February 25, 2010 15:25 WSPC/S0219-6352 179-JIN

S0219635209002319

Journal of Integrative Neuroscience, Vol. 8, No. 4 (2009) 453–469

c Imperial College Press
DOI: 10.1142/S0219635209002319

Research Report

PRINCIPLES OF CELLULAR-MOLECULAR
MECHANISMS UNDERLYING NEURON FUNCTIONS

ALEXANDER S. RATUSHNYAK∗ and TATIANA A. ZAPARA

Design Technological Institute of Digital Techniques SB RAS
Novosibirsk, 63009, Russia
∗
ratush@mail.ru

Received 27 October 2009

Revised 2 December 2009

In the present work, it was experimentally shown that a neuron in vitro was capable
of responding in a manner similar to habituation, Pavlov’s reflex and avoidance of the
reinforcements. The locality of plastic property modifications and molecular morphology,
as well as the connection between functional activity and cytoskeleton have been revealed.
A hypothesis is formulated that the neuron is a molecular system which may exercise
the control, forecast, recognition, and classification. The basic principles of the molecular
mechanisms of the responses underlying integrative activity, learning and memory at the
neuronal level are discussed.

Keywords: Neuron information functions; recognition; classification; forecasting; learning.

1. Introduction
For the past ten years, vast amount of experimental data have been accumulated
in neurobiological and cell molecular studies. The list of pathways of signal trans-
duction of molecules, enzymes and genes taking part in the information processes
is enormous. It allows understanding of the main principles of the organization and
function of molecular information devices — the neurons. However, unequivocal
answers to the main conceptual questions of bioinformatics have remained elusive
still. There is an opinion that such “analytic approach has been exhausted, we have
another and possibly more difficult problem. We literally sink in data. You can
connect all data in the world, but without the model these data will always be
insufficient” [23].
The main part of current neurobiological investigations is devoted to analysis
of the processes taking place in synapses during learning and memorization and
executed with the use of reduced neuronal systems as models, including those of
mollusks [1, 2, 5, 8, 10, 15, 16, 18, 22, 33]. Plasticity is a fundamental property
of the nervous systems. One of the best characterized forms of synaptic plasticity
is long-term potentiation (LTP) [7]. Currently, 1124 proteins have been identified

453
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

454 Ratushnyak & Zapara

in the synaptic terminal and only 466 of them were validated by detection in two
or more studies [9]. DNA microarray techniques allowed examination of the char-
acteristics of the activity-regulated genes (ARG) which alter their expression dur-
ing tetanus-induced LTP. The ARG-associated processes are complicated, including
signal transduction, transcription regulation, and modulation of synaptic structure
and functions. Structural changes at the synapse and new synapse formation are
thought to be critical for longer lasting LTP [25, 32]. A significant number of ARGs
are involved in neurite outgrowth and synapse formation. Many ARGs are impli-
cated in the cytoskeleton regulation and cell-to-cell or cell-to-extracellular matrix
interactions. Activity-regulated expression of these genes may directly contribute
to the synaptic modification or formation of the synapse, because some ARG pro-
teins which regulate cell–cell/extracellular matrix interactions and cytoskeletons are
known to be enriched at the synaptic regions [27]. ARGs also reveal the novel molecu-
lar processes underlying LTP. For example, CDC25B is an ARG that is a well-known
cell cycle regulator [13] but it plays a role in the regulation of synaptic plasticity as
well [27].
Unlike an analytical path there is a direction in which the study of a cell and
the brain as a whole is directed first of all to creation of imitation models with their
subsequent experimental verification mainly under the black box scheme. In this
case, for the most part the verifications are restrained by creation of mathematical
models of different complexity levels. Until recently, a neuron was thought as a
conductor of information and a simple threshold linear adder. Such ideas were used
as the basis for most of the neurocomputer metaphors. An attempt to bring the
neuron model nearer to the real object was undertaken in the Blue Brain project
[23]. The notion of neuronal cells as rather complicated molecular device is the
base of this project [14, 17, 20]. However, in this case undertaken attempts are
also reduced to creation of a cell model based not on conceptual principles but on
rather conflicting and inexhaustible experimental data. The results of such attempts
clearly demonstrate the problem of knowledge synthesis from the infinite body of
information while the conceptual model is absent. The big set of neurobiological
investigations of neurons were undertaken in the Blue Brain project to increase the
data volume, specify the data and take away the contradictions of known data. These
studies mainly touched on the function of ion channels and membrane mechanisms.
Using these data and supercomputer IBM, a model of neocortical columns containing
several tens of thousands of model neurons has been created. According to the
author’s opinion, the model under development is similar to the rat brain neurons
and is exhausted by the limiting/existing possibilities of computer systems. However,
it is not clear how fully this model reflects the real possibilities of simulated structure,
because the function of this structure of the brain is still not fully revealed. Thus
distinction of power consumption and speed on several orders from prototypes can
serve by the indirect evidence of considerable mismatches (not only “technological”
ones). Unfortunately the real informational neuron properties were not the focus of
study in this mega project. Works which showed that the simple neuronal structures
and separate cells are capable, in the experimental conditions, to generate enough
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

