Site hosted by Build your free website today!

The Hippocampus


The hippocampus is one of a group of structures forming the limbic system and is a part of the hippocampal formation, which also includes the dentate gyrus, subiculum, and entorhinal cortex.  Different components of the limbic system have been shown to play a critical role in all aspects of emotions, fear, learning and memory (Geinisman et al. 2000a; Geinisman 2000b; McMillan et al. 1987; Cardinal et al. 2002) . 

The initial insights on the role of the hippocampus came from studies of amnesia in human patients following removal of the hippocampus plus neighboring medial temporal structures (Scoville and Milner 1956).  Extensive evidence implicates the hippocampus and related structures in the formation of episodic memories in humans (Reilly 2001; Aggleton and Brown 1999) and in consolidating information into long-term declarative memory (Mumby et al. 1999) .   

Hippocampal Inputs and Outputs

The hippocampus has direct connections to the entorhinal cortex (via the subiculum) and the amygdala. Outputs from these structures can then affect many other areas of the brain.  For example the entorhinal cortex projects to the cingulate cortex, which has a connections to the temporal lobe cortex, orbital cortex, and olfactory bulb. Thus, all of these areas can be influenced by hippocampal output, primarily from CA1.

The entorhinal cortex is a major source of inputs to the hippocampus collecting information from the cingulate cortex, temporal lobe cortex, amygdala, orbital cortex, and olfactory bulb (Johnston and Amaral 1998).  The hippocampus receives inputs via the precommissural branch of the fornix from the septal nuclei.  Figure 1.4 gives an overview of the major inputs and outputs of the hippocampus.

Intrahippocampal Pathways

The main input to the hippocampus (perforant pathway) arises from the entorhinal cortex and passes through to the dentate gyrus. From the granule cells of dentate gyrus connections are made to area CA3 of the hippocampus proper via mossy fibers.  CA3 sends connections to CA1 pyramidal cells via the Schaeffer collateral (SC) as well as commissural fibers (comm.) from the contralateral hippocampus.  The major neurotransmitter in these three pathways is glutamate.  The final output from the two CA fields passes through the subiculum entering the alveus, fimbria, and fornix and then to other areas of the brain.  For the purposes of this dissertation I will focus on the synapse between the CA3 to CA1 neuron.  Figure 1.5 is a



Figure 1.4.  Schematic representation of hippocampal connections.

Information leaving the entorhinal cortex can enter any of the following layers: CA3, CA1 or the subiculum.  Information entering the dentate gyrus predominantly follows the mossy fiber pathway to CA3.  Information from CA3 leaves via the Schaffer collateral pathway for the CA1 region.  Information from CA1 travels to the subiculum and then to the entorhinal cortex.

Figure 1.5.  Hippocampal pathways and their stimulation

Signals from the entorhinal cortex (EC) enter the dentate gyrus (DG) via the perforant path (PP).  From the DG information travels to the CA3 pyramidal neurons via the mossy fibers.  From the CA3 neurons the signal leaves via the Schaffer collaterals and joins with the commissural fibers (Comm.) from the contralateral CA3 making connections with CA1 pyramidal neurons.  Signals leaving CA1 then travel to neurons within the subiculum. A bipolar stimulating electrode was placed on the Schaffer collateral and commissural (comm.) fibers.  Recording electrodes placed in the dendritic layer and/or the pyramidal layer of CA1 will record an Excitatory Postsynaptic Potential (EPSP) or a population spike (PS) following stimulation, respectively.  As discussed in the text the EPSP represents the response at the CA3-CA1 synapse and the PS represents the number of pyramidal cells firing and the contribution of the EPSP at that location. The top portion of the figure demonstrates the four layers that the CA1 pyramidal neuron lies within (S. denotes Stratum).  The small neuron with a letter “I” represents an inhibitory interneuron.  The pathway diagramed in the top portion of the figure corresponds to the recurrent inhibitory loop in area CA1.


diagrammatic representation of the pathways entering the hippocampus and the pathways within it. 

     The highly organized and laminar arrangement of synaptic pathways makes the hippocampus a convenient model for studying synaptic actions in vivo and in vitro (Andersen et al. 1971). 

     Electrophysiology of the CA3ŢCA1 Synapses

     Extracellular field recordings represent the summed responses from a number of neurons in the vicinity of the recording electrode.  Because of the orderly arrangement of the pyramidal neurons and their dendrites, electrical field recordings offer valuable information about the temporal arrangement of responses from apical dendrites to cell bodies.  Following stimulation of the SC and commissural fibers (Figure 1.5 – stimulating electrode), an extracellular recording electrode in the stratum radiatum (Figure 1.5 – S. radiatum) containing synapses, would record a small negative potential that results from the action potentials generated in the presynaptic fibers (Fiber Volley, FV).  Following the FV a slow negative potential, corresponding to the population excitatory postsynaptic potential (pEPSP), would be recorded (Figure 1.5 shows a representative pEPSP, which will be referred to as an EPSP from now on).  The EPSP represents depolarization at the postsynaptic membrane, indicating that transmission took place at the CA3-CA1 synapse.  Placing the recording electrode in the stratum pyramidale (Figure 1.5 – S. pyramidale) would allow us to record a positive deflection due to current exiting the basal dendrites near the cell body.  If the magnitude of the depolarization is sufficient to bring the pyramidal cell to threshold, it will fire one or more action potentials.  These action potentials will be recorded as a negative potential overlapping the positive potential.  This type of recording is known as a population spike (PS) and is represented in Figure 1.5.  While the EPSP is affected by changes occurring at the synapse the PS is affected by combination of 3 factors:  1) the amplitude of the EPSP, 2) the passive properties of the CA1 pyramidal cell (from dendrites to axon hillock), and 3) the level of inhibition produced by the GABAergic interneurons innervating the CA1 pyramidal neurons.  Changes in the PS give a great deal of information about the number and excitability of neurons involved in the final output from the hippocampus. 

