Synaptic transmission: reliability and variability
7:30am-10:30am, 4:00pm-7:00pm,
Saturday Dec 7, 1996
Snowmass (Aspen), Colorado
Organizers
Background
Synaptic transmission is an essentially probabilistic process: when a
nerve terminal is depolarized, vesicular fusion and neurotransmitter
release do not always occur from each release site. At some
synapses the stochastic nature of this process may lead to 'unreliable'
transmission. For example, at some classes of synapses in the
mammalian CNS, release occurs with very low probability; in
hippocampal neurons, release probability averages less than about 0.4.
There are also other sources of variability, including the fluctuations
in the amplitude of the synaptic response when it occurs, and the
spontaneous release of neurotransmitter.
Until very recently, most neural network models assumed (either
explicitly or implicitly) reliable synaptic transmission and low
variability of single synaptic responses. This assumption is
particulary inappropriate in the mammalian CNS, where synaptic
transmission is neither reliable nor stationary. Further, synaptic
efficacy can be modified by many factors and on many time scales by a
large number of factors (eg, spike train history, neuromodulators).
We believe that a careful consideration of the stochastic nature of
synaptic transmission may lead to new insights into computational
principles.
Goals of the workshop
The primarly goal of this workshop is to explore the
consequences of the stochastic nature of synaptic transmission
for neural computation. We will encourage presentation of both
experimental and theoretical work, but our main perspective will be
neuroscientific. The target audience comprises researchers interested
in coding principles in the nervous system, as well as those
specifically interested in synaptic transmission.
Specific issues to be addressed in this workshop will include:
How do we quantify synaptic variability? Can we obtain
information theoretic measures of the reliability of synaptic
transmission from real synapses? An example of this approach is the
work of de Ruyter and Laughlin in non-spiking neurons in the fly
visual system.
Can the apparent stochasticity be circumvented by specific kinds
of spike trains? In other words is synaptic transmission truly
probabilistic for more 'realistic' spike sequences?
What are the consequences of postsynaptic response variability on
the fidelity of synaptic transmission?
Scheduled Presentations
1. Rob de Ruyter van Steveninck and Simon Laughlin.
NEC Research, Princeton, and Cambridge University, Cambridge.
Signal shaping and reliability in the blowfly photoreceptor-LMC
synapse: Possible implications for vesicle release.
2. Roberto Malinow.
Cold Spring Harbor Labs, Cold Spring Harbor.
Signal to noise: effect of synaptic plasticity.
3. Lynn Dobrunz and Chuck Stevens.
Salk Institute, San Diego.
Reliably unreliable synaptic transmission in the hippocampus.
4. Henry Markram and Misha Tsodyks.
Weizmann Institute, Rehovot, Israel.
Redistribution of synaptic efficacies and implications for
information transfer in cortical synapses.
5. Larry Abbott and Sacha Nelson.
Brandeis University, Waltham.
Synaptic depression enhances information transfer at visual
cortical synapses.
6. Ken Stratford.
Oxford University, Oxford.
Visual cortical synapses.
7. Karel Svoboda.
Bell Labs, Lucent Technologies, Murray Hill.
Seeing is believing: imaging synaptic function in vitro and in vivo.
If you need further information about NIPS*96 and the Workshops here are
the links: