Download INTRAANALYZER CONDITIONED REFLEX PROPERTIES OF TWO

ACTA NEUROBIOL. EXP. 1981, 41: 543-552
Lecture delivered a t the Symposium "Brain a n d behavior"
held i n Jablonna near Warsaw
May 1981
INTRAANALYZER CONDITIONED REFLEX PROPERTIES OF TWOWAY CONNECTIONS OF CORTICAL NEURONS IN CATS
U. S. GASSANOV and A. G. GALASHINA
Institute of Higher Nervous Activity a n d Neurophysiology
USSR Academy of Sciences
Butlerowa 5a, 117485 Moscow, USSR
Key words: two-way conditioned connection, auditory cortex, intraanalyzer conditioning
Abstract. One of the principles in the formation of integrating activity
of the brain appears to be a two-way character of neuronal connections.
Forward and backward conditioned connections demonstrate the acquired properties of two-way relations in the cerebral cortex. Interneuronal
cortical connections and their modification i n conditioned reflex activity
were studied in a alert cats with chronically implanted electrodes using
a statistical method for analysis of two impulse flows. Multineuronal
activity was recorded in the auditory cortex. All of the investigated
neurons were divided into pairs according to their functional relations:
(1) with two-way connections, (2) with one-way connections and (3) con-.pletely independent pairs. The functional interrelation in the cortical
mlcroarea, and between cortical microareas, were defined according t o
the space distribution of neuronal pairs. Elaboration of the motor conditloned reflex led to a n intensification of two-way connections both
in the cortical microarea and between remote neurons. Extinction of
the conditioned reflex exerted almost no effect on the intensity of twoway connections, but considerably decreased the number of independent
neuronal pairs at the expense of reduction in the number of one-way
connections between cortical microareas. Thus, acquisition and extinct-
ion of the conditioned reflex were accompanied by intensive interconnected neuronal activity. The features of interneuronal connection properties during extinction inhibition were the attenuation of succesive
information transfer in cortical cells, which prevented the impulse activity from paissing to effector pathways.
INTRODUCTION
At the beginning of this century, Lorento de No (7) ascertained two
basic principles of neuronal cell connection, multiplicity and two-sidedness. The two-way principle of functional connections in the nervous
system underlies the integrative activity of the brain. The elaboration
of forward and backward connections in learning illustrates the acqwred
properties of two-way neuronal connections. Numerous experiments on
dogs and cats have shown a universal character of the acquired p1.operties of two-way connections (1, 8).
The investigation of interneuronal connections at the level of cortical
neurons appears to be the province of morphologists. Neurophysiological
work is confined to studies of synaptic mechanisms lying a t the root of
interneuronal interactions. Studies on the properties of interneuronal
connections require a reliable recording of several neurons simultaneously' and the proper objective analysis of their mutual activity.
At present, a cross-correlation analysis of two impulse series has
been widely adopted for objective studies of the functional connections
between cortical neurons. The published data indicate that interneurorial
relations identified statistically characterize the interconnections both
in brain microsection and between remote cortical projection zones
(2, 4, 5, 8). According to available experimental data, the functional
connections of cortical neurons in their complexity and diversity may
be comparable to morphological data on the multiplicity and reciprocity
of synaptic connections in the cerebral cortex. In the present work we
studied interneuronal connections and their modifications in conditioned
reflex activity through the method of the statistical analysis of two pulse
flows recorded both in one cortical mlcroarea and in the adjoining micl-oareas of the cerebral cortex.
METHODS
The experiments were conducted on four alert cats in which a m o t ~
conditioned reflex to sound was established under unrestrained conditions. Sound clicks were presented for a 5 s with a frequency of 10,'s,
and were immediately followed by 5 strokes of current of supraliminal
power with the same frequency applied to the foreleg. Prior to conditioning, a cluster of 7 nichrome electrodes with tips of 50 pm each
was implanted into the auditory cortex, under pentobarbital anesthesia.