Principles of Cellular-Molecular Mechanisms in Neuron Functions 455

complicated responses were not practically considered [3, 8, 11, 16, 22, 24, 34]. It is
necessary to note that the data could not be used because of their incompleteness
and inconsistency.
The integration of the many known facts and the experimental results obtained
in the past years allow us to suggest that the neuronal cell has rather complicated
functions. However, on the conceptual level, the questions of the main informational
neuron properties are not been yet solved. There are no answers to the most concep-
tually important questions. Namely, what is a neuron? Are its functions limited in
the nervous system by the simple operations of signal processing and signaling? Is
it possible to selectively change the effectiveness of separate input of a neuron? Is it
possible to receive specific neuronal response after signaling from receptors to effec-
tors by the non-selective intracellular pathways? Without solving these questions,
the creation of models for both single cell system and higher level systems will only
lead to creation of simulations weakly resembling real neuronal systems.
Solving these problems would permit development of the conceptual models of
biological information systems. That in turn will bring closer possibility purpose-
fully to solve problems of medical correction pathological and cognitive changes in
the nervous systems, and with another to use principles (and also molecular mecha-
nisms) functioning of such biological prototypes of information systems in the field
of neurocomputers nano- and molecular electronics.

2. Methods and Experimental Model

In this study, we attempt to solve the above questions. It is necessary to carry
out direct investigations of isolated neurons with the aim to show the functional
and information possibilities of such cells. It was necessary to detect experimentally
a change of the weights for separate inputs and fixation of new functional mean-
ings by recording of spatial distribution of neuron activity. It is important on the
reduced system (the isolated neurons) to determine experimentally the properties of
processes of the functional plasticity and the participation of separate intracellular
systems in such processes.
To achieve these tasks, the selection of the experimental model is important.
The subject must be rather simple, but it must permit organization of several infor-
mation inputs to control and modify the molecular organization. The methodolog-
ical possibility of storage of the main properties is necessary. One of the subjects
which fits the majority of these requirements is isolated mollusk’s neurons culti-
vated outside of the organism. The large size of such cells permits organization of
rather simple system of informational inputs. The present work was carried out
using this model. The functional activity of isolated cultivated mollusk’s neurons
was recorded using the imaging, microelectrode and patch-clamp techniques. Sev-
eral types of the plasticity paradigms to be described below were developed. The
formation of external electric inputs and registration of ion channel functioning
(molecules forming the output signal) on the micro-areas was performed using the
glass coaxial micropipettes with 5 µm end diameter. Supplying the substances was
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

456 Ratushnyak & Zapara

Fig. 1. Isolated neuron of Lymnaea stagnalis mollusk with micropipettes positioned for simulating
the system of inputs and recording of the biological properties. The chosen areas for recording the
imaging signals are marked on the cell soma by number. The bottom part of the figure shows the
imaging signal from some chosen areas. On the left, the imaging signal from the chosen areas in
another time scale (the scale is 10 minutes is shown). The scale of image is 10 µm. The time-signal
scale is 10 seconds.

performed using the microcapillaries located inside these pipettes. Using these meth-
ods, several electric and chemical inputs were simulated. The cell response and ionic
currents were recorded.
Recording and comparison of ionic currents of stimulated and testing areas of
a cell were conducted at the various stages of formation of the neuron response.
The transformation of molecular morphology was estimated by redistribution of
the cell structures decorated by endogenous pigments to be characterized for the
chosen mollusk type. The mechanisms of interaction between various cells’ zones
were investigated by influencing on the sites of a cell, located on different distances.
For this, several micropipettes were put in the immediate vicinity, limited only by
the thickness of their wall (3–5 µm) (Fig. 1).

3. Results
3.1. Information characteristics of the isolated neurons
3.1.1. Experimental models of plastic responses
Many cellular analogs of learning paradigms to study the membrane mechanisms
of the plastic responses were developed using mollusk’s neurons [1, 5, 15, 18, 33].
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

