The entorhinal cortex is part of the medial temporal lobe or hippocampal memory system and constitutes the major gateway between the hippocampal formation and the neocortex. The entorhinal cortex has initially attracted attention because of its strong reciprocal connections with the hippocampal formation and its involvement in certain brain disorders.
Contents |
The entorhinal cortex (Brodmann area 28; Brodmann, 1909) constitutes the major gateway between the hippocampal formation and the neocortex. The name entorhinal (inside rhinal) cortex derives from the fact that it is partially enclosed by the rhinal (olfactory) sulcus. Together with the hippocampal formation and neighboring portions of the parahippocampal region it forms a major substrate mediating conscious (declarative) memory.
Interest in the entorhinal cortex arose around the turn of the 19th century when Ramón y Cajal in his seminal studies on the anatomy of the nervous system described a peculiar part of the posterior temporal cortex which was so strongly connected to the hippocampal formation that he suggested that the physiological significance of the latter structure would relate to that of the entorhinal cortex, and at that time he assumed that the entorhinal cortex, and therefore the hippocampal formation, was part of the olfactory system, processing smell information (Ramón y Cajal, 1902). This latter functional connotation did not fit well with ideas emerging in the late nineteen-fifties that the hippocampus was a main player in conscious memory processes in humans (Scoville and Milner, 1957). Interest in the entorhinal cortex was kindled again by reports describing widespread reciprocal cortical connections of the entorhinal cortex in monkeys in the nineteen-seventies (Van Hoesen and Pandya, 1975; Van Hoesen et al., 1975) and spurred even more by reports that the entorhinal cortex is a site of early pathology in Alzheimer’s Disease (Braak and Braak, 1985). The discovery of spatially modulated cells (Fyhn et al., 2004), later known as grid cells (Hafting et al., 2005), sparked a renewed interest.
By now it is well accepted that the entorhinal cortex is part of a strongly interconnected set of cortical areas that together form the parahippocampal region, which in turn is closely associated with the hippocampal formation on the one hand and with a variety of multimodal association areas of the cortex such as parietal, temporal, and prefrontal cortex (Witter et al., 1989; Burwell et al., 2002). The entorhinal cortex is thus uniquely positioned as an interface between the neocortex and the hippocampal formation (Buzsaki, 1996; Lavenex and Amaral, 2000; Witter et al., 2000b).
Prominent species differences are apparent with respect to surface area and complexity of the cortical mantle, but the anatomy and overall functional role of the entorhinal cortex appear to be largely conserved throughout the animal realm. The entorhinal cortex is generally subdivided into two domains, the so-called lateral and medial entorhinal cortices (Figure 1; a note of caution: multiple, quite different schemes to subdivide the entorhinal cortex have been proposed, occasionally resulting in substantial confusion among students of this part of the cortex (Witter et al., 1989). The two subdivisions, initially differentiated on the basis of overall differences in morphological features such as types of neurons, packing density and cell sizes and shapes (Brodmann, 1909; Krieg, 1946) are characterized by largely different input-output connectivity (Witter et al., 2000a). The lateral entorhinal cortex, for example, is strongly connected to the perirhinal cortex, olfactory and insular cortex and the amygdala. The medial entorhinal cortex preferentially connects with the postrhinal cortex, the presubiculum, visual association (occipital) and retrosplenial cortices. The two parts of entorhinal cortex also connect differentially to the hippocampal formation. Both target the same neurons in the dentate gyrus and field CA3, while they reciprocally connect to different groups of cells in CA1 and subiculum (Figure 1B).
The connectivity of the entorhinal cortex shows a strong topographical organization along a gradient that runs from its border with the rhinal sulcus towards the border between the entorhinal cortex and the adjoining hippocampal formation. Multimodal sensory cortex providing inputs representing the outside world preferentially target a strip of entorhinal cortex which is adjacent to the rhinal sulcus, including portions of both the lateral and medial divisions. Parts of the two divisions that are more distant from the rhinal sulcus receive a different class of information, for example from olfactory areas and the central and medial components of the amygdala. Outputs of the entorhinal cortex are similarly organized (Figure 1E). Interestingly, the intrinsic connectivity of the entorhinal cortex is biased to connect portions that belong to the same strip and to connect deep to superficial layers (Figure 1E).
The different sets of inputs are mapped onto the hippocampus with a striking topographical organization as well, such that the zone closest to the rhinal fissure preferentially connects with the dorsal (non-primate) or posterior (primate) part of the hippocampus whereas portions of the entorhinal cortex positioned more distant from the rhinal fissure connect more to ventral (non-primate) or anterior (primate) part of the hippocampus (Figure 1C and D; Witter et al., 2000b).
Similar to other cortices, in the entorhinal cortex, neurons are grouped into different layers that are characterized by a dominant cell type. Six layers are commonly distinguished, of which layers I and IV are relatively free of neurons. The principal neurons of the entorhinal cortex, i.e., the neurons that are among the main recipients of incoming axons and constitute the major source of entorhinal output to a variety of cortical and subcortical structures, are generally pyramidal cells or modified versions, the so-called stellate cells (in LEC these are often referred to as fan cells; Canto et al., 2008; Moser et al., 2010). These mainly utilize glutamate as an excitatory neurotransmitter. A second group of neurons are the interneurons that mainly provide intrinsic, local connections that use GABA as an inhibitory transmitter. Recently, a third group of excitatory local neurons have been described as well (Figure 2).
