Basic Neurosciences · Paper A
Hippocampal anatomy
Follow the floor of the temporal horn of the lateral ventricle and your finger traces the hippocampus, a curled strip of grey matter with essentially one road running through it. Fresh experience arrives from the cortex, is bound into an episode along that road, and leaves again for storage elsewhere. Traffic runs overwhelmingly in one direction, which makes the structure efficient and fragile in the same breath: interrupt the road anywhere and new declarative memories stop being made, while memories that left years ago sit safely in the cortex, untouched.
A stem naming the sequence entorhinal cortex, dentate gyrus, CA3, CA1 and subiculum, or asking which subfield suffers most after a hypoxic event such as a cardiac arrest, is testing the internal anatomy directly. When the patient in front of you has medial temporal damage — surgical, hypoxic, or from herpes encephalitis — and cannot form new memories yet still rides a bicycle and recalls childhood clearly, you are looking at the anterograde-predominant pattern with its temporally graded retrograde tail, which you separate from a Korsakoff presentation by the absence of confabulationFilling a gap in memory with a made-up answer the person genuinely believes is true, not a deliberate lie.Read more → and by the lesion site if one is given.
CA1 vs dentate gyrus vulnerability — CA1 is the subfield most easily lost to hypoxic-ischaemic injury, while the dentate gyrus is comparatively resistant.,Anterograde vs retrograde amnesia after hippocampal damage — anterograde loss dominates and is the more severe deficit, while the retrograde component is temporally graded and spares older memories.,Declarative vs procedural memory — hippocampal damage takes conscious recall of facts and events, while skills and habits survive because they depend on the basal gangliaA group of deep brain structures that help control movement and are affected in Parkinson's disease and by antipsychotic drugs.Read more → and cerebellumThe brain structure at the back of the skull that coordinates balance and smooth movement.Read more →.,Hippocampal amnesia vs Wernicke-Korsakoff syndromeA chronic memory disorder, usually from thiamine deficiency, marked by severe short-term memory loss and confabulation.Read more → — hippocampal amnesia follows focal medial temporal damage, whereas Korsakoff's follows thiamine deficiency damaging the mammillary bodies and dorsomedial thalamusA deep brain structure that relays and filters sensory information on its way to the cortex.Read more →, and confabulationFilling a gap in memory with a made-up answer the person genuinely believes is true, not a deliberate lie.Read more → is classical there but not a defining feature of a pure hippocampal lesion.,Trisynaptic circuit vs Papez circuit — the trisynaptic circuit is the internal relay inside the formation, while the Papez circuit is the larger loop joining the hippocampusA brain structure central to forming new memories.Read more → to the mammillary bodies, anterior thalamus and cingulate cortex.
The classic error is scrambling the route, or dropping the entorhinal cortex from either end of it, since stems test the exact order rather than the bare list of structure names. Close behind sits the assumption that every subfield stands up to hypoxia equally well, when CA1's fragility is itself an examined fact. Mixing this syndrome up with Korsakoff's is the third, because both present with severe memory loss and only the anatomical target and the presence of confabulationFilling a gap in memory with a made-up answer the person genuinely believes is true, not a deliberate lie.Read more → separate them. Calling H.M. a case series is the fourth, and it was one patient. A fifth is subtler: hunting for the guideline that governs 'managing the hippocampusA brain structure central to forming new memories.Read more →'. What actually gets managed is whatever damaged it, and each of those causes carries its own guidance — parenteral thiamine for suspected Wernicke's encephalopathyAn acute brain condition from thiamine deficiency causing confusion, ataxia, and abnormal eye movements.Read more →, set out in NICE CG100 on the physical complications of alcohol-use disorders, and intravenous aciclovir for suspected herpes simplex encephalitis.
Where a sequence of structures is offered, check it against the fixed order entorhinal cortex, dentate gyrus, CA3, CA1, subiculum before choosing. Where a vulnerability or an amnesia pattern is described, ask two things: which subfield the history implicates, with hypoxia pointing at CA1; and whether the memory loss is purely anterograde with a graded retrograde tail, which points at the hippocampusA brain structure central to forming new memories.Read more →, or comes with confabulationFilling a gap in memory with a made-up answer the person genuinely believes is true, not a deliberate lie.Read more → and a thalamic or mammillary lesion, which points at Korsakoff's.
Run a finger along the floor of the temporal horn of the lateral ventricle and you are following the , a curled strip of a few centimetres long. Everything on this page is a journey: where the structure sits, the road that runs through it, and the bill that arrives when the road is cut. A man known in the literature as H.M. paid that bill in 1953, and what he could and could not do afterwards taught neuroscience most of what it knows about how memory is built.