Principles of Cellular-Molecular Mechanisms in Neuron Functions 457

However, these experimental models and methods did not reveal the answer to the
question: Are there any homogeneous modifications of conductance of the whole
membrane or these modifications can be local and specific for different zones of
somatic membrane? In particular, the question of whether whole neuron is modified
or the selective change of effectiveness of separate input is possible during training
remains to be elucidated? The goal of the present study included search for a model
which would allow investigation of the local changes of plastic properties of a somatic
membrane.
Plastic response modifications of isolated neurons occur by electric stimulation of
soma’s zones. The modifications of initial response, transition from action potential
(AP) to the local responses and transition from local response to AP were investi-
gated. It is known that type, dynamics and maintenance time of plastic modifica-
tions of the response depend on individual properties of a neuron and parameters
of applied influences (namely, intensity, sequence and frequency of stimuli).
In our experiments, among the possible variety of applied stimuli were picked
those which caused necessary changes of the response. The new neuron response had
to be preserved during the period, which would exceed membrane processes. Sev-
eral inputs onto the cell were performed. The stimuli were supplied to these inputs
according to classical rules of plastic response development. The influences which
increase or decrease the effectiveness of such responses to the signal supplied region-
ally to the selected inputs were used. The paradigms of unassociated, associated and
auto-reinforcement stimulation have been chosen.
The parameters of the stimuli for the paradigm of unassociated stimulation were
selected such that the stimuli evoked the AP of a neuron. The single extracellular
impulses of a current (0.1–0.5 nA) were supplied to one or several inputs with an
interval of 1–2 seconds. The time course of response changes and AP restoration
after the stimulation termination were recorded.
The parameters of the stimuli for the paradigm of associated stimulation were
selected such that the first in the pair of stimuli (applied to a micropipette) caused
the local response, namely, a change of the membrane potential (MP) on 1–3 mV.
The second stimulus was supplied on an intracellular electrode with some delay. Its
parameters (i.e., amplitude and duration) were selected so that the second stimulus
caused AP to every presentation. Thus, the second stimulus caused activation of
ionic channels of the whole cell.
The experimental model with autostimulation (biofeedback) has been chosen as
the third paradigm to produce the neuronal plastic response [28–30]. Such feedback
allowed correction of the stimuli presentation by means of the neuron output signal.
In the experiments performed on spontaneously active cells, the duration of neuronal
interspike (pulse-to-pulse) interval (ISI) was the operating parameter. Dynamics of
the ISI averaged values was estimated as the coefficient of reinforcement (Cr). Cr =
taisi /tcisi , where taisi is averaged interspike interval of background and tcisi is the
current interspike interval.
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

458 Ratushnyak & Zapara

3.1.2. Plasticity of somatic areas of isolated neurons under

local unassociated influences
At the first stage, two or more micropipettes were placed on a somatic membrane
and then the microelectrode was inserted into a cell for registration of intracellu-
lar activity. The stimuli parameters were selected such that initially every stimulus
generate identical number of AP. Hyper- and depolarization stimuli were used. The
multiple applications of the same extracellular non-periodic stimulus with stable
parameters and the interval accidentally changing from 3 to 15 seconds in many
cases evoked transition from the generation of spike responses to local responses.
The probability and time course for decreasing the spike responses depended on the
intensity of the initial response of a neuron. The time course of the reorganization
of the responses to stimulus causing generation of four to seven APs was studied
in detail. The multiple applications of local extracellular stimuli on the soma area
(limited by the tip of the micropipette) led to the gradual decrease of the number
of neuronal APs. The number of APs was reduced to one after the application of
30–50 stimuli. Then the first local responses appeared. At this stage (during 8–10
stimulus presentation), local responses were irregular and the periods of the APs
generation were observed. Starting with the 40–60th presentation of the same stim-
ulus, generation of APs completely stopped but APs can be evoked by modification
of stimulus parameters such as amplitude, duration and polarity. Such dynamics of
the neuron responsibility to the stimulus with stable parameters was developed in
86% of cases. In the remaining 14% of cases, the multiple applications of stimuli on
the neuron somatic areas did not lead to the appearance of local responses, although
reduction of AP number up to one or two was observed.
To define the time of maintenance of the neuron response modification, the stim-
uli with interval 0.5–1 minute was applied to the soma area. It were revealed that the
generation of a single AP in neuronal response was resumed after 1–2 minutes. The
restoration of the initial number of APs was observed after 2–5 minutes. In repeated
stimulation, the above mentioned stages of modification of neuronal responses could
be evoked by lesser number of stimuli presentations. In particular, a reduction of AP
number occurred after 8–10 stimuli, spike responses became irregular after 15–20
stimuli and only local responses were only observed after 20–25 stimuli.
Test stimuli were supplied to other areas of the membrane in three stages reduc-
tion the AP numbers in a response, appearance of the first new (local) responses
and stable generation of these responses. Such stimuli did not caused the change of
the resting potential (RP) to be more than background variations (1–2 mV).

3.1.3. Plasticity of the soma areas of isolated neurons under associated

local and general influences
The stimulation of different soma areas by means of local influences was carried
out in combination with intracellular stimulation by an algorithm similar to the
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

