Hippocampal connectivity: within and between brain regions

Within the hippocampal formation, DG and the CA regions form a trisynaptic circuit first described by Santiago Ramon y Cajal (Figure 4). The circuit forms a loop where information through principal glutamatergic cells travels from the EC to DG, from DG to CA3 and from here to CA1 through synaptic transmission between principal cells of these regions. The classical description consists of three major projections: 1) The Layer II neurons form the entorhinal cortex transmit polymodal sensory information by projecting and making synapses with granule cells dendrites in the DG via the perforant path. 2) Granule cells project their axons, the mossy fibers, to CA3 where the contact dendrites of pyramidal cells. 3) CA3 project ipsilaterally to pyramidal cells of the CA1 via Shaffer collaterals and contralaterally through commissural fibers that contact CA3 and CA1 pyramidal cells. The trisynaptic pathway of the hippocampus is one of the most studied region over the last two decades. It is the major brain site used to study synaptic plasticity mechanisms and memory processes (developed in chapter 4).

Figure 4. Basic circuits of the hippocampus, adapted from Neves et al. (2008).

The diagram shows the traditional trisynaptic pathway. The perforant path conveying polymodal sensory information from entorhinal cortex layer II neurons to the dentate gyrus carries out the main input to the hippocampus. Perforant path axons make excitatory synaptic contact with granule cells’ dendrites: axons from the lateral and medial entorhinal cortices innervate the outer and middle third of the dendritic tree, respectively. Granule cells send their projections (the mossy fibers), to the proximal apical dendrites of CA3 pyramidal cells. CA3 principal cells project to ipsilateral CA1 pyramidal cells through Schaffer collaterals and to contralateral CA3 and CA1 pyramidal cells through commissural connections. Additionally, there is also an important associative network interconnecting CA3 cells on the same side. CA3 pyramidal cells are also innervated by a direct input from entorhinal cortex layer II neurons (not shown). The distal apical dendrites of CA1 pyramidal neurons receive a direct input from entorhinal cortex layer III neurons. There is also substantial modulatory input to hippocampal neurons.

Hippocampus Entorhinal cortex of synaptic plasticity, so that subsequent activation by incoming stimulation of only a sub-component of the assembly will lead to activation of the whole assembly, thereby recapitulating the activity elicited by the original event. (LTP is a Hebbian process, since its induction requires coincident activity of the pre- and postsynaptic neu-rons.) The immediate problem is to identify such cell assemblies in the hippocampal encoding of memory.

Place cells

Single-unit recordings from neurons in the hippocampus of freely moving rodents reveal that pyramidal and granule cells show a preference for firing in a particular loca-tion of an explored environment, regardless of the direction from which the animal enters the location^{33(BOX 1)}. Hundreds of such ‘place cells’ fire in concert as a rat reaches a particular location, and place cells fire in sequence as the animal moves Synaptic plasticity in the hippocampus

The hippocampus has been a major experimental system for studies of synaptic plasticity in the context of putative informa-tion-storage mechanisms in the brain. Its simple laminar pattern of neurons and neural pathways ^{(FIG. 1)} enables the use of extracellular recording techniques to record synaptic events for virtually unlimited peri-ods in vivo¹². The much-studied model of synaptic plasticity, long-term potentiation^13,14 (LTP; see ^{FIG. 2a}), was first identified in the hippocampus and has been extensively characterized using electrophysiological, biochemical and molecular techniques¹⁵. Several recent studies have detected LTP-like synaptic changes in the hippocam-pus^{16,17(FIG. 2b)} and the amygdala¹⁸ following learning. Other forms of activity-dependent plasticity have been found, including long-term depression (LTD)¹⁹, EPSP-spike (E-S) potentiation^20,21, spike-timing-dependent plasticity (STDP)²², depotentiation^23–25 and de-depression^25,26. The transverse hippo-campal slice preparation^{27(FIG. 2a)} has been of major importance to this field, enabling

pharmacological agents to be rapidly washed on and washed off and allowing intracellular and patch-clamp recordings.

