Synaptic Mechanisms Underlying Classical Conditioning

Written by: Jennifer Beauharnois –Psychobiology 325

In order to understand the neural mechanisms by which an organism 
acquires and retains information identification of the site(s) of 
learning and memory storage must occur. It now appears that in mammals 
there are different forms or aspects of memory that are subserved by 
different brain structures. For example, the hippocampus seems to be 
important for spatial, contextual and relational memories, whereas the 
cerebellum is necessary for classical conditioning (Davis, 1992).
Classical or Pavlovian conditioning is the simplest form of 
associative learning by which animals, including humans, learn 
relations among events in the world so that their future behaviors are 
better adapted to their environments (Rescorla, 1988). Eyeblink or 
nicitating membrane conditioning provides a perfect display of such 
learning mechanisms since the unconditioned stimulus (US) (usually and 
air puff to the eye), and the conditioned stimulus (CS)(usually a tone) 
are well defined and can be precisely controlled.
Converging lines of evidence from lesioning, recording, 
stimulation, reversible inactivation and brain-imaging studies indicate 
that the cerebellum (an important site for motor learning) is essential 
for eyeblink conditioning (Davis, 1992). In brief, Thompson (1990) 
concluded that selective lesions (electrolytic or chemical) of the 
cerebellum prevent the acquisition and retention of conditioned 
eyeblink responses. 
The cereberal cortex consists of two layers of neuronal cell 
bodies, the Purkinje cell layer and the granular cell layer. The 
Purkinje cells are of main interest because their axons synapse on 
neurons in the deep cerebellar nuclei, which are the major afferent 
cells of the cerebellum (Bear, Connors, & Paradiso, 1996). Purkinje 
cell dendrites are directly contacted by one of the two major inputs 
into the cerebellum –axons of the inferior olive in the medulla called 
climbing fibers and the axons from the pontine nuclei called the mossy 
fibers (Bear et al. ,1996). Indirect transfer of impulses comes from 
cells that run along an opposite plane than the mossy or climbing 
fibers, these cells are termed parallel fibers and they consist of the 
axons from the granular cells. 

Electrophysiological studies done by Thompson (1990) indicate that the 
Purkinje cells undergo learning-induced changes during eyeblink 
conditioning. It was presumed that the joint stimulation of the two 
sets of synapses with the dendrites of the purkinje cells strengthens 
the one transmitting the auditory information of the conditioned 
stimulus. This presumption was displayed when Steinmetz, Rosen, 
Chapman, Lavond and Thompson (1986) classically conditioned the 
nicitating membrane response, only using electrical stimulation of the 
pontine nucleus (CS) and the inferior olive (US). This displayed a 
successful classical conditioning excluding the information entering 
through the auditory system and the sensorimotor system of the eye.
	In order to monitor the effectiveness of parallel fiber synapses 
on Purkinje cells Masao Ito (1990) applied electrical stimulation to 
the parallel fibers and measured the size of the EPSP in the Purkinje 
cells. Then, to induce synaptic plasticity he paired stimulation of the 
climbing fibers with stimulation of the parallel fibers. It was found 
that after this pairing procedure, activation of the parallel fibers 
alone resulted in a smaller postsynaptic response in the Purkinje cell. 
This type of modification could last at least one hour and was 
therefore termed long- term depression (LTD).
	Masao Ito recognized that an important property of long-term 
depression was that it will only occur in those parallel fibers that 
are active at the same time as the climbing fibers (input specificity). 
There is an input- specific modification when activation of the 
parallel fiber-Purkinje call synapse occurs at the same time as 
depolarization of the postsynaptic Purkinje cell by the climbing 
fibers. After classical conditioning, the sensory-motor synapse became 
more effective (known as long-term potentiation), while after pairing 
parallel and climbing fiber activity, the parallel-Purkinje synapse 
becomes less effective (long-term depression), (Bear et al.,1996).
	The evidence that was available for mentioning summarized the 
three intracellular signals that occur at the same time: (1) a rise in 
calcium ions due to climbing fiber activation, (2) a rise in sodium 
ions due to AMPA receptor activation on the Purkinje cell dendrite, and 
(3) activation of protein kinase C due to metabotropic receptor 
activation (Bear et al., 1996). According to this model of long-term 
depression, learning occurs when rises in calcium and sodium ions 
coincide with the activation of protein kinase C. Memory occurs when 
AMPA channels are modified and excitatory postsynaptic currents are 
depressed.
	It was also found that long-term potentiation-related enhancement 
of sensorimotor synapses plays a significant role in classical 
conditioning of the eyeblink response (Frost, Clark & Kandel, 1988) A 
summarized mechanism for long-term potentiation is mentioned below in 
order to avoid the possibility that one might only assume that long-
term depression is necessary for classical conditioning. The mechanism 
goes as follows: Stimulation of the tone (CS) activates sensory 
neurons. The stimulation of the air puff to the eye (US) causes a 
strong depolarization of the nicitating membrane motor neurons (caused 
by indirect activation of excitatory interneurons), (Frost, et 
al.1988). The release of the presynaptic neurotransmitter-glutamate 
(Dale & Kandel 1993)-together with the strong postsynaptic 
depolarization activates postsynaptic NMDA or NMDA-type receptors. 
Activation of these receptors causes a postsynaptic influx of Ca2+ 
through the cell membrane's ion channels (Mackey, Kandel & Hawkins, 
1989). This increase in intracellular calcium initiates the activation 
of Adenylate cyclase, which in turn converts ATP into Cyclic AMP 
(giving off inorganic Phosphates as a byproduct (Thompson & Kim, 1997). 
Cyclic AMP then continues to activate one or more protein kinases that 
participate in the phosphorylation of a receptor protein in the 
postsynaptic membrane. This phosphate addition alters the formation of 
the protein, which can either result in depolarization or 
hyperpolarization of the membrane. This biochemical chain- reaction 
within the interneurons, paired with presynaptic activity produces both 
presynaptic and postsynaptic changes that contributes to the 
strengthening of sensorimotor connections involved in classical 
conditioning