Principles of Cellular-Molecular Mechanisms in Neuron Functions 459

conditioning. The parameters of local extracellular stimulus were chosen such that
originally the stimulus applied though each of the micropipettes produces a local
response. The amplitude of local responses originally was not more than 1–3 mV
(RP of neurons was equal to −55–60 mV). The parameters of stimulus through
the intracellular electrode were chosen so that they produced one or two APs. The
paired stimuli were used. In each pair, the intracellular stimulus was applied after
the extracellular stimulus with interval 50–100 ms. This interval was determined in
each concrete case as a period to restore the neuron membrane potential up to RP
after AP. The paired stimuli were applied irregularly with 5–15 seconds interval.
In 69% of cases, the repeated application of local electric stimulus (coupled with
a threshold stimulus) was accompanied by alteration of the neuronal response. The
application of the first 15–20 associated stimuli lead to increase of the local response
amplitude. When the stimulation was continued, the neuron began to generate AP
on initially sub-threshold stimulus.
At this stage (from 20–30 stimuli) the spike responses alternated with the sub-
threshold ones (Fig. 2I, 2C). After 30–35 associated stimuli, the generation of the AP
in response to previously ineffective stimulus applied to the micropipettes became
regular (Fig. 2I, 2D). The stimulus provoked the generation of AP (in this case
response to the second stimulus of the pair did not changed). In our experiments,
in 14% of cases, such stimulation did not imply response reconstruction after appli-
cation of 40–60 coupled stimuli; in this case, further stimulation was ceased. In 17%
of cases, the response to the second stimulus of a pair was changed, in this case the
further stimulation was ceased. Single and associated stimulation did not changed
RP of neuron soma.
To define the time of maintenance of the neuron response modification, a single
stimuli (with interval of 0.5–1 minute) were applied to the soma area. It was revealed
that the response modification (to the stimulus with initially local response) was kept
for 5–7 minutes. While the application of the single stimuli has the same frequency
as in paired stimulation, the generation of AP was kept in response to 10–18 stim-
uli. Then the responses to the stimulus applied through the micropipettes became
local (as it was before using the pair stimulation) (Fig. 2III). At repeated paired
stimulation the change of response (generation of AP to the first stimulus) ensued
with the result of application of less number of associated stimuli (6–10).

3.1.4. Response plasticity of the isolated neurons under autostimulation

This experiment was carried out on the spontaneously active cells that generated
APs without stimulation. The interspike interval of such background activity was
varied as a rule from spike to spike. At the initial stage, the background current tcisi
was calculated and according to the obtained value, the reinforcement threshold was
determined. Two experimental protocols were used to form the external feedback
by autostimulation/reinforcement of a particular interval, namely long and short
intervals. In the first case, the stimulus was applied to the cell when interspike
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

460 Ratushnyak & Zapara

Fig. 2. Examples of plastic response modifications of the isolated neurons: I: responses to stimuli
(AP) under associative stimulation (A: at the beginning of the experiment, B: after 10 stimuli,
C: after 25 stimuli, D: after 50 stimuli, E: after withdrawal of reinforcing stimulus, F: at the
end of experiment); II: unassociative stimulation (A: at the beginning of the experiment, B: after
10 stimuli, C: after 25 stimuli, D: after 40 stimuli, E: after 50 stimuli, F: changing the stimulus
amplitude and duration); III and IV: the averaged time course of development and restoration of
responses under associative (III) and unassociative (IV) stimulation (↑: beginning of stimulation,
↓: end of stimulation). The ordinate axis corresponds to the percentage of responses to a stimulus.
The abscissa axis denotes the time.

interval duration was more than the predetermined threshold. In the second case,
the stimulus was applied in the time period equal to 5–10% of the value of tcisi
calculated before the stimulation. In such mode of stimulations, the reinforcement
of the long intervals switched the cell in the regime of more frequent generation of
AP (decrease of ISI, Fig. 3). At reinforcement of the short intervals, the value of ISI
was increased. In a number of the cases, the cell was transferred to the regime of
the generation of the burst consisting of 3–7 AP with a short ISI (less than 5% from
taisi ). Between the bursts, the ISI was higher than the predefined threshold. Such
differently directed activity modifications could be generated on the same cell (or on
the different areas of a neuron) many times during the experiment. The dynamics
of the modification of the reinforcement coefficient is given in Fig. 4.

3.1.5. Local plasticity of somatic membrane of isolated neurons

The local character of the response modification was determined by stimulation of
several soma areas of the same neuron. The same series of stimuli were applied to the
first and second micropipettes. If the stimulation of the first area caused the response
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

Principles of Cellular-Molecular Mechanisms in Neuron Functions 461

Fig. 3. The examples of the isolated neuron responses to auto-reinforcement of the intervals
(between the spontaneous action potentials) which are shorter than the threshold determined
according to the current averaged interspike interval in the background period (↑: the beginning of
stimulation). The ordinate axis corresponds to the value of interspike interval. The abscissa axis
denotes the time.

Fig. 4. The averaged temporal course of the response of the isolated neuron to auto-reinforcement
of intervals (between the spontaneous action potentials). RSI : reinforcement of the short intervals;
RLI : reinforcement of long intervals; ↑: the beginning of stimulation; ↓: the end of stimulation. The
ordinate axis denotes the value of the coefficient of reinforcement (Cr). Cr = taisi /tcisi , where taisi
represents the averaged interspike interval during full recording period and tcisi represents currrent
interspike interval. The abscissa axis corresponds to the time.

modifications then, after restoration of initial response to these stimuli, we began to

stimulate the area under the second micropipette. During the successive irritation of
two somatic areas, it was shown that multiple local influences on these areas could
have different effects on the neuron responses. The neurons whose responses were
changed under stimulation of two areas or only one chosen area were revealed. In
14% of cases, the multiple stimulations did not cause changes of initial responses to
one of the neuron soma areas. The stimulation through one of micropipettes did not
change the neuron responses to the stimulus applied to other membrane areas.

