Nmj preparation* Note multiple release sites *Frog most common for earlier studies, this image is actually from a mouse.

nmj preparation*Note multiple release sites

*Frog most common for earlier studies, this image is actually from a mouse

Katz & Miledi 1965Frog neuromuscular junction

No Calcium

Focal applicationof calcium

Focal applicationof a little less calcium

Back to no Calcium

postsynaptic response—release of neurotransmitter

presynaptic action potential

averaged responses

Calcium is required for exocytosis

The “calcium hypothesis”:

Ca2+ entry into the axon terminal rapidly triggers

exocytosisby binding to a “calcium

sensor” for release

Katz & Miledi 1967These experiments proved it

At the Nernst Potential*, the balance of electrical and diffusion tendencies creates an electrochemical equilibrium between the

opposing chemical (concentration) and electrical forces—

NO NET MOVEMENT OF IONS

At ~20, simplifies to:

Em = membrane potentialR = Gas ConstantT = Absolute Temperaturez = valance of ion (charge)F = Faraday’s Constant

*AKA Reversal Potential, Equilibrium Potential

V = I x R

I = g x (Vm – Veq)

The amount of current carried by a particular ion across a membrane is a function of how many channels that are permeable to that ion are open (the “conductance”, or “g”), and the “driving force” for that ion, that is, how far the membrane potential is from the equilibrium potential for that ion

1/R = g

ECa is very positive (>+100 mV), and gCa is voltage-dependent, so increasing membrane potential (depolarization) will open more and more calcium channels, increasing the conductance for calcium, but will also bring the membrane potential closer to ECa, reducing the driving force.

For the calcium current: ICa = gCa x (Vm – ECa)

driving force (decreasing)

calcium current (U-shaped)

g (increasing)

(membrane potential)

resting potential

Katz & Miledi 1967

“suppression” potentialis approached

note the growing “off response”

what is this?

what is this?

full suppression is achieved during pulse

Voltage-clamp recording of squid giant synapse

Llinas, 1982

Release depends on

[Ca]Ext

0.2 mM

0.25 mM

0.3 mM Dodge & Rahamimoff, 1967

more calcium = more release

“Cooperativity” of the calcium dependence of release

Dodge & Rahamimoff, 1967

multiple calcium ion must bind to trigger release

R ≈

[Ca]/KCa

1 + [Ca]/KCa + [Mg]/KMg

n

n is typically estimated between 2 - 4

NOT linear

How much calciuminside the cell is required for

release from mammalian CNS neurons?-giant presynaptic terminal (Calyx of Held) is filled with “caged” calcium

-a flash of light “uncages” the calcium; the brighter the flash, the more calcium is uncaged

-a fluorescent calcium-indicator dye reports the concentration of calcium in the terminal

EPSCs

-release of neurotransmitter is monitored as the postsynaptic response

Schneggenburger & Neher, 2000

Only a little bit of intracellular calcium (9-10 M) is required to trigger release

Schneggenburger & Neher, 2000

This is how much transmitteris released by an action potential

C C

CC

C

How fast is release?

Very Fastless than 60 sec delay between start of ICa and Ipost at 38C

Sabatini & Regehr, 1996

ICa

Ipost

time

room temp

fluorescent dyes

current recording

Calcium channels controlling release

There are many different subtypes of voltage-gated calcium channels in neurons, including:

L(onglasting)-, T(ransient)-, N(either)-, P(urkinje)-,

Q(cool letter, in the right region of the alphabet)-, and

R(esistant)-type channels

…and this doesn’t even include ligand-gated calcium channels!

How can we start to distinguish between them and identify their roles (if any) in controlling synaptic

transmission?

Voltage-dependent calcium channels differ in their activation/inactivation ranges and kinetics

(biophysical properties)

Miller, 1987

note differenttime courses

T-Type

L-Type

Calcium channels differ in their sensitivity to pharmacological

reagentsL- dihydropyridines (nifedipine, nimodipine, enhanced by BayK); rarely controls release

N- 1M Conotoxin fraction GVIA (CTx-GVIA)

T- 100 M Nickel (Ni2+); mostly localized to dendrites

Q- 1 M Agatoxin fraction IVA (Aga-IVA)

1.5 M Conotoxin fraction MVIIC (CTx-MVIIC)

R- 5 M CTx-MVIIC

P- 30 nM Aga-IVA

All are blocked by 10 mM cobalt (Co2+) and cadmium (Cd2+) works pretty well, too.

We can use these tools to identify the channels that control neurotransmitter release

Wheeler et al, 1994

(N-type) (Q- & R-)

(P-) (N-)

time (min)

Rel

ativ

e re

spo

nse

si

ze

Release at these hippocampal CA3-CA1 synapses is controlled by N-, and Q and(/or) R-type calcium channels; P-type channels do not control release at these synapses

How many calcium channels does it take to release a vesicle?