The spontaneous multiunit activity was recorded on the alert cats a t
any time for a t least 30 s in the preconditioning period and for 30 s
just before the presentation of the conditioned signal during acquisition
and extinction of the motor reaction. The methods for implanting electrodes and techniques for statistical processing of background impulse
activity were described in detail previously (4). In the present work,
impulse series were singled out according to the amplitude of spikes
from neurograms recorded both from one electrode and from two electrodes with a distance between the centers of 70-200 ym (Fig. 1). Cross-
Fig. 1. Multineuronal recordings from the acoustic cortex in alert cats registered
by seven chronically implanted electrodes (nichrome, diameter of each electrode
50 ,urn). Digits, numbers of glued electrodes.
correlation interval histograms (CIH) were made up by means of a n
AI-128 analyzer. The analysis epoch was 20 ms, bin 2 ms. As a rule,
deviation from the mean value was observed in the first three bins,
i.e., in the first 6 ms (Fig. 2).
Fig. 2. Types of relations in neuronal pairs. Displaced relative to "On,cross-interval histograms. In the upper part, two-way nonsymmetrical connection (in one
direction excitatory effect, in the other one - inhibitory); in the middle, one-way
excitatory connection; in the lower part, independent pair of neurons.
The analysis of relations between two impulse flows of two neurons
singled out from multineuronal activity recorded with one electrode
involved the examination of 132 CIH before acquisition, 112 CIH after
acquisition and 52 CIH during extinction of the reflex. The relations
between impulse flows of two neurons from two electrodes were studied
in 20 CIH prior to acquisition, in 46 CIH after stabilization of the reflex
and in 20 CIH during extinction.
Conditioning according to myogram data lasted several days (up to
30-90 days). For this reason the stages of acquisition and extinction of
the reflex could not be followed up by using the same pair of neurons
in all of the animals. In order to elicit the domineering tendency in
interneuronal relations, the number of connections observed izl a ,oiv.?n
form of reflex activity was summed. The literature contains experimentally substantiated data, according to which a direct neuronal connection
may be identified by using certain parameters of cross-correlation histograms. It is customary to assume that histograms with nonsymmetrical
and short-term indices of dependent relations point to the existence of
the functional connection between the neurons under study (2). Observation of the conditions indicated enable us to regard cross-interval
histograms as statistical indices of interneuronal functional connxtions.
According to their functional relations all of the investigated rieurons
were divided into pairs: (i) with two-way connections, (ii) with one-way
connections and (iii) entirely independent (Fig. 2). In this case wc did
not take into account the type (i.e., excitatory or inhibitory) of connections.
RESULTS
Activity from all the seven chronically implanted electrodes in the
normal, alert cats were observed infrequently. Most often the activity
was manifested under three to five electrodes. In such cases not all of
the active points showed multineuronal background pulsation. This
picture of multineuronal activity is presented in Fig. 1. Amplitude
discriminators singled out from one to three pulse series, depending on
the character of the neurograms in each of the active points. In order
to construct neuronal CIH, the records were selected which made it
possible to define three levels of spikes: with the highest amplitude
(neuron 1 or NI), average amplitude (N2) and low amplitude (N3). In
the process of selection we tried to maintain a divisible ratio of amplitudes for the respective neurons (4 : 2 : I), where the amplitude of N3
spikes was taken as 1. This procedure permitted a distinct classification of three neuronal types. Differentiation of neurons in one multineuronal record was necessary to determine the character of functional
connections inside the microarea of the auditory cortex. In crder to
characterize the functional connection between microareas, i.e. between
neurons, recorded by two adjacent electrodes in a set, it was possible
to use the neurograms with diferent number of recorded cells. In neuronal pairs the dependent relations could be both one-way (when the connection occured only i n one direction from one cell to another and was
absent in the opposite direction) and two-way.
As was mentioned earlier symmetrical (similar in both directions)
connections were excluded from the analysis, while the types of con3
- Acta Neurobiol. Exp.
6/81
nections (excitatory, inhibitory, mixed) were considered in terms of
direction. Depending on the conditions of multineuronal activity recording, all CIH were divided into three groups: prior to acquisition, after
stabilization of the reflex and during extinction.
The percentage of neuronal pairs with two-way, one-way and entirely independent CIH types was determined in each group. Figure 3.