In addition, hippocampal neurons can be cultured^28,29, either as transverse ‘organo-typic’ slices or as populations of dissociated neurons, for periods of months, facilitating molecular manipulations such as over-expression or RNAi-based knock-down of specific proteins. These in vitro techniques have greatly enhanced our understanding of the molecular mechanisms that underlie synaptic plasticity^15,30. In the hippocampus it has been possible to track effects such as the phosphorylation of a protein at a specific residue at multiple levels of organization, from isolated synaptic membranes all the way through to the behavioural analysis of intact animals with specific molecular defects³¹. Nevertheless, the larger picture of how synaptic plasticity in extensive networks of cells leads to the storage and recall of information remains dimly illumi-nated. The Canadian psychologist Donald Hebb posited a role for such assemblies as engrams or memory traces³². His famous Figure 1 | Basic anatomy of the hippocampus. The wiring diagram of the

hippocampus is traditionally presented as a trisynaptic loop. The major input is carried by axons of the perforant path, which convey polymodal sensory information from neurons in layer II of the entorhinal cortex to the dentate gyrus. Perforant path axons make excitatory synaptic contact with the dendrites of granule cells: axons from the lateral and medial entorhinal cortices innervate the outer and middle third of the dendritic tree, respec-tively. Granule cells project, through their axons (the mossy fibres), to the proximal apical dendrites of CA3 pyramidal cells which, in turn, project to ipsilateral CA1 pyramidal cells through Schaffer collaterals and to contra-lateral CA3 and CA1 pyramidal cells through commissural connections. In

addition to the sequential trisynaptic circuit, there is also a dense associa-tive network interconnecting CA3 cells on the same side. CA3 pyramidal cells are also innervated by a direct input from layer II cells of the entorhinal cortex (not shown). The distal apical dendrites of CA1 pyramidal neurons receive a direct input from layer III cells of the entorhinal cortex. There is also substantial modulatory input to hippocampal neurons. The three major subfields have an elegant laminar organization in which the cell bodies are tightly packed in an interlocking C-shaped arrangement, with afferent fibres terminating on selective regions of the dendritic tree. The hippocampus is also home to a rich diversity of inhibitory neurons that are not shown in the figure. For a full description of hippocampal anatomy, see REF. 90.

P E R S P E C T I V E S

66 | JANUARY 2008 | VOLUME 9 www.nature.com/reviews/neuro

“where”

“what”

Entorhinal cortex

The major input of the hippocampal formation is from the EC. In 1893, Ramón y Cajal first described projections from the entorhinal cortex, in particular layer II, to the dentate gyrus and named it the perforant pathway. Layers III, V and VI also project to the hippocampal formation but to a lesser extent. These projections also target the CA fields and the subiculum, the temporoammonic pathway. Perforant path terminals specifically target the apical dendrites and form synapses with principal cells and interneurons in the molecular layer of the DG or the stratum lacunosum-moleculare of CA3 (Nafstad, 1967; Hjorth-Simonsen and Jeune, 1972). Thus, subsets of cells that do not have dendrites in this specific layer are not innervated by these projections. This is particularly the case of hilar mossy cells that receive little or no perforant input. The MEC and LEC projections share similar anatomical targets though not completely identical. In particular, layer II of LEC projections target the outer third of the molecular layer of the DG, while MEC project more to the middle third of the same layer. Similarly, layer II of MEC projects deeper in the stratum lacunosum-moleculare of the CA3 whereas LEC projects more superficially in this layer. Another difference is that layer III of MEC projects to the proximal stratum lacunosum-moleculare of CA1 and distal molecular layer of subiculum while the same layer of LEC project to distal CA1 and proximal subiculum.