				References:

1. Davis, M.(1992) Annual Review of Neuroscience,  
   Vol.15, 353-375.
2.	Rescorla, R.A. (1988) Annual Review of Neuroscience,
Vol.11, 329-352.
3.	Thompson, R.F. (1990) Philos. Trans. R. Soc. London
  Ser. B 329, 161-170.
4.	Bear, M., Connors, B., & Paradiso M., Neuroscience:
  Exploring the Brain, Baltimore: Williams & Wilkins, 1996;        
  Vol.1, 250-258.
5.	Steinmetz, J.E., Rosen, Chapman, 
6.	Ito, M., The Cerebellum and Neural Control. New York: Raven 
Press,1984.
7.	Frost, W.N., Clark, G.A., & Kandel, E.R.,(1988) Neurobiology, 
Vol.19, 297.
8.	Dale, N. & Kandel, E.R., (1993) Proc. Natl. Acad. Sci. U.S.A., 
Vol.90, 7163.
9.	Mackey, S.L., Kandel, E.R., & Hawkins, R.D., (1989) Journal of 
Neuroscience, Vol.9, 4227.
10.	Thompson, R.F. & Kim, J.J. (1997) Cerebellar    
     circuits and synaptic mechanisms involved in 
     classical eyeblink conditioning, Trends in 
     Neuroscience. Vol.20, 177-181.

