Mednotes

L6 Nervous System Neuroanatomy 5

Properties2
tagsnervoussystem, lecture, anatomy, memory, alzheimersdisease
aliasesNervous System Neuroanatomy 5

16 min read

Overview

This lecture covers the neuroanatomy of learning and memory, centred on the hippocampus and the medial temporal lobe (MTL) memory system, and then applies this anatomy to Alzheimer’s disease (AD). It opens with a classification of memory systems, then works through hippocampal and MTL anatomy, the classic lesion/clinical evidence (Penfield, patient HM, patient RB) and cellular physiology (place, grid, head-direction cells) that established the hippocampus’s role in memory and spatial navigation. It then turns to AD: its clinical presentation and staging, its two hallmark pathologies (senile plaques from amyloid precursor protein processing, and neurofibrillary tangles from tau), the regional spread of pathology (Braak stages), genetic risk factors, the amyloid cascade hypothesis and the substantial evidence against it, alternative (vascular) hypotheses, current drug treatments, and the case for early biomarkers and lifestyle-based prevention.

Classification of Memory Systems

(Bartsch & Butler, Nature Reviews Neurology, 2013, 9:86-97)

Memory divides first into:

  • Short-term memory (working memory)
  • Long-term memory, which divides into:
    • Declarative (explicit) memory — the focus of this lecture — divided into:
      • Episodic: personal episodes in time and space -> relies on hippocampus, medial temporal lobe, neocortex
      • Semantic: facts, meanings, concepts, knowledge about the external world -> relies on lateral and anterior temporal cortex, prefrontal cortex
    • Non-declarative (implicit) memory, divided into:
      • Procedural (skills and habits) -> striatum, cerebellum, motor cortex
      • Priming and perceptual learning -> neocortex
      • Simple classical conditioning -> amygdala and cerebellum
      • Non-associative learning -> reflex pathways

Hippocampal Anatomy

The hippocampus is named for its seahorse-like shape when dissected free. It lies within the temporal lobe, medially tracing the inferior horn of the lateral ventricle, with the fornix running above/anterior to it.

The fornix is the hippocampus’s main fibre output, with parts (in order from the hippocampus forward):

  • Fimbria — fibres on the surface of the hippocampus (the origin point)
  • Crus — posterior limbs, also linked to the hippocampal commissural fibres
  • Body — the arching midsection
  • Columns — anterior limbs descending to the mammillary bodies (of the hypothalamus)

So the connection sequence is: fimbria -> crus -> body -> columns -> mammillary bodies.

Medial Temporal Lobe (MTL) Memory System

The medial temporal lobe houses the parahippocampal gyrus alongside the hippocampus. The MTL memory system has two major, reciprocally interconnected components:

  • Hippocampal formation (“where” — spatial memory role):
    • Hippocampus, itself composed of the CA fields (Cornu Ammonis) and dentate gyrus
    • Subiculum
  • Parahippocampal region (“what” — object/content memory role):
    • Pre-subiculum and para-subiculum
    • Entorhinal cortex
    • Perirhinal cortex (bordered by the rhinal sulcus)
    • Parahippocampal cortex

The subiculum links the hippocampus proper to the pre-/para-subiculum, which in turn connect to the entorhinal, perirhinal, and parahippocampal cortices.

Clinical Evidence for the Hippocampus’s Role in Memory

These cases are the classical evidence base for localising declarative memory to the hippocampus/MTL.

  • Wilder Penfield performed unanaesthetised neurosurgeries, identifying and removing epileptogenic scar tissue; stimulating the temporal lobes during surgery caused patients to recall memories of places and events.
  • Patient HM: knocked down by a bicycle at age 9, a minor seizure at age 10, a major seizure at age 16, and an experimental bilateral MTL resection at age 27 (1953) for focal epilepsy (Scoville & Milner, 1957). The anterior two-thirds of the hippocampus and adjacent cortex were removed. HM became profoundly amnesic:
    • Retrograde amnesia (recall of the past) — but remote memory stayed intact
    • Anterograde amnesia (forming new knowledge) — blocked
    • IQ, recognition, and motor abilities remained normal; spatial memory deficits were present
  • Patient RB (Zola-Morgan et al., 1986): a 52-year-old man who sustained a cardiac arrest (anoxia), producing permanent anterograde amnesia. Histology showed selective CA1 cell loss/gliosis — demonstrating that damage confined to CA1 alone is sufficient to cause permanent anterograde amnesia.