The hippocampal formation is the proper (the Cornu Ammonis fields, CA1 to CA4) together with the dentate gyrus and the subiculum, all packed into the medial . Its job is to take heavily processed information arriving from your cortex, bind it into a single episode, and hand a stabilised version back out for storage somewhere else.
Your is needed to make new declarative memories, not to keep them. As a memory consolidates it becomes less dependent on this structure and more widely represented across the neocortex, which is why damage here takes last week far more thoroughly than it takes a childhood holiday.
Information from every sensory modality converges on the entorhinal cortex, the gateway, and then travels through the formation along a route with a strong preferred direction. Dominant direction is the honest description rather than a one-way street: CA3 sends back-projections, and a direct route from entorhinal cortex to CA1 skips the main relay altogether. Traffic still runs overwhelmingly one way.
A road with few alternatives is efficient and fragile in the same breath. Cut it anywhere and your material stops arriving at the far end, however intact the cortex feeding it. That design fact sits behind every clinical consequence below.
Your signal leaves the entorhinal cortex along the perforant path (the bundle that perforates the hippocampal boundary to make the first ) and arrive at the dentate gyrus. From there the mossy fibres (short, thick axons with a beaded, moss-like look down the microscope) carry the signal to CA3, the second synapse. CA3's Schaffer collaterals (side branches of its main axons) hand on to CA1, the third — three synapses, hence the trisynaptic circuit. CA1 passes the signal to the subiculum, the formation's main output stage, which returns it to the entorhinal cortex and out to wider cortex.
The road continues into a longer loop, the Papez circuit: fornix, mammillary bodies, mammillothalamic tract, anterior thalamic nucleus, cingulate cortex, and back by way of the parahippocampal and entorhinal cortex. Proposed originally as the anatomy of emotion, it matters at least as much for memory, and it is where the second amnesic syndrome in this topic does its damage.
One stop is weaker than the rest. CA1 pyramidal cells are the most vulnerable in the to hypoxic-ischaemic injury, for example after a cardiac arrest or a long spell of severe hypotension, while the dentate gyrus tolerates the same insult comparatively well. The leading explanation points to NMDA-receptor properties and calcium handling in those cells, which leave them open to excitotoxic injury when oxygen delivery fails — well supported, not fully settled. Timing matters as well: a memory depends on this route for a limited consolidation window before it becomes stable enough to survive damage here, which is why a fresh injury wipes out recent memories far more completely than memories from a decade ago.
| Hippocampal amnesia | Wernicke- | |
|---|---|---|
| Damage sits at | Medial , especially CA1 | Mammillary bodies and dorsomedial |
| Usual cause | Hypoxia, surgery, herpes encephalitis | Thiamine (vitamin B1) deficiency |
| Not a defining feature | Classical | |
| Position on the Papez loop | At the start of the loop | Further round the same loop |
A second contrast is between kinds of memory rather than kinds of lesion. Declarative memory, meaning the facts and events you can consciously report, collapses after bilateral damage here, while procedural memory for skills and habits survives, because it runs on the and . A patient who cannot tell you they have met you before can still get better at a mirror-drawing task day after day.
No blood test names this for you. Neuropsychological testing carries the diagnosis: a striking, disproportionate failure to acquire new declarative material, set against preserved attention, general intelligence and procedural learning. Structural imaging, usually MRI, shows medial temporal pathology where there is any to see. The shape of the loss is characteristic in itself — anterograde amnesia dominates, and the retrograde component is temporally graded, sparing older memories in the way Ribot's law describes.
Anatomy is not treatable; the insult behind it often is, and urgently. Suspected in someone who misuses alcohol gets parenteral thiamine, which NICE CG100 on the physical complications of alcohol-use disorders sets out. Suspected herpes simplex encephalitis gets intravenous aciclovir started before confirmation, because waiting costs tissue. After a cardiac arrest the work is neuroprotective and supportive. For your patient left with an established amnestic syndrome, what follows is rehabilitation: external memory aids, errorless learning, and structured routines that lean on the procedural memory the damage spared.
Two things recur. The first is H.M.: one patient, not a case series, whose bilateral medial temporal resection for intractable epilepsy produced severe anterograde amnesia and a temporally graded retrograde tail while working memory, procedural learning and intelligence stayed intact. The second is CA1, which stems use to check whether you can localise an amnestic presentation to a subfield rather than to the whole structure as an undivided lump. Test yourself now: which subfield, which pattern, and what is the name for the sparing of old memories?
New information runs entorhinal cortex, dentate gyrus, CA3, CA1, subiculum through the hippocampal formation in one dominant direction, and because CA1 is the link most easily lost to hypoxia, damage there costs new declarative memory while old memories and practised skills remain.