Synaptic Specializations of Classical Conditioning
    By: Melissa Jessup


	Classical conditioning is a form of associative learning in which 
an unconditional stimulus (US) and its subsequent unconditional 
response (UR) are paired with a conditional stimulus (CS) to form a 
conditional response (CR).  An example of classical conditioning is the 
nictating membrane response often tested in rabbits, in which case the 
unconditional stimulus is an air puff to the eye and the conditioned 
stimulus is usually a tone.  For such associations to be learned, many 
cellular changes must occur concurrently within the brain, including 
changes and specializations within synapses.
	Studies have shown that synaptic changes involved with classical 
conditioning take place within both the MGm (medial division of the 
medial geniculate nucleus) of the thalamus and the amygdala (Carlson, 
1998).  A single unit recording study showed that neurons in the MGm 
increased their responsiveness to the auditory CS after it was paired 
with an aversive US (Weinberger, 1982).  In recording neural activity 
in the lateral amygdala of rats throughout stages of pairing a tone 
with a shock, it was found that within a few trials, neurons became 
more responsive to the tone, and there was an increase in the number of 
neurons that responded altogether (Quirk, Repa, and LeDoux, 1995).
	The synaptic changes within the thalamus and amygdala have been 
shown to occur as a result of long term potentiation (LTP) like 
phenomena that occurs during classical conditioning (Cowan, Sudhof, and 
Stevens, 2001).  Rogan et al., (1997b), monitored the evoked potentials 
elicited by an auditory CS in the lateral nucleus of the amygdala in 
rats before and after fear conditioning.  Paired presentations of the 
auditory CS and the foot shock resulted in and an increase in freezing 
behavior in response to the tone along with a parallel potentiation of 
the CS- evoked potential.  The removal of the foot shock after the CS 
led to the extinction of the conditioned fear and the fall of the CS 
evoked response back to the baseline levels.  NMDA receptors are the 
induction mechanism underlying CS and US association within the 
amygdala.  Using fear potentiated startle to visual and auditory CS?s, 
it was found that the infusion of AP5 (a drug that blocks NMDA 
receptors) into the amygdala blocks the acquisition, but not the 
expression, of fear conditioning (Miserendino et al., 1990; Campeau et 
al., 1992).  This shows that AP5 can interrupt associative learning by 
blocking synaptic plasticity in the amygdala, but cannot affect changes 
that have previously occurred.
	The increase in synaptic strength during long-term potentiation 
is a result of changes in both the pre- and postsynaptic neurons.  
Presynaptically, such changes could be produced through an increased 
release of transmitter substance.  Postsynaptic changes could occur 
through an increased ability of the receptors to activate changes in 
the permeability of the postsynaptic membrane, and an increased 
communication between the region of the postsynaptic membrane and the 
rest of the neuron.  Changes could also occur through an overall 
increased number of synapses, which would involve both pre and 
postsynaptic changes (Carlson, 1998).  
	One major postsynaptic property of plasticity is dendritic 
spines.  In a study by Hosokawa et al. (1995), it was shown that LTP 
causes spines to change structurally, growing in length and changing 
their orientation toward the shaft of the dendrite. LTP also serves to 
increase the number of perforated synapses, which are characterized by 
two enlarged active zones within the synapse that occur when the spine 
projects into the terminal button.   Buchs and Muller (1996), found 
that after long term potentiation, certain spines formed perforated 
synapses with presynaptic terminals.  The development of these 
perforated synapses causes the number of postsynaptic AMPA receptors to 
increase (Edwards, 1995).  When the active zones are enlarged, the 
terminal button inserts more mechanisms for the release of 
neurotransmitter into the presynaptic membrane.  The dendritic spine 
also inserts an increased amount of AMPA receptors into the 
postsynaptic membrane.  This increase in AMPA receptors makes the 
postsynaptic membrane of the dendritic spine more sensitive to the 
release of glutamate by the presynaptic terminal button (Carlson, 
1998).  
	Postsynaptic specialization may also cause presynaptic changes 
through the activity of an enzyme called nitric oxide synthase (NOS), 
which produces nitric oxide (NO), a highly soluble gas that can 
communicate messages from one cell to another.  The entry of calcium 
ions through NMDA channels causes activation of NOS.  (Carlson, 1998). 
Nitric oxide diffuses out of the dendritic spine back towards the 
terminal button causing a rise in cyclic GMP.  It has been shown that 
NOS inhibitors specifically block the rise of cGMP.  Cyclic GMP then 
triggers an increased amount of glutamate, which in turn causes an 
increased amount of neurotransmitter to be released from the terminal 
button. (Cowan, Sudhof, and Stevens, 2001).  
	Long term potentiation requires many series of changes with 
contributions from both pre and postsynaptic cells.  Together, these 
changes create plasticity of synapses, allowing for the retention of 
associative learning in classical conditioning.

References:

Carlson, N. Physiology of Behavior.  Massachusetts:  Allyn and Bacon, 
1998, pp. 417 to 425.

Cowan, W., Sudhof, T., and Stevens, C., Synapses.  John Hopkins 
University, 2001, pp. 461 to 464, 457, 519 to 570.

Zimmerman, H., Synaptic Transmission; Cellular and Molecular Basis.  
1993, pp. 111 to 119.

