Question

. Electrical stimulation is applied to entorhinal cortex, and neural recording is made at the output of hippocampus. After repeatedly applying stimulation, it is found that synapses at the recording site did not change in any way (i.e., did not show greater activity when activated). Explain why change in synaptic activity did not occur, and suggest what occurred at the level of the synapse. Describe all relevant events.

Answer 1

stimulation of the perforant pathway will result in activation of neurons in the entire hippocampal formation. Because of intrinsic connections within the hippocampal formation, principal neurons in the dentate gyrus (i.e., granule cells) receive mainly monosynaptic inputs, whereas principal cells and interneurons in the hippocampus proper and subiculum (i.e., pyramidal cells) not only receive monosynaptic inputs via perforant pathway fibers, but also receive additional inputs via the hippocampal trisynaptic circuit. Because there is no, or at least no substantial, backpropagation of signals from the CA3 region to granule cells in the dentate gyrus (Scharfman, 2007), a single electrical pulse applied to the perforant pathway will elicit one electrophysiological response; at low intensities this results in an fEPSP and, in addition, at higher intensities an fEPSP with a population spike component can be evoked (Figs. 23.3 and 23.4). However, backprojection from CA3 pyramidal cells to hilar mossy cells and GABAergic interneurons occurs causing a feedback inhibition of granule cells during repetitive stimulations.Thus, the stimulation protocol defines the input activity, and, therefore, the most reasonable region to study the relationship between neuronal activity and resultant BOLD response is, in this experimental setup, the dentate gyrus. In particular, when the recording electrode is located in the granule cell layer, the recorded population spike amplitude can be related to the amount of synchronized spiking of granule cells. By this means, the spiking activity of principal cells can be directly related to the resultant BOLD response, whereas the incoming activity can be directly related to, and adjusted by, the stimulation protocol used

application of one single electrical pulse may not be sufficient to elicit a substantial change in BOLD signal intensities in the hippocampal formation, at least if biphasic pulses with a length of 0.1–0.2 ms are used. It turned out that at least several pulses (i.e., 5–10 consecutive pulses within 8s) are required to elicit detectable changes in BOLD signal intensities in the hippocampal formation. The actual threshold depends on the present anesthetic/sedative (Krautwald and Angenstein, 2011) and may be even lower in awake animals. Increasing the number of consecutive pulses within a constant stimulation period (tantamount to an increased stimulation frequency) increases the BOLD response, however, only up to a certain level, which is using a field strength of 4.7 T about 4%. Thus, only a small frequency range (under medetomidine sedation between 0.625 and 5 Hz) exists, in which the strength of the BOLD response directly relates to the input activity (Krautwald and Angenstein, 2011). It rather appears that in the naïve, i.e., previously electrically unstimulated hippocampus, an initial “default”-like BOLD response to complex stimulus pattern is generated that adjusts during subsequent stimulation periods (Riemann et al., 2017). Evidently the quality of induced BOLD responses changes with repetition of the same stimulus protocol. Thus, the input activity, defined by the pulse protocol, determines not only the strength of the BOLD response but also how the incoming pulses are processed.

Another way to modify input activity to the hippocampal formation is to vary the pulse intensity. When the stimulation electrode is located in the perforant pathway, pulses of higher stimulation intensity will affect more fibers, thus more neurons in the hippocampal formation will be activated. Again, a minimal pulse intensity exists that is required to induce detectable BOLD responses. For low-frequency pulse trains, pulse intensities are required that also elicit population spikes (around 100 μA). Increasing the pulse intensity will not only lead to increased neuronal responses (according to the measured field potential, i.e., steeper slope of the fEPSP and bigger population spike amplitude) but also lead to an increased hemodynamic response. Again, there is a limited range (roughly between 100 and 400 μA) for a linear relationship between applied pulse intensity and resultant BOLD response (see Fig. 23.3). Although a further increase of pulse intensities will still cause activation of additional neurons, the resultant maximal BOLD response remains similar; however, the spreading of a significant BOLD response still increases. Thus, under this condition, changes in the spatial distribution are more closely related to variations in pulse intensities than to the magnitude of the BOLD response. When, however, the number of identical pulses within one stimulation period is increased, then BOLD signal intensities also increase and the apparent maximal BOLD response is reached at lower pulse intensities (around 200 μA, Figs. 23.3 and 23.4). Consequently, based on measured BOLD responses in the hippocampus, a clear conclusion about the underlying changes in neuronal activities is almost impossible, at least, if one only analyses the initial response. This may be different if identical stimulation trains are subsequently applied and the development of individual BOLD responses is considered (Riemann et al., 2017).

