presynaptic code. If we consider the spike sequence of a presynaptic neuron as the
representation of a stimulus, the PSPs produced at the synaptic level will be the
synaptic representation of that stimulus. The meaning of stimulus, however, does
not only refer to the codification of an environmental stimulation. The spike
sequences, in several neurons, are not only the codification of stimuli but participate also to the high-level performances connected to memory recall, thought,
reasoning, and so on. Whatever is the role of the spike sequence, it represents an
information which, transmitted to other neurons, is translated at the synaptic level
in a sequence of PSP. How this will be further recoded into a postsynaptic spike
sequence depends on a complex integration of all the inputs arriving to the neuron
in a compatible time window.
Although a large effort is spent in the last five decades for its understanding, the
way the neurons really code, manipulate, and share information remains a mystery.
What seems to be generally accepted is that the code of a neuron, for a given event,
is formed by a sequence of elementary bits (spikes) in a given time window. The
difficulty in understanding the code for a given stimulus rises because this sequence
often seems to be randomly distributed in time (irregular and non-repetitive
interspike intervals) also when generated for the same stimulus. So far, two main
ideas have been affirmed on the possible nature of the code, and both of them are
supported by many strong experimental evidences. According to one of them, the
codification of the stimulus occurs in terms of frequency of the spikes in a given
time window. Many different time sequences of the spikes can give the same
frequency since it depends on the number of spikes given in the chosen time
window. The alternative one assumes that the coding is embedded in the precise
timing of the spike occurrence.
The difficulty in understanding the relationship between the code generated by
neurons in sequences of spikes (either as frequency or precise timing) rises essentially by the lack of the precise knowledge on how the neuron generates spikes
thanks to the thousands of synaptic inputs it receives. In turn, this lack of knowledge depends on the still low level of knowledge on how the synapses code the
presynaptic information into a sequence of PSP. The understanding of the basic
mechanisms of synaptic transmission is fundamental in all fields of neurosciences
including the genesis of important brain diseases involving memory impairment
and other brain performances as Parkinson [1], Alzheimer [2], and Autism [3]. Not
surprisingly then a big effort is spent nowadays worldwide to study synaptic transmission with the most diverse experimental approaches but also with mathematical
modeling and computer simulations since, for the structural conformation, not all
the properties of the synapses can be unveiled by the experimental approaches.
In the present chapter, we will use the most common type of excitatory synapse
in the brain, the glutamatergic synapse, to outline, after a brief simple explanation
of its functioning, how many and how complex are the mechanisms controlling the
flow of information among the neurons operated by these synapses.
A typical pyramidal neuron of the cortex or of the hippocampus subfields receives
thousands of synaptic inputs (3000–30,000) [4–6]. The larger parts of these inputs
(80%) are excitatory inputs which use glutamate (Glu) as neurotransmitter. It is then
reasonable to assume that these synapses are the most important way of information
transfer and elaboration. Probably, the most important regulatory system of the
activity of the glutamatergic pyramidal neurons is given by the inhibitory neurons
which use the γ-aminobutyric acid (GABA) as neurotransmitter [4–6].
GABAergic synapses represent between 10 and 20% of the synapses inputting on
a pyramidal neuron, and they are located in strategic positions on the shaft of the
dendritic branches among the excitatory glutamatergic synapses [4–6].
Glutamatergic synapses are normally located on spines (a sort of elongation)
92
Advances in Neural Signal Processing
representation of a stimulus, the PSPs produced at the synaptic level will be the
synaptic representation of that stimulus. The meaning of stimulus, however, does
not only refer to the codification of an environmental stimulation. The spike
sequences, in several neurons, are not only the codification of stimuli but participate also to the high-level performances connected to memory recall, thought,
reasoning, and so on. Whatever is the role of the spike sequence, it represents an
information which, transmitted to other neurons, is translated at the synaptic level
in a sequence of PSP. How this will be further recoded into a postsynaptic spike
sequence depends on a complex integration of all the inputs arriving to the neuron
in a compatible time window.
Although a large effort is spent in the last five decades for its understanding, the
way the neurons really code, manipulate, and share information remains a mystery.
What seems to be generally accepted is that the code of a neuron, for a given event,
is formed by a sequence of elementary bits (spikes) in a given time window. The
difficulty in understanding the code for a given stimulus rises because this sequence
often seems to be randomly distributed in time (irregular and non-repetitive
interspike intervals) also when generated for the same stimulus. So far, two main
ideas have been affirmed on the possible nature of the code, and both of them are
supported by many strong experimental evidences. According to one of them, the
codification of the stimulus occurs in terms of frequency of the spikes in a given
time window. Many different time sequences of the spikes can give the same
frequency since it depends on the number of spikes given in the chosen time
window. The alternative one assumes that the coding is embedded in the precise
timing of the spike occurrence.
The difficulty in understanding the relationship between the code generated by
neurons in sequences of spikes (either as frequency or precise timing) rises essentially by the lack of the precise knowledge on how the neuron generates spikes
thanks to the thousands of synaptic inputs it receives. In turn, this lack of knowledge depends on the still low level of knowledge on how the synapses code the
presynaptic information into a sequence of PSP. The understanding of the basic
mechanisms of synaptic transmission is fundamental in all fields of neurosciences
including the genesis of important brain diseases involving memory impairment
and other brain performances as Parkinson [1], Alzheimer [2], and Autism [3]. Not
surprisingly then a big effort is spent nowadays worldwide to study synaptic transmission with the most diverse experimental approaches but also with mathematical
modeling and computer simulations since, for the structural conformation, not all
the properties of the synapses can be unveiled by the experimental approaches.
In the present chapter, we will use the most common type of excitatory synapse
in the brain, the glutamatergic synapse, to outline, after a brief simple explanation
of its functioning, how many and how complex are the mechanisms controlling the
flow of information among the neurons operated by these synapses.
A typical pyramidal neuron of the cortex or of the hippocampus subfields receives
thousands of synaptic inputs (3000–30,000) [4–6]. The larger parts of these inputs
(80%) are excitatory inputs which use glutamate (Glu) as neurotransmitter. It is then
reasonable to assume that these synapses are the most important way of information
transfer and elaboration. Probably, the most important regulatory system of the
activity of the glutamatergic pyramidal neurons is given by the inhibitory neurons
which use the γ-aminobutyric acid (GABA) as neurotransmitter [4–6].
GABAergic synapses represent between 10 and 20% of the synapses inputting on
a pyramidal neuron, and they are located in strategic positions on the shaft of the
dendritic branches among the excitatory glutamatergic synapses [4–6].
Glutamatergic synapses are normally located on spines (a sort of elongation)
92
Advances in Neural Signal Processing