Hippocampus as a Cognitive Map

From “The Hippocampus As a Cognitive Map” (O’Keefe & Nadel, 1978), three cell types contribute to spatial navigation and orientation:

  • Place cells (hippocampus): each fires when the animal is in a specific location, producing a location-specific firing field.
  • Grid cells (entorhinal cortex): provide a complementary spatial coordinate system [slide does not elaborate on firing pattern beyond location in entorhinal cortex].
  • Head direction cells (post-subiculum): fire when the animal orients its head in a particular direction.

Together, place, grid, and head-direction cells determine location and orientation in space and support navigation.

Alzheimer’s Disease: Overview and Clinical Features

Dementia is an umbrella term for a range of symptoms associated with cognitive impairment. Approximate proportional causes of dementia: Alzheimer’s 50-75%, Vascular 20-30%, Lewy body 10-25%, Frontotemporal 10-15%, plus Parkinson’s disease dementia (no percentage given).

Alzheimer’s disease (first described by Dr. Alois Alzheimer in patient Auguste Deter) is a progressive, degenerative brain disease characterised by:

  • Increasing memory loss
  • Other cognitive decline
  • Changes in behaviour, personality, judgment, and activities of daily living

Early warning symptoms:

  • Memory loss affecting job skills
  • Difficulty performing familiar tasks
  • Problems with language
  • Disorientation of time and place
  • Poor or impaired judgment
  • Problems with abstract thinking
  • Misplacing things
  • Changes in mood or behaviour
  • Changes in personality
  • Loss of initiative

In severe AD, the person cannot: communicate verbally; understand words or instructions; recognise themselves in a mirror or pictures; recognise family members; or care for themselves. Patients usually die within 15 years.

Gross anatomical changes: atrophy (neuronal loss), most severe in the frontal and temporal lobes, and enlarged ventricles.

Neuropathology of AD

The two hallmark lesions are senile plaques (SPs) and neurofibrillary tangles (NFTs), alongside loss of acetylcholine (ACh)-producing cells in the basal forebrain.

Amyloid precursor protein (APP) processing — APP, a membrane protein, is cleaved via two competing pathways:

  1. Non-amyloidogenic (non-toxic) pathway: APP is cleaved by α-secretase into α-APPs and a membrane-bound stub (C83); C83 is then cleaved by γ-secretase to generate the short p3 peptide.
  2. Amyloidogenic pathway: APP is cleaved by β-secretase into β-APPs and a fragment (C99); C99 is then cleaved by γ-secretase, producing Aβ40 and Aβ42. These toxic Aβ peptides aggregate into insoluble neuritic (senile) plaques and are associated with reduced synapses around deposits.

Tau protein normally promotes microtubule assembly and stabilisation. In AD, tau becomes problematic: it detaches, microtubules disintegrate and their subunits fall apart, and tau forms tangled clumps (neurofibrillary tangles), leading to cell death.

Progression of Pathology

Tangles spread in a defined regional sequence:

  1. Entorhinal and perirhinal cortices (earliest)
  2. Hippocampus
  3. Frontal and parietal lobes

The amount of tangles parallels the duration and severity of AD. Tangles never occur in the cerebellum [slide does not elaborate on why].

Braak stages (Braak & Braak, 1991) formalise this progression across three stage groups, mapped to clinical status:

  • Transentorhinal stages I-II -> Asymptomatic
  • Limbic stages III-IV -> Mild symptoms
  • Neocortical stages V-VI -> Dementia

A separate four-stage schema (Braak et al., 2000) traces spread from the trans-entorhinal region (also involving uncus, para-subiculum, pre-subiculum, subiculum, CA1) outward through perirhinal cortex to temporal isocortex by Stage IV.

Entorhinal cortex layer II neuron loss is already significant at the mild cognitive impairment (MCI) stage: mean neuron counts were roughly 650,000 in controls (NCI) versus roughly 250,000 in MCI and roughly 280,000 in AD (both significantly reduced versus control) (Kordower et al., 2001) — i.e., most of the layer II loss has already occurred by MCI, with little further loss between MCI and AD.

