Microcircuit‐specific processing in the hippocampus

How does the hippocampus process information originating from the entorhinal cortex and transform it into the integrated output that ultimately underlies its functions in vivo? How does this type of activity remain within physiological boundaries despite the many recurrent excitatory loops that create a risk for epileptic seizures? These questions are important because entorhinal–hippocampal circuits serve as major centres for spatial navigation and memory and play key roles in temporal lobe epilepsy and other neurological disorders. Although the fundamental wiring diagram of the synaptic connections between entorhinal afferents and hippocampal principal neurons (i.e. pyramidal and granule cells) has been known since the pioneering studies of Ramon y Cajal, additional synapses and neuronal types play critical roles in the integrative process that shapes the physiological activity of specific assemblies of hippocampal principal cells in vivo. In particular, despite being a numeric minority and accounting for only 15–20% of the total hippocampal cell population, many diverse types of non-principal neurons are involved in several key processes including feedback and feedforward inhibition, the generation of oscillatory rhythms, the control of synaptic plasticity and excitation–neurogenesis coupling. Non-principal cells are mostly GABAergic interneurons, with very specific anatomical and physiological properties, and strategically located at key points in the hippocampal network. The fast integrative functions of other non-principal neurons such as Cajal-Retzius cells remain rather mysterious, but the co-alignment of their dendrites and axons with entorhinal afferent inputs suggests that they may also play a role. How these non-principal cells shape the properties of the entorhinal–hippocampal synaptic dialogue is critical to understanding the network mechanisms that control populations of principal neurons and prevent the onset of pathological activity. This requires progress on several levels. First, a rigorous definition of the anatomical and physiological properties of the many diverse non-principal cell subtypes is needed. Second, attention must be paid to how flexible specific microcircuits are, and how they can respond to neuromodulation occurring during different brain states in vivo. Third, experimental methods capable of directly observing the functions of microcircuits at high temporal and spatial resolution need to be developed and used. Fourth and last, the insights into the functions of microcircuits obtained from in vitro studies need to be integrated within the framework of experimental results obtained in vivo, and new testable hypotheses on their functions need to be proposed. To highlight recent progress in these specific points, The Journal of Physiology organized a symposium entitled ‘Microcircuit-specific processing in the hippocampus’ at the 2010 Society for Neuroscience Meeting in San Diego, California. This symposium brought together four leading speakers of this field, all of whom have made major contributions to our understanding of the entorhinal–hippocampal microcircuits. This issue of The Journal of Physiology publishes Symposium Reports from each of these speakers. Afferent inputs from the entorhinal cortex to the CA1 region are mostly localized in the stratum lacunosum-moleculare. Dr Capogna from the Medical Research Council, UK reviews his recent findings on the detailed anatomical and physiological properties of a specific GABAergic interneuron, the neurogliaform cell (Capogna, 2011). This interneuronal subtype is particularly abundant in the stratum lacunosum-moleculare and can be activated by the entorhinal cortex via the direct temporoammonic input (entorhinal–CA1 synapse). His results indicate that neurogliaform cells have unique output properties. In particular, he shows that neurogliaform cells generate slow synaptic inhibition mediated both by GABAA and GABAB receptors. The slow kinetics of the GABAA receptor-mediated inhibitory postsynaptic currents are likely to be due to a peculiar spatiotemporal profile of extracellular GABA released by these neurons, which produces a prolonged, low-level GABA transient at these synapses. This interpretation is well supported both by pharmacological experiments and modelling work. Dr Capogna proposes that the compact dendritic tree of these cells, coupled to their dense axonal arborization, makes them a prototypical feedforward interneuron very well suited to modulate entorhinal–hippocampal interactions. He concludes by suggesting that processing of entorhinal input by neurogliaform cell-mediated slow inhibition is involved in setting the temporal delay between population activities in the entorhinal cortex and in the CA1 hippocampus, which is observed during theta oscillations in vivo. Neuromodulation of networks stratum lacunosum-moleculare networks is reviewed by Dr Maccaferri, Northwestern University, Chicago, IL, USA (Maccaferri, 2011). Particular emphasis is put on recent work on two different types of modulators: the catecholamine