic bipolar terminals having vesicles of

number of vesicles that are immediately available for intermediate mobility.

release,allowing changes in Ca2?-dependent exocy-At nearly all synapses,synaptic vesicles are tethered

tosis to accurately represent changes in illumination. to the cytoskeleton through the membrane-associated

protein synapsin,which binds to actin(Hilfiker et al.,

Experimental Procedures

1999).Intracellular Ca2?and cAMP,via CaMKII and PKA,

respectively,can trigger the phosphorylation of synap-

Retinal Preparation and FM Loading

sin,decreasing its affinity for actin.Synapsin phosphor-Anolis segrei lizards maintained on a12:12hr light:dark cycle were

ylation,and consequently mobilization of tethered vesi-used for all experiments.The care and use of these animals was

cles,is one of several Ca2?-and cAMP-mediated events

approved by the UC Berkeley Animal Care and Use Committee.

Eyes were removed following decapitation and pithing,and dark-that have been implicated in modulation of secretion

adapted on ice for30min in normal lizard saline(in mM:NaCl149; from presynaptic terminals.Remarkably,vertebrate

KCl4;CaCl21.5;MgCl21.5;HEPES10;glucose10;pH7.5)before rods and cones from nearly all species lack synapsin

removal of the retina.Enucleation was carried out in normal saline (Von Kreigstein et al.,1999;Morgans,2000).This may in the dark under infrared light,using IR viewers attached to a

account for the high mobility of synaptic vesicles in dissecting microscope.The retina was separated manually from the

cones and may also help explain why Ca2?and cAMP

retinal pigment epithelium(RPE)and loaded with dye(FM1-43,FM4-

Streamlined Vesicle Cycle in Cone Photoreceptors

765

64,or AM1-43)in normal saline for15min in the dark.Unless other-References

wise stated,loading was followed by washing in0Ca2?/EGTA solu-

tion(same as normal saline,but with1mM EGTA substituted for Aravanis,A.M.,Pyle,J.L.,and Tsien,R.W.(2003).Single synaptic 1.5mM Ca2?).FM or AM dyes were removed from external mem-

vesicles fusing transiently and successively without loss of identity. branes by washing with1mM Advasep-7(Kay et al.,1999)for15Nature423,643–647.

min.About1/3of the retina was cut from a position lateral to the Ashmore,J.F.,and Copenhagen,D.R.(1983).An analysis of trans-optic nerve and fovea and prepared for imaging.For retinal slices,mission from cones to hyperpolarizing bipolar cells in the retina of the piece of retina was flattened onto nitrocellulose paper with the the turtle.J.Physiol.340,569–597.

photoreceptor side uppermost and cut into300?m thick slices

Berk,D.A.,Yuan,F.,Leunig,M.,and Jain,R.K.(1993).Fluorescence using a tissue chopper.In experiments examining the distribution

photobleaching with spatial Fourier analysis:measurement of diffu-of dye at given time points of endocytosis,the tissue was fixed in

sion in light-scattering media.Biophys.J.65,2428–2436.

2.5%glutaraldehyde before slicing,within2min after dye loading.

Betz,W.J.,Mao,F.,and Bewick,G.S.(1992).Activity-dependent For dissociated cell preparations,the piece of retina was treated

fluorescent staining and destaining of living vertebrate motor nerve with10U/ml papain plus1mg/ml cysteine for15min,transferred

terminals.J.Neurosci.12,363–375.

to10mg/ml BSA for2min,and finally triturated gently five to ten

Brodin,L.,Low,P.,and Shupliakov,O.(2000).Sequential steps in times using a narrow glass pipette(0Ca2?/EGTA solution used for

clathrin-mediated synaptic vesicle endocytosis.Curr.Opin.Neuro-all three stages).A suspension of dissociated cones was then trans-

biol.10,312–320.

ferred onto a coverslip for imaging.

Ceccarelli,B.,Hurlbut,W.P.,and Mauro,A.(1973).Turnover of trans-

mitter and synaptic vesicles and the frog neuromuscular junction. Imaging

J.Cell Biol.57,499–524.