3.2. Biophysical properties of areas of local neuron plasticity

To investigate the possible cellular mechanisms of the phenomena described above
we studied the properties of cell’s areas limited by micropipettes using these areas
as model of information inputs.
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

462 Ratushnyak & Zapara

3.2.1. Change of the integral ionic currents of the areas of somatic membrane
of neuron during development of plastic responses
The total ionic currents flowing through the areas of the somatic membrane limited
by end of micropipette (5 µm) were registered. The stimuli series were accompa-
nied by alternate registration of currents flowing through the zones of the somatic
membrane limited by micropipettes. In this experiment, the current’s registration
in these areas was carried out at different stages, namely: (1) before stimulation,
(2) at the stage of alternation of spike and local responses, (3) at the stage of stable
responses when the changed responses were recorded to 10–15 of stimuli.
At the stage of alternation of spike and local responses, the changes of the input
and output currents were observed if response changes were recorded. The ampli-
tudes of the input currents were increased to 50–200% in comparison with those
registered before stimulation.
At the stage of the stable responses, the amplitude of input currents was
decreased in comparison with the currents to be registered at the stage of response
alteration. The output currents were moderately changed relatively to the initial
ones (the changes in 80% of cases were not significant). In other experiments the
amplitude of the output currents was decreased to 10–15%.
The ionic currents in the control areas were not changed. If the multiple stim-
ulations did not cause the response change then the currents were not changed in
those areas of the neuron somatic membrane.
The main changes of the integral ionic currents during local plastic modifica-
tions for the neurons placed in the solid solution without sodium ions consisted of
increase of the amplitude of the input currents; the output currents were decreased
insignificantly.

3.2.2. Optical density in the zone of the local influences producing

the plastic responses
During local stimulation of neurons producing the plastic responses, the change of
relative optical density in the zone of influence and the fluorescent label (for actin
and tubuline) were registered. The relative optical density was increased from 5–15%
to 250%. On the average, the change was equal to 44% (in 15 experiments). The
change of optical properties correlated in time with the change of neuron responses
and transmembrane ionic currents. The quick restoration (during 1–2 minutes) of
initial optical density took place after ceasing the stimulation in 33% of cases, in
54% of cases the decreasing of optical density was observed during 5–40 minutes.

3.3. The influence of selective modification of molecular neuron

structures on the plastic changes of responsiveness
3.3.1. The influence of microfilament modification on the neuron plastic properties
The influence on the neuronal microfilament system was performed by injection of
phalloidin and cytochalasin B into the micropipette solution. The monomeric and
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

Principles of Cellular-Molecular Mechanisms in Neuron Functions 463

polymerized (microfilaments) cellular actin were in the dynamic equilibrium state.

Phalloidin interacts with the microfilaments by blocking the reaction of depolymer-
ization and stabilizing the cellular microfilaments. Cytochalasin B interacts with
monomeric actin, blocking the reaction of polymerization and formation of new
microfilaments.
To investigate the influence of actin polymerization blockade on the plastic prop-
erties of neurons, cytochalasin was injected into the micropipettes solution. In these
experiments, the check electric stimulation was carried out to find two neuronal
areas with the determined plastic properties. Cytochalasin was injected into one
of the pipettes only. The injection of cytochalasin B did not influenced the plastic
responses development during the first 30–45 minutes. During the following stages
after the injection, the single stimulation did not evoke transformation of spike
responses into local ones in 65% of experiments (such effect was revealed in 7% of
the control experiments). In 27% of cases, local responses appeared, but the stage
of response alternation was maintained even after application of 50–100 stimuli.
The associative stimulation while using cytochalasin B evoked AP in 9% of cases.
The plastic properties of the check neuronal area were not changed. The neuronal
areas at which the local stimulation evoked the response change were influenced by
phalloidin. Local stimulation was carried out on those two areas. If the stimulation of
each area caused the response change, one of these areas was processed by phalloidin.
The substance was injected at the various stages of experiment, namely: after
the series of stimuli evoking the response change; after the simulation causing the
irregular responses; before stimulation and after spontaneous restoration of the ini-
tial response. In the first case, we began to inject phalloidin into a micropipette
immediately after the response change and continued stimulation (20–25 minutes).
Then to reveal the time of maintenance of the response change, the stimuli were
applied to this neuron area at 1–2 minutes interval. The restoration of the initial
response to the stimulus did not occur during the time (1–1.5 hours) of observation

Fig. 5. The time course of the plastic response development under associative stimulation after
injection of phalloidin. The results of four groups of experiments (Ph1–Ph4) are accumulated. →: the
beginning of substance application, ↑: the beginning of stimulation. The ordinate axis corresponds
to the percentage of responses (AP) averaged on the group of neurons. The abscissa axis denotes
the time.
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

464 Ratushnyak & Zapara

(Fig. 5, Ph1). In the control area, the time of response restoration did not changed.
If phalloidin was injected before the stimulation or after the restoration of the initial
responses and the second series of stimulations was made, then the response change
was not observed even after application of 100–150 stimuli (Fig. 5, Ph4). The pro-
cessing by phalloidin of the neuron areas completely blocked the development of all
plastic responses during the 1–1.5 hours. In the process of testing the effects of phal-
loidin it was revealed that the maintenance time of the new response was increased
if the stimulation was carried out not earlier than 45–60 minutes after the substance
injection into a micropipette. Reactions with maintenance time of more than 15
minutes were revealed most often. In this case, the cell response was not restored
during all period of observation (up to several hours). The averaged dynamics of
responses to the associative stimulation is given in Fig. 5, Ph1–Ph4.