noradrenaline, and the chemokine stromal cell-derived factor 1α (SDF-1), which impact GABAergic microcircuits and Cajal-Retzius cells, respectively. Based on the effects of noradrenaline on the membrane potential and electrical synapses of different types of interneurons, he proposes that high and low levels of noradrenergic modulation may control the relative weights of feedforward vs. feedback inhibition of temporoammonic transmission. Noradrenergic modulation of GABAergic networks is then contrasted to the effects of SDF-1 on Cajal-Retzius cells. This neuronal subtype is shown to possess spontaneous firing activity, which is powerfully depressed by SDF-1. The intriguing correlation between inhibition of tonic firing in these neurons following exposure to SDF-1 and reduced transmission strength at the entorhinal–CA1 synapse suggests that spontaneous activity of Cajal-Retzius cells may regulate the entorhinal–hippocampal connections by releasing a yet unidentified neurotransmitter. The relevance of this modulation for disease is further discussed in the view that levels of SDF-1 are known to be affected in various pathological conditions. The role of specific microcircuits in preventing pathological epileptiform activity is reviewed by Dr Coulter and colleagues from the University of Pennsylvania, Philadelphia, PA, USA (Coulter et al. 2011). By taking advantage of patch clamp recordings combined with fast voltage-sensitive dye imaging in vitro, the authors highlight the highly controlled nature of entorhinal–hippocampal transmission to both the dentate gyrus and the CA1 area. Their data clearly indicate that stimulation of entorhinal cortical inputs to the dentate gyrus and to the CA1 subfield are usually ineffective in making granule or pyramidal cells fire. This is due to a combination of the intrinsic properties of dentate granule cells and distal dendrites of pyramidal neurons, together with the powerful activation of feedforward and feedback inhibitory networks. Feedforward inhibition activated by the temporoammonic pathway does not affect greatly the temporoammonic excitatory postsynaptic potential at the distal dendrites, but it does impair its propagation to more proximal compartments. In contrast, feedback microcircuits activated by properly patterned stimulation can directly suppress excitatory postsynaptic potentials measured with fast voltage-sensitive dye imaging at distal dendrites. Nevertheless, the tight control of entorhinal–hippocampal transmission by these microcircuits can be released under appropriate circumstances. For example, the integration of incoming transmission from the Schaffer collaterals with temporoammonic inputs provides a window of opportunity for effective propagation of temporoammonic excitatory postsynaptic potentials, which can now reach the soma. The specific vulnerability of inhibition during temporoammonic–CA1 transmission in models of temporal lobe epilepsy is further discussed and contrasted with preserved dentate granule filtering functions, most likely due to compensatory GABAergic responses. The reluctance to fire of granule cells observed both in vivo and in vitro is the basis of a new hypothesis proposed by Dr John Lisman from Brandeis University, Watham, MA, USA (Lisman, 2011) regarding the role of neurogenesis in the dentate gyrus. He starts by highlighting the rather puzzling observation that ∼95% of dentate granule cells do not fire in any environment in vivo, despite the continuous input received from the entorhinal cortex. Dr Lisman proposes that the interaction between constantly active entorhinal input, powerful inhibitory circuits and bi-directional synaptic plasticity results in a winner take all process. The most excited cells fire and therefore strengthen (most of the time, but not always) their synaptic weights, whereas cells that do not fire will progressively weaken their synaptic inputs and become part of a non-functional pool of neurons. Once in this pool, these cells will never have the chance of back-transitioning to the functional pool. In contrast, because the possibility of depression of synaptic weights in winners cannot be completely eliminated, a progressive ‘leak’ of cells from the functional to the non-functional pool would be expected. In the absence of a replenishing process, the functional pool could be greatly decreased or even completely lost. Therefore, Dr Lisman proposes the intriguing hypothesis that neurogenesis may specifically prevent this occurrence by providing newly generated cells with more excitable properties, which would have a good chance of entering the functional pool and maintaining a physiological balance between functional and non-functional neurons. Lastly, he suggests a few key experiments specifically designed to test this very intriguing hypothesis. As this brief overview shows, hippocampal processing of incoming information from the entorhinal cortex requires the cooperative activity of several complex microcircuits. Understanding the role of each microcircuit is an exciting challenge, which will maintain this lively field engaging and prosperous for the foreseeable future.