Images were obtained with a Zeiss Axioplan2upright microscope

Cochilla,A.J.,Angleson,J.K.,and Betz,W.J.(1999).Monitoring and Zeiss510confocal system equipped with a488nm argon laser.

secretory membrane with FM1–43fluorescence.Annu.Rev.Neu-Fluorescence was imaged with a63?achroplan,0.9NA water-

rosci.22,1–10.

immersion objective.Additional magnification,time series,and pho-

Cooper,N.G.,and McLaughlin,B.J.(1983).Tracer uptake by photo-tobleach protocols were controlled by Zeiss LSM acquisition soft-

receptor synaptic terminals.I.Dark-mediated effects.J.Ultrastruct. ware.For FRAP experiments,a2?m2region of the terminal was

Res.84,252–267.

continually scanned five times at maximum laser power(total pho-

tobleach time1.5s).For examination of dye localization within a Fuchs,P.A.,Glowatzki,E.,and Moser,T.(2003).The afferent syn-single terminal and for FRAP experiments,images were acquired apse of cochlear hair cells.Curr.Opin.Neurobiol.13,452–458. with a spatial resolution of33pixels per micron and optical section Gaal,L.,Roska,B.,Picaud,S.A.,Wu,S.M.,Marc,R.,and Werblin, thickness of1.7?m(?1Airy unit).Scion Image software(Scion F.S.(1998).Postsynaptic response kinetics are controlled by a gluta-Corporation,Frederick,MD)was used for image analysis.FM1-43mate transporter at cone photoreceptors.J.Neurophysiol.79, (or AM1-43)and FM4-64dyes were excited using the488nm laser190–196.

line,and emitted light was collected through a500-550BP or650LP

Gandhi,S.P.,and Stevens,C.F.(2003).Three modes of synaptic filter,respectively.In experiments examining FM4-64fluorescence

vesicular recycling revealed by single-vesicle imaging.Nature in double-loaded tissue,FM1-43bleedthrough into the650LP chan-

423,607–613.

nel was first assessed by imaging a sample loaded with FM1-43

Guatimosim,C.,Hull,C.,Von Gersdorff,H.,and Prado,M.A.(2002). alone,and the gain settings then reduced to below the threshold

Okadaic acid disrupts synaptic vesicle trafficking in a ribbon-type for FM1-43fluorescence.Variability among data is expressed as

synapse.J.Neurochem.82,1047–1057.

mean?SEM.Student’s t test was used to assess statistical signifi-

cance at p?0.05unless stated otherwise.Harata,N.,Ryan,T.A.,Smith,S.J.,Buchanan,J.,and Tsien,R.W.

(2001).Visualizing recycling synaptic vesicles in hippocampal neu-

rons by FM1–https://www.360docs.net/doc/8411067227.html,A98, Electron Microscopy12748–12753.

Slices were prepared from retina previously loaded with45?M

Henkel,A.W.,Lubke,J.,and Betz,W.J.(1996a).FM1–43dye ultra-AM1-43.Slices were then fixed in2.5%glutaralydehyde in0.1M

structural localization in and release from frog motor nerve terminals. phosphate buffer(PBS)overnight at4?C.After fixation,slices were

https://www.360docs.net/doc/8411067227.html,A93,1918–1923.

washed in100mM glycine solution for1hr to reduce glutaraldehyde

Henkel,A.W.,Simpson,L.L.,Ridge,R.M.,and Betz,W.J.(1996b). autofluorescence.Slices were then incubated in1mg/ml diamino-

Synaptic vesicle movements monitored by fluorescence recovery benzidine(DAB)solution(sonicated and filtered)in PBS at21?C for

after photobleaching in nerve terminals stained with FM1–43.J. 30min,before exchanging for fresh DAB solution.Photoconversion

Neurosci.16,3960–3967.

of DAB was achieved using a confocal microscope by focusing high

power,488nm light from an argon laser onto a60?m2region of Heuser,J.E.,and Reese,T.S.(1973).Evidence for recycling of synap-the slice to include a row of AM1-43-loaded terminals for8min.

tic vesicle membrane during transmitter release at the frog neuro-Following photoconversion,the slice was trimmed to enable location muscular junction.J.Cell Biol.57,315–344.

of the photoconverted region and prepared for electron microscopy Hilfiker,S.,Pieribone,V.A.,Czernik,A.J.,Kao,H.T.,Augustine,G.J., by dehydration and embedding as described previously(von Gers-and Greengard,P.(1999).Synapsins as regulators of neurotransmit-dorff et al.,1996).ter release.Philos.Trans.R.Soc.Lond.B Biol.Sci.354,269–279.