3.3.2. Influence of microtubules modification on the plastic

properties of neuron
The influence on the microtubules system was performed by injection of colchicine
or taxol into the micropipette solution. The monomeric and polymerized forms
of tubulin protein (microtubules) were in the cell in dynamic equilibrium state.
Taxol interacts with microtubules and blocks the reaction of depolymerization, sta-
bilizing the polymeric form of tubulin in a cell. Colchicine in turn interacts with
monomeric tubulin blocking the reaction of polymerization and the formation of
new microtubules.
Colchicine was injected in one of the pipettes and it was impacted on devel-
opment of the plastic responses for 10–15 minutes. The single stimulation did not
evoke local responses or they developed only at the stage of response alteration. In
43% of experiments (in the case when the stimulus initially evoked 1–2 AP), the
plastic response to the single stimulation did not develop. In 57% of experiments,
dynamic change of the response development took place. The first local response
appeared only after 50–100 stimulations. In this case, response alteration occurred
even after 300–500 stimuli. Cessation of stimulation led to restoration of the initial
response for 20–30 seconds. During associated stimulation, the local responses did
not transformed in AP even after 50–100 stimuli. The plastic properties of the check
neuron area did not change.
Taxol was injected in the micropipette solution and its effects were detected
after 20–30 minutes. The dynamics of response change were close to the check ones,
independently from the type of stimulation (Fig. 6, T1). The restoration of initial
response happened more quickly in 40–60% of cases. The second session of stimula-
tion through the pipette with taxol showed that it was necessary to use more stimuli
to change the responses. In the second session, the number of stimuli to develop a
new response was increased by 20–25% (Fig. 6, T2). The time of maintaining the
new type of response did not change relatively to the first series. During the third
and following sessions, the number of stimuli necessary for response change was
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

Principles of Cellular-Molecular Mechanisms in Neuron Functions 465

Fig. 6. The time course of the plastic response development during unassociative stimulation after
taxol injection. T1–T4: the series of stimulation, ↑: the begin of stimulation. The ordinate axis: the
percentage of responses (AP) averaged over the group of neurons. The abscissa axis: the number of
stimulus.

being increased further (Fig. 6, T3–T4). The plastic properties of the check neu-
ron area did not change. In the repeated session of stimulation through the check
pipette, the response change took place as the result of using lesser number of stim-
ulations. The dynamics of responses, averaged over the neuron group is shown in
Fig. 6, T1–T4.

4. Discussion
Up to now there is relatively small number of the investigations aimed at the direct
experimental analysis of information neuronal functions. It was shown that simple
neuronal systems and single cells in vitro were capable of forming complex responses
[3, 8, 11, 16, 23, 25]. According to these data, a neuron may be considered as a
complicated molecular information system.
The neuron response modifications revealed in our study are phenomenologically
similar to the behavioral type of habituation and Pavlov’s reflex, but they are gener-
ated by a single cell. As the plastic modifications were developed in evolution as the
system level response of molecular and cellular complexes to the environment influ-
ences, it can then be suggested that the common basic mechanisms are characteristic
for both the behavior level of an organism and for neuronal elements [2, 8, 10, 22, 33].
Thus, the results obtained confirm the hypotheses that the isolated neuron is
a complex information controlling system. It was shown that every cellular input
was modified according to preceding learning. A significant number of electrical or
chemical inputs (the size of one cellular region to be activated in our experiments
were less 1/1000 of the total surface of a cell) can be organized on a cell. By the
influence on each of such local inputs we can change (increase or decrease) the weight
of active input. Our results may be considered as indirect confirmation of the idea
that the principles of the neuron functions are defined by their morphology [4, 6,
19, 34]. Some structural elements of such molecular “nanocomputer” are defined.
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