Genetics and Risk Factors

Aging is described as a major risk factor for AD. AD is split into:

  • Early-onset familial AD (FAD): onset < 65 years, caused by genetic mutations
  • Late-onset sporadic AD: onset > 65 years, accounts for 95% of AD cases

Genetic loci implicated:

  • Chromosome 21 (APP): early-onset FAD; APP overexpression drives increased amyloid-beta
  • Chromosome 14 (PS1, presenilin 1): early-onset FAD, enhances Aβ production
  • Chromosome 1 (PS2, presenilin 2): [role stated only as the gene location; mechanism not elaborated on this slide]
  • Chromosome 19 (ApoE4): risk factor for late-onset AD (accounts for the 95% sporadic group)

AD prevalence rises steeply with age: 9% at age 80, 19% at 85, 32% at 90, 53% at age 95+.

Amyloid Cascade Hypothesis and Its Challenges

The amyloid cascade hypothesis (Hardy & Higgins, 1992) is the dominant but contested model of AD pathophysiology.

Proposed cascade (sequential steps):

  1. Overproduction, decreased clearance, or enhanced aggregation of Aβ42
  2. Aβ42 oligomerisation and deposition as diffuse plaques
  3. Subtle effects of Aβ42 oligomers on synapses
  4. Microglial and astrocytic activation (complement, cytokines)
  5. Progressive synaptic and neuritic injury
  6. Altered neuronal ionic homeostasis, oxidative injury
  7. Altered kinase/phosphatase activities
  8. Tangles form
  9. Widespread neuronal/neuritic dysfunction and cell death with transmitter deficits
  10. Dementia

Animal models used to test the hypothesis: transgenic models (e.g. APP, PS1, tau mutants) and non-transgenic models (e.g. brain injection of synthetic Aβ peptides).

Six challenges to the hypothesis:

  1. No direct evidence links APP, PSEN1, and PSEN2 mutations mechanistically.
  2. No β- or γ-secretase mutations are associated with FAD or protect against AD.
  3. Sporadic AD is linked to the APOE ε4 mutation without any mutations in APP or the secretases.
  4. Massive Aβ deposits are sometimes associated with few AD clinical symptoms.
  5. Transgenic mice (APP/PS) with defective APP genes and PSEN1/PSEN2 mutations develop amyloid plaques but do not develop overt neurofibrillary pathology.
  6. Passive immunisation against AD removes Aβ without cognitive improvement.

A 2014 commentary (Drachman) argues amyloid is “the downstream result, not cause” of AD, and that alternative hypotheses must explain (i) why amyloid toxicity is not the cause of AD, (ii) what alternative mechanisms cause the degeneration/dementia, and (iii) why amyloid accumulates in AD brains at all. A 2018 Nature feature likewise argues that, given repeated drug failures targeting amyloid-β, research should widen beyond the amyloid hypothesis. The lecture concludes more research into other mechanisms is needed.

Alternative Hypothesis: The Two-Hit Vascular Model

The two-hit vascular model of AD dementia (Nelson, Sweeney, Sagare & Zlokovic): cerebrovascular damage (“hit 1”) is an initial insult that is self-sufficient to initiate neuronal injury and neurodegeneration, but it can also promote accumulation of Aβ toxin in the brain (“hit 2”).

Mechanism as diagrammed: vascular risk factors (e.g. hypertension, diabetes) and genetic factors (e.g. APOE4), combined with aging, drive hit 1 (vascular damage) -> blood-brain barrier dysfunction and oligemia -> neurotoxins, reduced Aβ clearance, altered APP expression/processing, and microvascular hypoperfusion. Reduced Aβ clearance/altered APP processing feed into hit 2 (Aβ accumulation). Both hits converge on inflammation, Aβ accumulation, and phospho-tau, driving synaptic dysfunction, neuronal injury, and neurodegeneration, then disrupted structural and functional connectivity, culminating in dementia. The neurovascular unit involved comprises the neuron, oligodendrocyte, pericyte, endothelium, basement membrane, astrocyte, and microglia.