Fig. 3. Conditioned reflex recombinations of interneuronal connections in the
acoustical cortex in cats. Upper line, in the microarea of the cortex; lower line,
between microareas of the cortex. 1. before acquisition; 2. after stabilization of
the reflex; 3. during extinction of the reflex. Double hatched columns, two-way
connections; inclined hatched columns, one-way connections; white columns, independent pairs. Each three columns comprise the percentage ratios (the total sum
of neuronal pairs in the triplet is taken as 10O0Io).
shows summarized data obtained in all subjects. According to our criteria, the frequency of reccurring neuronal pairs with two-way arid
one-way connections in the microareas of the auditory cortex was
equally probable. Independent neuronal pairs were encountered with
the same probability. After acquisition of the motor reflex to the auditory stimulus, two-way connections became more frequent, the number
of neuronal pairs with one-way connections somewhat increased and
that of independent pairs decreased. As illustrated in Fig. 3, the relations between indices of interconnected and independent neuronal pairs
changed considerably. The overwhelming predominance of two-way connections in the microarea of the auditory cortex was observed during
extinction of the conditioned reflex, when the response after cessation
of the electrocutaneous reinforcement did not occur on three successive
presentations of the auditory signal. Neuronal pairs without any dependent relations, in this case, were entirely absent. The number of
neuronal pairs with one-way connections remained practically unchanged.
If we take the functional connections between the neurons in the
adjoining microareas, i.e., between the neurons spaced at 200 pm (measured by spaces between the electrodes of a set), then all of the conditions of alert brain activity show differences in the integration of auditory neurons inside the microarea and between microareas.
In the naive cats (that is, prior to conditioning), the independent
neuronal pairs between microareas prevailed over those pairs with oneway and two-way connections. After elaboration of the reflex a reversed
picture can be seen, namely, a greater predominance of two-way connections a t the expense of a considerable decrease of independent pairs.
When the conditioned reflex was inhibited, two-way relations appeared
to be the main form of interconnection between spaced acoustical neurons The number of pairs with one-way connections in this case dropped considerably and that of independent pairs rose somewhat.
Fig. 4. Conditioned reflex recombination of functional connections in the auditory
cortex microarea between different types of neurons. Vertical hatched coiu~nns,
hefore acquisition; horizontal hatched columns, after stabilization of the conditiuned reflex; white columns, during extinction of the conditioned reflex. Upper, twoway connections; lower, independent neuronal pairs. Types of neurons N1, N2, N3)
arc designated with the corresponding diameter of rings. Rations are calculated
as in Fig. 3.
The results of the statistical analysis of the functional integration
of cortical neurons were obtained by t h e quantitative distribution of
CIH i n neuronal pairs without taking into account the class of the cells
analyzed. As stated before, in some of the multineuronal records it was
possible to identify three types of neurons according to the amplitude
of spikes. The extent to which the defined classes of neurons participate
in the integrating processes is also of interest.
Figure 4 shows summarized data only on two-way connections of
cells of different classes as well as similar summarized data on independent neuronal pairs inside the microarea of the cortex.
In the naive cats, all of the neuronal pairs had two-way dependent
connections, but they were especially frequent between neurons 1 and 2.
Their frequency in this pair was twice as great as that of similar connections in neuronal pairs 1-3 and 2-3. After acquisition this index
leveled off in all neuronal pairs. Upon extinction of the conditioned
reflex, two-way connections started to predominate again in one of
the neuronal pairs. However, unlike the data on the naive cats, such
a pair in this series appeared to be neurons 1-3.
The distribution of independent type connections differed somewhat
from the previous indices. Complete ("two-way") independence in naive
cats was often observed between neurons 1-3 and 2-3. Aiter stabilization
of the conditioned reflex, the number of cases with the independent
work of two neurons fell to a minimum in all pairs. No cases of independent work of two neurons were observed upon extinction of the conditioned reflex.
DISCUSSION
The statistical analysis of cortical interneuronal connections based
on the cross-relations of two impulse series made in possible to quantitatively describe the functional manifestations of one of the main morphological principles in the integration of nervous cells - their twoway connection. According to morphological data of the multiplicity of
interneuronal connections, it may be assumed that the alert brain posseses a wide selection and diversity of the functional activity of contacts. Diversity of two-way contact manifestations between neurons is
also probable.
Our experimental data show that two-way connections not only
characterize one of the forms of cortical neuronal integration, but may
also attest to a directed systemic activity of the trained brain. Elaboration of both positive and inhibitory conditioned reflexes was accompanied by activation of two-way connections in the cortical projection
of a signal stimulus. Interconnected activity was intensified between
closely spaced neurons and those spaced a t tens and hundreds of microns.
However, the development of conditioned inhibition was marked by a n
especially high level in the interaction of closely spaced neurons and
a considerable decrease in the activity of one-way connection; between
remote neurons.