466 Ratushnyak & Zapara

Besides, the influences on these molecular systems permitting to accelerate and block
the weight changing for local inputs are found.
The compilation of the results obtained and known modern data in molecular
cellular organization allow us to formulate (heuristic to a great extent) a hypothe-
sis about the leading role of regrouping of neuronal receptors and effectors in the
process of learning. It was shown that the isolated neuron can control external stim-
ulation by changing output signals to minimize reinforcement. All these permit us
to consider the cell as a molecular information system capable of modifying its input
and output elements according to a definite program genetically determined but can
be influenced by environmental factors.
The mechanisms of such system operations have not been sufficiently investigated
yet. It is known that ionic channels and receptors are as a rule connected with the
cytoskeleton by the specialized protein domains. In the initial concept of signal trans-
mission it was suggested that all main components, namely, receptors, G-proteins
and their targets, protein kinases, and ionic channels are freely distributed, and
information transfer in the intercellular space takes place during the occasional col-
lisions and interactions of process partners [26]. A huge number of receptors for
various neurotransmitters, hormones and sensor stimuli interacting with the system
of the secondary mediators are known. Nevertheless, the neurons are capable of
distinguishing one stimulus from another within microseconds. A new concept of
signal transmission in the eukaryotes based on the results of previous investigations
appeared. It implies that the domains of protein–protein recognition allow physi-
cal interactions between molecules partners. Various components such as receptors,
enzymes and their substrates, and the final targets are structurally combined in the
functional complexes — the microdomains. The formation of microdomains takes
place on the structure-forming proteins, which due to their genetically conditioned
properties, can form ordered connections with receptors, channels, proteins, and can
interact with cytoskeleton and molecular movers [12].
Various protein–protein interactions permit formation of clusters of ionic chan-
nels and/or receptors on specific subcellular sites of neurons which locate near to
the proteins of intracellular signal cascades. Such structure-module protein organi-
zation (named as microdomains) permits the input signal to be concentrated and
combined with signal transduction and transport pathways of a cell [12, 21]. The
specialized receptors and ionic channels of the cellular plasmatic membrane have a
unique individually-arranged structure of interactions with their partners (Fig. 7).
This microstructure organization permits the receptors and ionic channels of
the plasmatic membrane to react differentially on the outward influences using the
same systems of the secondary mediators. The data about the mechanisms of the
neuronal (non-synaptic) plasticity obtained by us on the neurons of the mollusk —
evolutionally older subject than the mammals, allow us to consider the principles
of the response of the cell system to the changing environment and the role of
the cytoskeleton in the formation and maintenance of the plastic responses. Based
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

Principles of Cellular-Molecular Mechanisms in Neuron Functions 467

(A) (B)
Fig. 7. A possible scheme of the functional domain formation by rearrangement of the receptors
and effectors proteins and attaching them to the cytoskeleton. A: before formation of the domain,
B: after the structural rearrangements.

on literary evidence and our experimental data, it is possible to assume that the
basic element of a neuron while realization of its main information function (record
and storage of the information) is the structural microdomain formed during inter-
actions of a neuron with the environment and maintained by the systems of the
cellular structure stabilization. The set of such domains being formed by regroup-
ing during the leaning process permits the cell to recognize, classify and use the
multidimensional vectors of outward signals for control (on the basis of previous
experience).
We can suggest that the molecular information neuron system on the basis of
previous leaning performs positional coded fixation of the multidimensional vector. It
forms a memory matrix consisting of the microdomains. The matrix being formed in
such a manner can include all input signals (conditioned and unconditioned stimuli)
as elements because neuron has the responses to them. The recognition of the image
(even in the absence of some elements of a multidimensional vector) can lead to
“predicting” previously connected events. Such identification can be performed as
distribution dynamics (defined by the matrix) of a wave of excitation from one
to another structure-functional cellular element. The activation of effectors of cell
included in the vector of recognition can lead to the formation of response to avoid
or minimize the external influences.
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

468 Ratushnyak & Zapara

Thus the main function of a neuronal cell is recognition, classification, forecast

and control, similar to biological systems of higher level of organization [3, 31]. Using
the above conceptual neuron model and taking into account the neuron’s main
information properties when creating imitation systems may bring such systems
closer to their biological prototypes.

References
[1] Alkon DL, Changes of membrane currents during learning, J Exp Biol 112:95–112,
1984.
[2] Alkon DL, Rasmussen H, A spatial-temporal model of cell activation, Science 239:998–
1005, 1988.
[3] Anokhin PK, Sketches on Physiology of Functional Systems, M: Medicine, p. 444, 1975
(in Russian).
[4] Bailey CH, Kandel ER, Structural changes accompanying memory storage, Annu Rev
Physiol 55:397–426, 1993.
[5] Balaban PM, Korshunova TA, Bravarenko NI, Postsynaptic calcium contributes to rein-
forcement in a three-neuron network exhibiting associative plasticity, Eur J Neurosci
19:227–233, 2004.
[6] Beattie EC, Carroll RC, Yu X, Morishita W, Yasuda H, von Zastrow M, Malenka RC,
Regulation of AMPA receptor endocytosis by a signaling mechanism shared with LTD,
Nat Neurosci 3:1291–1300, 2000.
[7] Bliss TV, Lomo T, Long-lasting potentiation of synaptic transmission in the dentate
area of the anaesthetized rabbit following stimulation of the perforant path, J Physiol
(Lond) 232:331–356, 1973.
[8] Burrone J, O’Byrne M, Murthy VN, Multiple forms of synaptic plasticity triggered by
selective suppression of activity in individual neurons, Nature 420:414–418, 2002.
[9] Collins MO, Husi H, Yu L, Brandon JM, Anderson CN, Blackstock WP, Choudhary
JS, Grant SG, Molecular characterization and comparison of the components and mul-
tiprotein complexes in the postsynaptic proteome, J Neurochem 97:16–23, 2006.
[10] Crow T, Pavlovian conditioning of Hermissenda: Current cellular, molecular, and circuit
perspectives, Learn Mem 11:229–238, 2004.
[11] Daoudal G, Debanne D, Long-term plasticity of intrinsic excitability: Learning rules
and mechanisms, Learn Mem 10:456–465, 2003.
[12] Delmas P, Crest M, Brown DA, Functional organization of PLC signaling microdomains
in neurons, Trends Neurosci 27:41–47, 2004.
[13] Draetta G, Eckstein J, Cdc25 protein phosphatases in cell proliferation, Biochim Bio-
phys Acta 1332:53–63, 1997.
[14] Druckmann S, Banitt Y, Gideon A, Schurmann F, Markram H, Segev I, A novel multi-
ple objective optimization framework for constraining conductance-based neuron mod-
els by experimental data, Frontiers in Neurosci 1:7–18, 2007.
[15] D’iakonova TL, Two types of neurons differing in their plastic properties: Study of ionic
mechanisms, Zh Vyssh Nerv Deiat Im I P Pavlova 35:552–560, 1985 (in Russian).
[16] Grechenko TN, Zints P, Formation of conditioned responses in isolated neurons of Helix
pomatia, Zh Vyssh Nerv Deiat Im I P Pavlova 33:957–959, 1983 (in Russian).
 February 25, 2010 15:25 WSPC/S0219-6352 179-JIN
S0219635209002319