Treatments

There is no cure for AD. Approved drug classes:

  • Acetylcholinesterase (AChE) inhibitors — increase ACh levels by blocking its breakdown, to increase brain “tone”:
    • Donepezil (Aricept), approved 1996
    • Rivastigmine (Exelon), approved 2000
    • Galantamine (Razadyne), approved 2001
  • Drug attenuating nerve cell over-excitation:
    • Memantine (Ebixa/Namenda), approved 2003, for moderate-to-severe AD
  • Immunotherapies (target Aβ, for mild AD):
    • Lecanemab (Leqembi), approved 2023
    • Donanemab (Kisunla), approved 2024

Biomarkers, Prevention, and Remaining Challenges

Field challenges: biomarkers, early detection, early diagnosis, prognosis. Progression is described as “a slow march.”

Biomarkers become abnormal in a defined temporal sequence as disease stage advances from Normal -> Preclinical -> MCI -> Dementia:

  1. Amyloid-β accumulation (CSF/PET) — earliest to change
  2. Synaptic dysfunction (FDG-PET/fMRI)
  3. Tau-mediated neuronal injury (CSF)
  4. Brain structure (volumetric MRI)
  5. Cognition
  6. Clinical function — latest to change

Earlier biomarkers (amyloid, synaptic dysfunction) fall in the window for preventive and diagnostic intervention; later measures (cognition, clinical function) fall in the window addressed by new therapies.

Prevention emphasis: promote successful aging via lifestyle interventions — the lecture frames this as the best current preventive strategy, citing a 2017 Nature commentary (McDade & Bateman) arguing for stopping AD before onset rather than relying on drugs after the disease starts, and noting that roughly one in ten people aged 65+ has the disease.

Slide 24 (genetics of AD) was recovered from text extraction only — its visual layout/figures were not rendered in the transcript, so only the bulleted gene list above is confirmed; no diagram content is available for that slide.

Self-test

  1. List the branches of the long-term memory classification from long-term memory down to non-declarative subtypes, and name the primary brain region(s) associated with each.
  2. Describe the fibre pathway connecting the hippocampus to the mammillary bodies, in anatomical order.
  3. Distinguish the hippocampal formation from the parahippocampal region in the medial temporal lobe memory system, including their proposed functional roles (“where” vs “what”).
  4. Describe what happened to patient HM’s memory following his bilateral MTL resection, distinguishing which abilities were preserved and which were impaired.
  5. Explain why patient RB’s case is significant for localising anterograde amnesia, given the extent of his brain damage.
  6. Distinguish place cells, grid cells, and head direction cells by the brain region each is found in and what triggers their firing.
  7. Define dementia and list the approximate proportion of dementia cases attributable to each major subtype.
  8. Describe the two APP processing pathways and explain which one produces the peptides associated with plaque formation.
  9. Explain how tau contributes to neurofibrillary tangle formation and neuronal death.
  10. Describe the order in which neurofibrillary tangles spread through the brain in AD, and state which brain region is spared.
  11. Describe the relationship between Braak stages and clinical AD status.
  12. Explain what the entorhinal cortex layer II neuron count data show about when neuron loss occurs relative to the transition from MCI to AD.
  13. Distinguish early-onset familial AD from late-onset sporadic AD in terms of age of onset, cause, and proportion of cases.
  14. List the amyloid cascade hypothesis’s proposed sequence of events from Aβ42 overproduction to dementia.
  15. List three of the six challenges raised against the amyloid cascade hypothesis.
  16. Describe the two-hit vascular model of AD dementia, including what constitutes “hit 1” and “hit 2” and how they interact.
  17. A patient is started on donepezil for early AD. Explain the pharmacological rationale (what the drug does and why that is expected to help).
  18. Explain why, according to the biomarker-staging curve described in the lecture, a treatment aimed at “new therapies” targeting cognition or clinical function is being applied later in the disease course than a “preventive & diagnostic” strategy aimed at amyloid or synaptic biomarkers.
  19. A patient’s family reports progressive memory loss, difficulty with familiar tasks, and disorientation, but the patient still has normal motor abilities and recognises their family. Using the concepts from this lecture, explain roughly what disease stage this most likely reflects and why.
  20. Integrative: explain how evidence from patients HM and RB, together with the anatomy of the medial temporal lobe memory system, supports the conclusion that the hippocampus (and specifically CA1) is necessary for forming new declarative memories.

Answers

Graph View

Backlinks