Holt,M.,Cooke,A.,Wu,M.M.,and Lagnado,L.(2003).Bulk mem-

brane retrieval in the synaptic terminal of retinal bipolar cells.J. Acknowledgments

Neurosci.23,1329–1339.

We thank Sue Yeon Choi for help with TAMRA-40K-DEX experi-

Holt,M.,Cooke,A.,Neef,A.,and Lagnado,L.(2004).High mobility ments and Fred Lanni and Diek Wheeler for assistance with FRAP of vesicles supports continuous exocytosis at a ribbon synapse. analysis.This work was supported by NIH grants to R.H.K.

Curr.Biol.14,173–183.

(EY12608),E.S.L.(NS32385),and P.S.(EY08124)and a Fight for Kamermans,M.,and Spekreijse,H.(1999).The feedback pathway Sight grant(to R.H.K.).E.S.L.was a Miller Visiting Professor at from horizontal cells to cones.A mini review with a look ahead. UC Berkeley.Vision Res.39,2449–2468.

Kay,A.R.,Alfonso,A.,Alford,S.,Cline,H.T.,Holgado,A.M.,Sak-

mann,B.,Snitsarev,V.A.,Stricker,T.P.,Takahashi,M.,and Wu, Received:September24,2003

L.G.(1999).Imaging synaptic activity in intact brain and slices with Revised:January23,2004

FM1–43in C.elegans,lamprey,and rat.Neuron24,809–817. Accepted:February10,2004

Published online:February12,2004Klingauf,J.,Kavalali,E.T.,and Tsien,R.W.(1998).Kinetics and regu-

Neuron

766

lation of fast endocytosis at hippocampal synapses.Nature394,Stevens,C.F.,and Williams,J.H.(2000).“Kiss and run”exocytosis at 581–585.

hippocampal https://www.360docs.net/doc/8411067227.html,A97,12828–12833.

Takei,K.,Mundigl,O.,Daniell,L.,and De Camilli,P.(1996).The Kraszewski,K.,Daniell,L.,Mundigl,O.,and Decamilli,P.(1996).

synaptic vesicle cycle:a single vesicle budding step involving Mobility of synaptic vesicles in nerve endings monitored by recovery

clathrin and dynamin.J.Cell Biol.133,1237–1250.

from photobleaching of synaptic vesicle-associated fluorescence.

J.Neurosci.16,5905–5913.Teng,H.,and Wilkinson,R.S.(2000).Clathrin-mediated endocytosis

near active zones in snake motor boutons.J.Neurosci.20,7986–Lagnado,L.,Gomis,A.,and Job,C.(1996).Continuous vesicle cy-

7993.

cling in the synaptic terminal of retinal bipolar cells.Neuron17,

957–967.von Gersdorff,H.,and Matthews,G.(1997).Depletion and replen-

ishment of vesicle pools at a ribbon-type synaptic terminal.J.Neu-Launay,P.,Fleig,A.,Perraud,A.L.,Scharenberg,A.M.,Penner,R.,

rosci.17,1919–1927.

and Kinet,J.P.(2002).TRPM4is a Ca2?-activated nonselective cat-

ion channel mediating cell membrane depolarization.Cell109,von Gersdorff,H.,and Matthews,G.(1999).Electrophysiology of 397–407.synaptic vesicle cycling.Annu.Rev.Physiol.61,725–752.

Lenzi,D.,Crum,J.,Ellisman,M.H.,and Roberts,W.M.(2002).Depo-von Gersdorff,H.,Vardi,E.,Matthews,G.,and Sterling,P.(1996). larization redistributes synaptic membrane and creates a gradient Evidence that vesicles on the synaptic ribbon of retinal bipolar neu-of vesicles on the synaptic body at a ribbon synapse.Neuron36,rons can be rapidly released.Neuron16,1221–1227.