Principles of Cellular-Molecular Mechanisms in Neuron Functions 469

[17] Hines ML, Eichner H, Schürmann F, Neuron splitting in compute-bound parallel

network simulations enables runtime scaling with twice as many processors, J Comput
Neurosci 25:203–210, 2008.
[18] Kandel ER, Schwartz JH, Molecular biology of leaning: Modulation of transmitter
release, Science 218:433–442, 1982.
[19] Kim CH, Lisman JE, A role of actin filament in synaptic transmission and long-term
potentiation, J Neurosci 19:4314–4324, 1999.
[20] Kozloski J, Sfyrakis K, Hill S, Schurmann F, Peck C, Markram H, Identifying, tab-
ulating, and analyzing contacts between branched neuron morphologies, IBM J Res
Develop 52:43–55, 2008.
[21] Lamprecht R, LeDoux J, Structural plasticity and memory, Nature Rev Neurosci 5:
45–54, 2004.
[22] Lin XY, Glanzman DL, Effect of interstimulus interval on pairing-induced LTP of
Aplysia sensorimotor synapses in cell culture, J Neurophysiol 77:667–674, 1997.
[23] Markram H, The blue brain project, Nat Rev Neurosci 7:153–160, 2006.
[24] Malinow R, Malenka RC, AMPA receptor trafficking and synaptic plasticity, Annu Rev
Neurosci 25:103–125, 2002.
[25] Matsuzaki M, Honkura N, Ellis-Davies GC, Kasai H, Structural basis of long-term
potentiation in single dendritic spines, Nature 429:761–766, 2004.
[26] Orly J, Schramm M, Coupling of catecholamine receptor from one cell with adenylate
cyclase from another cell by cell fusion, Proc Natl Acad Sci USA 73:4410–4414, 1976.
[27] Park CS, Gong R, Stuart J, Tang S-J, Molecular network and chromosomal clustering of
genes involved in synaptic plasticity in the hippocampus, Biol Chem 281:30195–30211,
2006.
[28] Ratushniak AS, Voskresenskaia LV, Pankova TM, Shtark MB, Adaptive trace reactions
of hippocampal neurons in tissue culture, Dokl Akad Nauk SSSR 228:1479–1481, 1976
(in Russian).
[29] Vasilevskii NN, Trubachev VV, Suvorov NB, Vasivevskii NN, Changes in the impulse
activity of cortical neurons with selective reinforcement of a chosen diapason of their
interspike intervals, Neirofiziol 4:339–348, 1972 (in Russian).
[30] Vasilevskii NN, Suvorov NB, Trubachev VV, Stable changes in frequency and synchro-
nism of cortical neuronal population dischargs in experiments with feedback, Dokl Akad
Nauk SSSR 206:510–512, 1972 (in Russian).
[31] Vityaev EE, Extraction of Knowledge from Data. Computer Knowledge. Modelling Cog-
nitive Process. Novosibirsk, NSU, p. 293, 2006 (in Russian).
[32] Yuste R, Bonhoeffer T, Morphological changes in dendritic spines associated with long-
term synaptic plasticity, Annu Rev Neurosci 24:1071–1089, 2001.
[33] Zapara TA, Ratushnyak AS, Shtark MB, Local changes of transmembrane currents
at plastic reorganizations of electrogenesis of isolated neurones of the snail, Neurosci
Behav Physiol 19:140–145, 1989.
[34] Zapara TA, Simonova OG, Zharkikh AA, Ratushnyak AS, The effects of the dynamic
state of the cytoskeleton on neuronal plasticity, Neurosci Behav Physiol 30:347–355,
2000.