The layered organization of the entorhinal cortex looks deceivingly simple, but in reality is rather complicated. Cortical inputs mainly target neurons in layers II and III, which give rise to the so-called perforant path, projecting to all subdivisions of the hippocampal formation, providing them with their major cortical input. Layer II neurons mainly project to the dentate gyrus and hippocampal field CA3, and cells in layer III distribute their axons largely to field CA1 and the subiculum. Subcortical inputs, such as cholinergic and monoaminergic inputs from the septal complex and brainstem and inputs from the thalamus, amygdala, and claustrum exhibit an overall diffuse terminal distribution in the entorhinal cortex, although some prevalence is apparent, that, with the exception of inputs from the amygdale, has not been assessed in detail yet. Output of the hippocampal formation preferentially targets layers V and VI of the entorhinal cortex, which in turn are the origin of widespread reciprocal cortical projections and subcortical projections to the septum, striatum, amygdala and thalamus. Although this apparent separation between input and output layers provides a simple functional concept, it has recently been challenged by a number of findings indicating that reciprocal interactions between deep and superficial layers are quite substantial, and that main cortical inputs also target apical dendrites of the deeper located neurons (Figure 1; Canto et al., 2008).
The entorhinal cortex is a relevant node in the network mediating learning and memory. However, the unique contribution of the entorhinal cortex to higher order cortical processing is as yet only partially understood. The entorhinal cortex, in conjunction with the hippocampal formation, appears to specifically deal with the translation of neocortical exteroceptive information into higher order complex representations that, when combined with motivational and interoceptive representations, will serve cognitive functions, in particular conscious memory (Eichenbaum et al., 2007). The overall differences in cortical connectivity between the lateral and medial entorhinal cortex are reflected in recent findings that the medial entorhinal cortex, but not the lateral, is a major hub in the brain’s circuitry for spatial navigation. A key component of this network is the grid cell (Hafting et al., 2005). When rats run around in two-dimensional environments, grid cells express a specific pattern of firing that tiles the entire environment covered by the animal, almost like the cross points of graph paper, but with an equilateral triangle as the unit of the grid rather than a square. It has been proposed that this universal spatial representation might be recoded onto a context-specific code in hippocampal networks, and that this interplay might be crucial for successful storage of episodic memories (Fyhn et al., 2007). Grid cells are predominant in layer II of the entorhinal cortex, but exist also in layers III and V. Grid cells in layers III and V intermingle with cells that code for the direction the animal is looking, the head-direction cells, as well as cells with conjunctive grid and head-direction properties (Sargolini et al., 2006) and border cells (Savelli et al., 2008; Solstad et al., 2008). It is quite likely, although not unequivocally established, that the directional firing of entorhinal cells depends on signals from the head direction cells in the presubiculum. In the lateral entorhinal cortex, neurons exhibit little spatial modulation (Hargreaves et al., 2005), but many respond to olfactory, visual, or tactile stimuli with a high degree of selectivity (Eichenbaum et al., 2007).
The notion of two functionally different parts of the entorhinal cortex should be tempered in view of the strong connectivity between the lateral and medial entorhinal cortex. It is thus likely that at the level of the entorhinal cortex relations between the two sets of inputs will already occur. This is in line with findings that in the entorhinal cortex cells respond to both object and place stimuli and that lesions of the entorhinal cortex do not result in impairments in for example object recognition but do impair the relational organization of memory (Eichenbaum et al., 2007).
Entorhinal functions are most likely modulated by a number of subcortical inputs, including those from the thalamic midline structures, and cholinergic inputs. High levels of acetylcholine might set the appropriate dynamics to facilitate the storage of stimuli, whereas removal of the cholinergic inputs to the entorhinal cortex dramatically interfere with the memory performance of the animal (Hasselmo, 2006). These subcortical inputs also play an essential role in occurrence of oscillatory activity that is an elementary component of normal entorhinal function (Lopes da Silva et al., 1990; Mitchell et al., 1982). It is debated whether or not the basal forebrain dependent theta rhythm in entorhinal cortex is required for the presence of stable grid cells (Brandon et al., 2011; Koenig et al. 2011; Yartsev et al., 2011). Functional imaging experiments in humans have as yet failed to provide clear indications for specific entorhinal contributions to memory processes, but no doubt with time and increased sensitivity, functional notions emerging from animal studies will be testable in healthy human subjects.
Severe alteration of the entorhinal cortex is associated with several disorders of the human brain, importantly Alzheimer’s disease, temporal lobe epilepsy and schizophrenia. In case of Alzheimer's disease, the initial pathological changes reportedly occur in layer II of the entorhinal cortex (Braak and Braak, 1985), and volume reduction of the entorhinal cortex is now considered a relevant and reliable measure to identify individuals at risk for Alzheimer’s disease. Entorhinal atrophy is associated with mild memory loss as seen in individuals with mild cognitive impairment and it precedes hippocampal volume reduction seen in Alzheimer patients (deToledo-Morrell et al., 2004). Temporal lobe epilepsy is associated with marked degeneration in layer III (Du et al., 1993), and in case of schizophrenia an overall miss-wiring of the entorhinal cortex or volume reductions have been proposed as a possible contributing factor (Arnold, 2000; Baiano et al., 2008).
The entorhinal cortex has been a focus for research in the early periods of neuroanatomy. Subsequently, interest diminished but was kindled again in the late 70s of the last century. With the discovery of spatially modulated cells, such as the grid cells in the medial entorhinal cortex and the striking involvement of entorhinal cortex in a variety of brain diseases, the interest has become stronger than ever. What still needs to be established though is an overarching concept of its functional relevance, taking into account the striking differences between the lateral and medial entorhinal subdivisions, the fact that these two are interconnected by way of the well developed intrinsic associational connections that appear to be organized in a way that is in concert with the topographical organization of entorhinal-hippocampal reciprocal connections. Such an overarching functional notion is a requirement to efficiently probe the functional relevance of the entorhinal cortex in humans.