Johnson, D., Baker, J., and Azorlosa, J, Acquisition, extinction, and 
reinstatement of Pavlovian fear conditioning: the roles of the NMDA 
receptor and nitric oxide. Science Direct: Brain Research.  Volume 857, 
Issues 1 and 2, 2000 pp. 66 to 70.

      James Davis 
      Psychobiology Project

The neurological circuit of classical conditioning
	
The mammalian brain is a complex collection of tissues, 
chemicals and electricity, that all combine in a profound 
organization to produce behavior. What's more, this organization 
is constantly changing as these organisms learn and grow. The 
survival of a species is dependant upon its ability to learn and 
adapt.  One of the mechanisms that have evolved to accomplish 
this is classical conditioning. 

Classical conditioning occurs when an (UCS) unconditioned 
stimulus is paired repeatedly with a conditioned stimulus so that 
the conditioned stimulus (CS) produces the (UCR) unconditioned 
response. In the nictitating response mechanism of rabbits, the 
UCS is usually the puff of air to the eye, and the UCR is the 
withdrawal of the eye through the nictitating membrane. Any CS 
that is repeatedly paired with the UCS can elicit this UCR. 

Thompson proposed a circuit by which the air puff (UCS) and 
the auditory tone (CS) produce the nictitating membrane response. 
The CS is presented to the rabbit, then a few milliseconds later 
the UCS is presented. The information about the air puff is 
transmitted to the sensory trigmental nucleus. The information 
about the auditory tone is sent to the ventral cochlear nucleus. 
The sensory trigmental nucleus sends information about the UCS to 
the pontine nucleus. At the same time the ventral cochlear 
nucleus sends information about the CS to the inferior olive 
nucleus. (Thompson 1990) The inferior olive nucleus and pontine 
nucleus send their axons to the dendrites of the purkinje cells. 
The purkinje cells send their axons to the interpositus nucleus. 
The interpositus nucleus sends its axons to the red nucleus, 
which sends its axons to the accessory abducens nucleus, which 
causes the retractor bulbi muscle to contract bringing the 
rabbits eye through the nictitating membrane   

Thompson tested his hypothesis by electrically stimulating 
the inferior olive nucleus and the pontine nucleus. He found that 
whatever response was caused by stimulation of the inferior olive 
nucleus became classically conditioned to the pontine simulation. 
Stimulation of the pontine served as the CS and stimulation of 
the inferior olive nucleus served as the UCS. (Thompson 1989) 

When classical conditioning occurs the synapses from the 
pontine nucleus and the inferior olive nucleus change at the 
purkinje dendrites. The weak synapse from the pontine nucleus 
could not alone, produce the necessary action potential to cause 
the nictitating membrane response. When fired at the same time as 
the already strong synapse from the inferior olive nucleus the 
weak synapse is strengthened; and if this process is repeated, 
eventually the pontine synapse can cause the response. 

References:

1. Davis, M.(1992) Annual Review of Neuroscience,  
   Vol.15, 353-375.
2.	Rescorla, R.A. (1988) Annual Review of Neuroscience,Vol.11, 
329-352.
3.	Thompson, R.F. (1990) Philos. Trans. R. Soc. London  Ser. B 
329, 161-170.
4.	Bear, M., Connors, B., & Paradiso M., Neuroscience:  
Exploring the Brain, Baltimore: Williams & Wilkins, 1996; 
Vol.1, 250-258.
5.	Steinmetz, J.E., Rosen, Chapman, 
6.	Ito, M., The Cerebellum and Neural Control. New York: Raven 
Press,1984.
7.	Frost, W.N., Clark, G.A., & Kandel, E.R.,(1988) 
Neurobiology, Vol.19, 297.
8.	Dale, N. & Kandel, E.R., (1993) Proc. Natl. Acad. Sci. 
U.S.A., Vol.90, 7163.
9.	Mackey, S.L., Kandel, E.R., & Hawkins, R.D., (1989) Journal 
of Neuroscience, Vol.9, 4227.
10.	Thompson, R.F. & Kim, J.J. (1997) Cerebellar    
     circuits and synaptic mechanisms involved in 
     classical eyeblink conditioning, Trends in 
     Neuroscience. Vol.20, 177-181.

Copyright © 2001, Dr. John M. Morgan, All rights reserved - This page last edited March 12, 2001
If you have any feedback for the author, E-mail me