In contrast to electrophysiological recordings of neuronal responses in the vicinity of the recording electrode in the dentate gyrus, fMRI also visualizes, although only indirectly, altered neuronal activities in the entire hippocampal formation and in the whole brain. This opens up the possibility of following the spread of activation within the hippocampal formation and the possible propagation of neuronal activities to target regions of the hippocampal formation during repetitive stimulations. Thus, the combination of electrical stimulation of the perforant pathway during an fMRI scan allows the detection of complex brain-wide neuronal circuits that are controlled by certain hippocampal activities.

As perforant pathway fibers only connect neurons inside the hippocampal formation, i.e., on the one hand, the entorhinal cortex and on the other hand, the dentate gyrus, hippocampus proper and subiculum, any detected stimulus-related changes in BOLD signal intensities outside the hippocampal formation relate to outgoing activities generated within the hippocampal formation. The entorhinal cortex is connected to the amygdala, nucleus accumbens, claustrum, striatum, and prefrontal cortex (Hoover and Vertes, 2007; Insausti et al., 1997; Swanson and Kohler, 1986; Witter et al., 1989). The CA1 and especially the subiculum project to several cortical regions (e.g., the medial and ventral orbitofrontal cortex, the pre- and infralimbic cortex, the agranular insular cortex, the retrospinal cortex, and the anterior cingulate cortex) as well as to subcortical structures, namely the thalamus, nucleus accumbens, amygdaloid complex, bed nucleus of the stria terminalis, and the medial mammillary nucleus (Cenquizca and Swanson, 2007; Witter, 2006). Because there are a number of target regions that receive inputs from both, the entorhinal cortex and CA1/subiculum projections, BOLD responses that are detected in these regions cannot be unambiguously allocated to neuronal activities originating in the entorhinal cortex or CA1/subiculum. One possible experimental approach to distinguish the actual involvement of the entorhinal cortex or hippocampus proper/subiculum in the activation of target regions of the hippocampal formation is the transection of the perforant pathway (Angenstein et al., 2007), or stimulation within the hippocampus proper (see Section 3.2).

Significant BOLD responses in target regions of the hippocampal formation may not be detectable during the initial perforant pathway stimulation but may develop during repetitive stimulations (Helbing et al., 2013, 2016) or, vice versa, may present at the beginning and disappear during subsequent stimulations (Riemann et al., 2017). These variations are caused by stimulus-induced changes in signal processing either within the hippocampal formation (and by that altered output activity) or in the target region itself. It should be emphasized that absence of significant BOLD responses in target regions of the hippocampal formation indicates, by no means, an absence of altered neuronal activities; it only indicates that the hemodynamic in this region is not altered.

3.1.2 Activation of Modulatory Transmitter Systems

The experimental setup may also be suited to study the effect of modulatory transmitter systems on hippocampal signal processing. Based on the assumption that fMRI-BOLD responses rather reflect the quality of incoming signal processing than sole incoming and/or outgoing activity, any modulatory effect on hippocampal signal processing may also become detectable by fMRI. The hippocampal formation is controlled by various modulatory transmitter systems, such as the cholinergic, dopaminergic, serotonergic, and histaminergic systems. In general, these modulatory transmitter systems can be directly activated by either electrical or optogenetic stimulation of the region of origin. For example, activation of the mesolimbic dopamine system can be triggered by electrical stimulation of the ventral tegmental area (VTA) or by expression of laser light–sensitive opsins in dopaminergic neurons in this structure (Helbing et al., 2016; Scherf and Angenstein, 2017). Alternatively, specific agonists or antagonists can be applied to mimic an increase or to inhibit the current activity of these systems (Helbing et al., 2017). To test the efficacy of a modulatory transmitter system in modifying signal processing in the hippocampal formation, the perforant pathway is stimulated again with a specific stimulation protocol. This is done in the absence or presence of activators/inhibitors of the modulatory transmitter system. An effect of modulatory transmitter systems on signal processing (measured as an altered BOLD response during activation of a modulatory transmitter system) can either occur within the hippocampal formation or in target regions of the hippocampal formation. Whereas putative changes of BOLD responses within the hippocampal formation obviously indicates altered signal processing in the same structure, changed BOLD responses in target regions of the hippocampal formation during activation of a modulatory transmitter system may indicate an altered activity of hippocampal projecting neurons, and/or an altered signal processing in the target regions. If an activation of one modulatory transmitter system results in no detectable changes in BOLD signal intensities, the interpretation is ambiguous. As mentioned above, similar BOLD responses can be induced by different neuronal activation patterns, thus there might be an alteration in signal processing during activation of a modulatory transmitter system; but this has not necessarily an effect on local hemodynamic parameters. Again, only electrophysiological recordings in the appropriate region can confirm the presence or absence of altered neuronal activity.