649–659.Von Kriegstein,K.,Schmitz,F.,Link,E.,and Sudhof,T.C.(1999). Luby-Phelps,K.,Castle,P.E.,Taylor,D.L.,and Lanni,F.(1987).Distribution of synaptic vesicle proteins in the mammalian retina Hindered diffusion of inert tracer particles in the cytoplasm of mouse identifies obligatory and facultative components of ribbon synapses. https://www.360docs.net/doc/8411067227.html,A84,4910–4913.Eur.J.Neurosci.11,1335–1348.

Mack,A.F.,Behrens,U.D.,and Wagner,H.J.(2000).Inhibitory con-Zenisek,D.,Steyer,J.A.,Feldman,M.E.,and Almers,W.(2002).A trol of synaptic activity in goldfish Mb bipolar cell terminals visual-membrane marker leaves synaptic vesicles in milliseconds after ized by FM1–43.Vis.Neurosci.17,823–829.exocytosis in retinal bipolar cells.Neuron35,1085–1097. McNeil,P.L.,and Terasaki,M.(2001).Coping with the inevitable:how Zucker,R.S.,and Regehr,W.G.(2002).Short-term synaptic plastic-cells repair a torn surface membrane.Nat.Cell Biol.3,E124–E129.ity.Annu.Rev.Physiol.64,355–405.

Meyers,J.R.,MacDonald,R.B.,Duggan,A.,Lenzi,D.,Standaert,

D.G.,Corwin,J.T.,and Corey,D.P.(2003).Lighting up the senses:

FM1–43loading of sensory cells through nonselective ion channels.

J.Neurosci.23,4054–4065.

Morgans,C.W.(2000).Presynaptic proteins of ribbon synapses in

the retina.Microsc.Res.Tech.50,141–150.

Murthy,V.N.,and Stevens,C.F.(1998).Synaptic vesicles retain their

identity through the endocytic cycle.Nature392,497–501.

Paillart,C.,Li,J.,Matthews,G.,and Sterling,P.(2003).Endocytosis

and vesicle recycling at a ribbon synapse.J.Neurosci.23,4092–

4099.

Pyle,J.L.,Kavalali,E.T.,Piedras-Renteria,E.S.,and Tsien,R.W.

(2000).Rapid reuse of readily releasable pool vesicles at hippocam-

pal synapses.Neuron28,221–231.

Rao-Mirotznik,R.,Buchsbaum,G.,and Sterling,P.(1998).Transmit-

ter concentration at a three-dimensional synapse.J.Neurophysiol.

80,3163–3172.

Richards,D.A.,Guatimosim,C.,and Betz,W.J.(2000).Two endo-

cytic recycling routes selectively fill two vesicle pools in frog motor

nerve terminals.Neuron27,551–559.

Richards,D.A.,Guatimosim,C.,Rizzoli,S.O.,and Betz,W.J.(2003).

Synaptic vesicle pools at the frog neuromuscular junction.Neuron

39,529–541.

Rieke,F.,and Schwartz,E.A.(1996).Asynchronous transmitter re-

lease:control of exocytosis and endocytosis at the salamander rod

synapse.J.Physiol.493,1–8.

Ripps,H.,Shakib,M.,and MacDonald,E.D.(1976).Peroxidase up-

take by photoreceptor terminals of the skate retina.J.Cell Biol.

70,86–96.

Rouze,N.C.,and Schwartz,E.A.(1998).Continuous and transient

vesicle cycling at a ribbon synapse.J.Neurosci.18,8614–8624.

Ryan,T.A.(2001).Presynaptic imaging techniques.Curr.Opin.Neu-

robiol.11,544–549.

Ryan,T.A.,Smith,S.J.,and Reuter,H.(1996).The timing of synaptic

vesicle https://www.360docs.net/doc/8411067227.html,A93,5567–5571.

Schacher,S.,Holtzman,E.,and Hood,D.C.(1976).Synaptic activity

of frog retinal photoreceptors.A peroxidase uptake study.J.Cell

Biol.70,178–192.

Schaeffer,S.F.,and Raviola,E.(1978).Membrane recycling in the

cone cell endings of the turtle retina.J.Cell Biol.79,802–825.

Slepnev,V.I.,and De Camilli,P.(2000).Accessory factors in clathrin-

dependent synaptic vesicle endocytosis.Nat.Rev.Neurosci.1,

161–172.