     The CA1 Pyramidal Neuron

     Activation of the CA3 neuron leads to an increase in glutamate release from the nerve terminals of the SC’s.  Glutamate released in the stratum radiatum and stratum lacunosum moleculare of CA1 activates either ionotropic or metabotropic receptors.  The ionotropic glutamate receptors are classified into three types AMPA, kainite, and NMDA receptors, named after the ligand initially used to characterize them.  AMPA and kainite receptors mediate the fast EPSP seen following SC stimulation (Karnup and Stelzer 1999).  NMDA receptors mediate slow-rising EPSP’s and are thought to be responsible for some forms of long-term synaptic plasticity (Tsien et al. 1996; Kullmann et al. 1996).    Metabotropic glutamate receptors, which are located at both the presynaptic and postsynaptic side act to modulate release of neurotransmitter presynaptically (Lie et al. 2000; Baskys and Malenka 1991) , and modify postsynaptic responses (Xiao et al. 2001).

     The major inhibitory neurotransmitter in the hippocampus is GABA (Roberts 1976).  Eliciting a single evoked potential via stimulation of the SC’s results in a characteristic sequence of excitation followed by inhibition when recorded from the stratum pyramidale.  In rats the excitation typically precedes the inhibition by a few milliseconds.  The inhibition arises from feedforward and feedback connections via inhibitory interneurons.  The inhibition corresponds to the release of GABA, which initiates two types of inhibitory responses, a fast inhibitory postsynaptic potential (IPSP) mediated by GABAA receptors and a slow IPSP brought on by GABAB receptor activation.

Hippocampal Synaptic Plasticity

     The hippocampus exhibits short and long term synaptic plasticity.  For the purpose of our discussion plasticity will be defined as a change in the efficiency of synaptic transmission following previous synaptic activity. 

     Short-term Plasticity

     Short-term synaptic plasticity lasting from a few milliseconds to a few minutes can be elicited, among other means by paired pulse stimulation.   

Activation of the CA1 pyramidal cell by a single pulse will lead to an action potential sent out of the hippocampus and to inhibitory interneurons within the hippocampus (Top insert in Figure 1.5; Arrows leaving S. pyramidale).  Activation of inhibitory neurons (Figure 1.5 – Top insert; Inhibitory interneuron labeled “I” in S. oriens) by the CA1 neurons will lead to the recurrent inhibition of the subsequent response initiated in the CA1 neuron by the second stimulus delivered shortly (10-13 ms) after the first one.  This type of pairing of two pulses in rapid succession leads to an inhibition known as paired pulse inhibition (PPI). 

Changes in the ratio of the amplitudes of the first and second evoked potentials (in a PPI experiment), can occur through changes in both GABA receptor sensitivity and GABA release.  This has been demonstrated experimentally through, an enhancement of PPI with GABA agonists (Rock and Taylor 1986) and a decrease in PPI by GABA antagonists (Kapur et al. 1989, 1997) . 

If the stimuli are further apart (15-40 ms) the second stimuli arrives, when the recurrent inhibitory loop has already been inactivated.  Therefore, the second response is not inhibited but facilitated due to residual Ca2+ increase after the first stimulus.  This is called paired pulse facilitation (PPF).  Changes in the ratio of amplitudes of first and second potentials are generally accepted as a modification in the presynaptic component of the synapse (Commins et al. 1998; Chen et al. 1996; Gottschalk et al. 1998), although alterations in postsynaptic AMPA receptors have also been reported during PPF experiments (Wang and Kelly 1995-1997, 2001).     

     Long-term Plasticity

     When a change in synaptic efficiency persists for long periods of time (hours to days) it is known as long-term plasticity.  In 1973 Bliss and Lomo observed that high frequency stimulation (HFS) of the perforant path in anaesthetized rabbits led to a potentiation of synaptic responses which could last for several hours (Bliss and Lomo 1973).  Later many other pathways in the brain including the SC’s in the hippocampus were shown to express similar phenomenon following HFS (see review by Recasens 1995).  The general consensus is that induction of LTP at CA1 synapses requires Ca2+ entry into the postsynaptic dendritic spine via the activation of the NMDA receptor (Collingridge et al. 1983), however increased Ca2+ concentration through NMDA-independent mechanisms may also lead to LTP (Grover 1998; Kullmann et al. 1992) .  Therefore Ca2+ which is directly involved in PPF, also initiates the mechanisms which maintain enhanced synaptic transmission for a long period of time (Impey et al. 1996; Solderling and Derach 2000; Suzuki 1996) .
     The mechanism responsible for LTP has been an area of intense debate.  Some of the more recent mechanisms proposed to explain LTP at hippocampal synapses include: 1) an incorporation of new AMPA receptors into the membrane (Pickard et al. 2001; Hayashi et al. 2000). 2) activation of previously silent synapses (Malinow 1995 and Konerth 1996), and 3) HFS induced splitting of dendritic spines allowing a synaptic response to be amplified (Jontes and Smith 2000). 

     Although some studies demonstrated a strong correlation between a deficit in LTP and poor spatial memory (Sakimura et al. 1995; Abel et al. 1997), several reports described normal spatial orientation in spite of impaired LTP (Meiri et al. 1998; Saarelainen et al. 2000; Jun et al. 1998; Bach et al. 1995).  These briefly mentioned conflicting results demonstrate some of the difficulties in accepting LTP as the molecular mechanism of memory formation.   


To go back to the main index click here.