Mintz et al., 1995

release

no release

release

Ca channel vesicle

[Ca] “profile”

Mintz et al., 1995

more than100%

Non-additive effects of toxins on release suggests that multiple calcium channels must open to trigger vesicle fusion at mammalian CNS

synapses

(note-higher conc. of Aga-IVA used here blocks P- and Q-type channels)

Messages of the Day

• Intracellular calcium triggers release of neurotransmitter

• Calcium ions act cooperatively to trigger release (calcium-release relationship is not linear)

• Very small increases in intracellular calcium (~10M) trigger release very quickly (<100sec)

• Different types of voltage-dependent calcium channels control release of neurotransmitter

• Entry of calcium through more than one calcium channel may be required to trigger release

How does a vesicle fuse?

synaptic vesicle proteins

Sudhof, 1995

SNARES and Synaptotagmin

Littleton et al., 2001

-synaptosomal-associated protein of 25 kDa (SNAP-25)-syntaxin-vesicle-associated membrane protein (VAMP aka synaptobrevin)

*NEM-sensitive factor (NSF)

Soluble NSF*Attachment Receptors

{

(aka VAMP)

SNAREs were originally identified in yeast trafficking assays

T-SNARES:

V-SNARES: Golgi/PM vacuole ER/golgi

plasmamembrane

vacuole

Golgi/ER

Only some combinations lead to fusion—leading to the hypothesis that SNAREs impart specificity to fusion reactions—and may be the

“minimal machinery” required for vesicle fusion

McNew et al., 2000

Model of SNARE mediated fusion

Tetanus toxin and various serotypes of Botulinum toxin cleave SNARE proteins:

Syntaxin is cleaved by BoNT/CSNAP-25 is cleaved by BoNT/A and EVamp (synaptobrevin) is cleaved by TeNT,

BoNT/B, D, F and G

nb—tight form is insensitive to toxin

Effects of toxins on exocytosis (spontaneous EPSCs)

(syntaxin)(SNAP-25)(vamp)

Capogna et al., 1997

Broadie et al., 1995

TNT = Flies expressing TeTX throughout their nervous system to cleave VAMP (as with mice, some spontaneous release still present)

Sys = Fly syntaxin knock-out

Knocking out syntaxin abolishes all exocytosis

Geppert et al., 1994

Knocking out Synaptotagmin 1 (SytI) abolishes the fast Ca2+-triggered component of exocytosis, but not the slower Ca2+-

dependent “asynchronous” component of release

asynchronous release remains in SytI KO

asynchronous release

10 mM Ca2+ 10 mM Sr2+

Synchronous vs. asynchronous release

asynchronous release

synchronous release

Wild-type

SytI KO

Spontaneous mEPSCs are also unchanged in the SytI knock-out

1.9 minis/synapse/min

1.4 minis/synapse/min

Geppert et al., 1994

Synaptotagmin I is required for fast Ca-triggered synaptic vesicle

exocytosis, but not slower Ca-dependent asynchronous release or Ca-independent spontaneous release—could synaptotagmin I

be a fast calcium sensor?

• Each C2 domain binds 3 Ca2+ ions w/affinities of 60 M, 400 M and >1 mM

• C2A domain binds phospholipids in a Ca2+-dependent & cooperative manner

• C2A domain binds syntaxin in a Ca2+-dependent manner (EC50=250 M)

• C2B domain mediates self-association of synaptotagmin I into multimers

Synaptotagmin I: the calcium sensor?

Synaptotagmin I binds phospholipids (membranes) in the presence of calcium

Earles et al., 2001

GST Control

Synaptotagmin

EGTA

Ca2+

Syt radiolabeledvesicle

heavy beads coupled to Syt aremixed with radioactive lipid vesicles, then spun down in a centrifuge, bringingalong anything the Syt has bound to(unbound lipid vesicles stay in solution)

Ca2+ binding to synaptotagmin I C2A domain: an “electrostatic switch”

How could Synaptotagmin I control Ca-dependent

exocytosis?

= SNARE complex

Model I: Shao et al., 1997

Model II: Littleton et al., 2001

Either, neither, or both of these models could be correct

(Could this be a fusion pore? )

Messages of the Day

• Presynaptic SNAREs are the minimal machinery required for both calcium-dependent and spontaneous fusion

• Synaptotagmin is currently the best candidate for the fast Calcium Sensor triggering calcium-dependent synchronous release

• A 2nd Calcium Sensor controls asynchronous release

• The mechanisms underlying SNARE and synaptotagmin mediated fusion are hot topics of research—likely to be worked out within the next 5 years