Thus, the high intensity of connections is characteristic of auy type
of learning, but inhibitory reactions caused local activation and attenuation of unidirectional spreading of excitation. Our previous investigations
showed that positive and inhibitory conditioned reflexes were characterized by predominately excitation forms of interneuronal connections
(5). In this connection it may be assumed that a decrease in the activity
of spreading unidirectional connections in the cortical projection apparently prevents the impulse flow from passing to effector pathways.
Selective disengagement of successive connections should be emphasized.
They were disrupted between relatively remote neurons, while in closely spaced cells (inside a microarea) they were maintained at high level.
Further investigations will show to what extent these hypotheses are
correct, but one fact is obvious that interconnected activity of acoustical
cells upon extinction remained as intensive as in the positive condit:.oned
reflex, while in microareas of the cortex it was even higher. All 01 these
data may attest to the intensification of local processes in the cortical
projection of conditioned stimuli during inhibition of the trained motor
reaction. These results are consistent with the concepts of the active
form of cortical activity under the condition of established internal
inhibition. They are described at length in one of our works (3).
In addition, the present results also attest to the conditioned reflex
properties of two-way cortical connections, their high reactivity when
subjected to adequate physiological study. Such acquired functional
variability in the interaction of cortical neurons located in various projection cortical zones has been shown in the research of Livanov (6), and
Merzhanova (8).
Attention is also drawn to the distribution and the flexibility of
two-way connections depending on the size and class of neurons. These
data that we obtained in separate microareas of the auditory cortex
revealed a n intensification of such connections with neurons generating
low-amplitude spikes. They manifested the most pronounced intensiiication of two-way connections after conditioning and especially after extinction of the conditioned reflex. It may be concluded that intraanalyzer
local integration of cortical neurons in learning is brought about mainly
by the activity of small neurons that play the role of a corinecting element in the functional combinations of neurons.
One more fact is noteworthy. According to ovr indices, the systemic
activity of neurons relative to interconnected, unidirectional and independent neuronal pairs, had a specific character in each experimental
situation. In describing the flexibility of interneuronal connections in
the microsystem of three classes, this specificity was also manifest, but
to a lesser degree. In our previous work we conducted a detailed analysis
of the three neuronal microsystem according to types and direction of
connections and revealed the specificity of systemic organization of cortical cells in learning by backward learning (4) and conditioned reflex
activity (5). The results of studies on two-way connections are consistent
with the concept of specificity of the systemic mechanism in cortical
activity in learning. At the same time they attest to the fact that twoway inerneuronal connections require special physiological investigations
and that their functioning depends to a large extent on the animal's
behavior.
REFERENCES
1. ASRATYAN, E. A. 1980. Reflex mechanisms of motivational behavior. In
R. F. Thompson, L. H. Hicks a n d V. B. Shvyrkov (ed.), Neural mechanisms
of goal-directed behavior a n d learning. Academic Press, New York, p. 11-38.
2. DICKSON, J. N. a n d GERSTEIN, G. L. 1974. Neuronal interactions i n auditory
cortex. J. Neurophysiol. 37: 1239-1261.
3. GASSANOV, U. G. 1972. Vnutrennee tormozhene. Nauka, Moscow, 142 p.
4. GASSANOV, U. G. a n d GALASHINA, A. G. 1980. Specific a n d unspecific neuron a l mechanisms of cellular operant conditioning. Acta Neurobiol. Exp. 40:
137-147.
5. GASSANOV, U. G., GALASHINA, A. G. a n d BOGDANOV, A. V. 1980. A study
of neuron systems activity i n learning. In R. F. Thompson, L. H. Hicks a n d
V. B. Shvyrkov (ed.), Neural mechanisms of goal directed behavior a n d
learning. Academic Press, New York, p. 341-352.
6: LIVANOV, M. N. 1975. Neural mechanisms of memory (in Russian). Vsp. Fiziol.
Nauk. 6: 66-89.
7. LORENTO d e NO'R. 1932. T h e regulation of eye positions a n d movements
induced by t h e labirinth. Laryngoscope 42: 233-330.
8. MERZHANOVA, G. 1979. Single unit activity in t h e visual cortex during cond i t ~ o n i g ni n cats. Acta Neurobiol. Exp. 39: 553-566.