Dentate Gyrus

DG projects to the CA3 via granule cells axons, the mossy projections. Mossy and GABAergic cells of the hilus project their axons to the molecular layer of both ipsilateral and contralateral DG. The DG receive different projections from different brain areas, but there is no evidence for DG to project outside of the CA areas. The major input to the DG is the projection form the layer II of the EC, the perforant pathway. DG receive septal projections arising from the medial septal nuclei (Mosko et al., 1973; Baisden et al., 1984);

(Amaral and Kurz, 1985; Wainer et al., 1985; Nyakas et al., 1987). Septal projections target the dentate gyrus closely to the granule cells layer and principally arise from the medial septal nucleus and the diagonal band of Broca. Most of these projections are cholinergic, but some are GABAergic. This heterogeneity matches the anatomical segregation where septal GABAergic cells projects preferentially to other GABAergic cells of the dentate gyrus whereas septal cholinergic cells mainly project to the molecular layer of the dentate gyrus where they contact dendrites of the granule cells. Interestingly, only around 5-10% of cholinergic projections contact interneurons. Other inputs to the molecular layer of the DG arise from the hypothalamus, specifically from the glutamatergic

neurons of the supramammilary area (Wyss et al., 1979; Vertes et al., 1993; Magloczky et al., 1994; Kiss et al., 2000). Also, the noradrenergic system from the locus coeruleus of the pons terminates in the polymorphic layer (Swanson and Hartman, 1975; Loughlin et al., 1986). DG also receive dopaminergic projections from ventral tegmental area (VTA) as well as GABAergic projections (Ntamati and Luscher, 2016). Another mid brainstem projection is the serotonergic system originating from the raphe nuclei that target the DG (Moore and Halaris, 1975; Kohler and Steinbusch, 1982; Vertes et al., 1999). Most of cells targeted by this input are GABAergic interneurons, basket cells as well as calbinding positive cells. There is significant intrinsic connectivity within the DG. This connectivity occurs mainly between principal granule cells and different types of glutamatergic and GABAergic cells giving rise to the different form of activity further described in chapter 5 (feedforward inhibition, feedback inhibition and disinhibition).

CA3

CA3 predominantly projects to CA1 principal cells, through the Shaffer collaterals. CA3 receives projections from the DG mossy fibers. The classical view is that CA3 is highly innervated by its own collateral axons (associational connectivity) and receive also contralateral CA3 connections (commissural connectivity). However, this view has been recently challenged by the study of Guzman et al. where they showed, by using electrophysiological ex-vivo slice recordings, that CA3 associational connectivity is rather sparse (Guzman et al., 2016). There is evidence for pyramidal neurons of CA3 to project back to the DG molecular layer (Laurberg, 1979; Li et al., 1994). The noradrenergic system from the locus coeruleus of the pons projects to the stratum lucidum of CA3 (Pickel et al., 1974; Swanson and Hartman, 1975; Loughlin et al., 1986). The basal nucleus of the amygdaloid complex mostly sends its projections to stratum oriens and radiatum of the CA3 (Pikkarainen et al., 1999; Pitkänen, 2000). The cholinergic projection from the septum also targets the CA3. CA3 bilaterally projects back to the lateral septal nucleus via the fimbria.

CA1

CA1 main inputs come from the CA3 projections (Shaffer collaterals) and the EC layer III, (temporoammonic pathway). The nucleus reuniens of the thalamus sends mainly glutamatergic projections to the stratum lacunosum-moleculare of CA1 where they contact apical dendrites from pyramidal neurons and GABAergic cells (Herkenham, 1978;

Dolleman-Van der Weel and Witter, 2000). Principal cells of CA1 project to a variety of

different regions comprising the subiculum, the entorhinal cortex, the perirhinal cortex, the retrosplenial cortex, the medial temporal lobe, the amygdaloid complex, the septal nucleus. Proximal principal CA1 pyramidal cells project to the MEA while distal pyramidal cells connect to the LEA. Connectivity to the amygdaloid complex consist of projections in order of its importance to the basal nucleus, the amygdalohippocampal area and all the nuclei including the lateral, central, accessory basal, medial and posterior cortical (Pitkänen, 2000). Differently form CA3 principal cells, CA1 pyramidal neurons have weak associational projections. Interestingly, evidence revealed that CA3 receive backprojections form the CA1 inhibitory cells found in the stratum oriens and radiatum (Laurberg, 1979; Amaral et al., 1991; Swanson et al., 1981).