View chapterPurchase book

Hippocampal Formation

Menno P. Witter, David G. Amaral, in The Rat Nervous System (Third Edition), 2004

Pre- and parasubiculum

As described above, a minor component of the perforant pathway courses through the molecular layer of the para- and presubiculum. In addition, fibers from cells in layer Va and, to a lesser extent, in layers III, Vb, and VI of the entorhinal cortex terminate weakly in layer I of the presubiculum and parasubiculum (Köhler, 1986, 1988; Van Groen and Wyss, 1990a, 1990c). This modest projection from the entorhinal cortex to the pre- and parasubiculum stands in marked contrast to the previously described dense projections to the entorhinal cortex from these two areas (see sections “Presubiculum” and “Parasubiculum”), indicating that the pre- and parasubiculum should be considered functionally as input structures to the entorhinal cortex.

View chapterPurchase book

Granule Cells: Granule-Cell Properties in Seizure Discharge☆

H. Beck, ... C.P. Fitzsimons, in Reference Module in Neuroscience and Biobehavioral Psychology, 2017

Background—Functional Anatomy of the Dentate Gyrus

The dentate gyrus receives its major input from the entorhinal cortex (perforant pathway) via glutamatergic synapses. The perforant path axons terminate on the dendritic tree of granule neurons within the molecular layer. The dentate granule cells (DGCs) project via the distinctive, nonmyelinated mossy fibers to the hilar and CA3 regions. Mossy fibers form en-passant synapses onto the proximal dendrites of the pyramidal neurons of the CA3 region. In the CA3 region, an intriguing feature of mossy fiber boutons is that they also contact numerous interneurons with filopodial-like processes. In fact, mossy fibers in the normal CA3 region contact many more inhibitory interneurons than CA3 pyramidal cell dendrites. Thus, the output of dentate granule neurons carries a powerful inhibitory component.

In addition to innervating CA3 neurons, mossy fibers collateralize in the hilar region and innervate various target-cell types. These cells may be glutamatergic (i.e., mossy cells) or one of numerous classes of interneurons utilizing the inhibitory neurotransmitter γ-aminobutyric acid (GABA). GABAergic neurons in the hilus and in other areas of the dentate gyrus provide a remarkably strong inhibition in this region. They can be subclassified into numerous categories on the basis of their soma location and dendritic arborization patterns, their axonal termination patterns, as well as their characteristic spiking patterns following depolarization. In addition, neurochemical criteria have been used to refine the definition of interneuron subclasses. In this context, a major criterion with perhaps the highest degree of functional relevance is onto which principal neuron compartments interneurons form synapses. it is important to note that the presence of different types of interneurons allows GABA release onto DGC soma and dendrites with precise spatiotemporal control, mediating powerful feed-forward and feed-back inhibition. In particular, recent observations have shown that parvalbumin-expressing interneurons in the DG suppress adult quiescent neural stem cell activation and project immature GABAergic synaptic connections onto proliferating hippocampal neural precursors, promoting the survival and development of their newborn neuronal progeny. These observations suggest that parvalbumin-expressing interneurons may be involved in aberrant induction of adult hippocampal neurogenesis associated with epilepsy.

Multiple lines of evidence, including studies of lesions of entorhinal cortex or perforant pathway (by lesioning cell bodies or axons/terminals transporting APP to terminals, respectively), indicate that removing the source of Aβ significantly reduces Aβ in target fields. Similarly, increasing local levels of degrading enzymes.