11. The Nervous System

Neurons, Synapses, and the Electrical Language of the Body

Learning Objectives

11.1 The Organization of the Nervous System

The nervous system has one job that no other system can do: it moves information fast enough to matter. The endocrine system also carries messages, and Chapter 16 will show it doing so with great precision, but its fastest messages take seconds and its slowest take days. The nervous system operates in milliseconds, and everything strange about its anatomy follows from that requirement.

You have roughly 86 billion neurons and a comparable number of glial cells. A single cortical neuron may receive input at 7,000 synapses; a cerebellar Purkinje cell receives up to 200,000. The system's power lies not in the number of cells but in the number of connections, and this is worth saying at the outset because it reframes what learning, memory, development, recovery from stroke, and aging actually are. They are almost never changes in the number of neurons. They are changes in the wiring.

The divisions, and why each division exists

Anatomists divide the nervous system by where the tissue is. Physiologists divide it by what direction information is travelling and what it is going to control. Both divisions are used constantly and they cut across each other, which is the single most common source of early confusion. Hold them apart.

                          ═══ NERVOUS SYSTEM ═══
                                    │
        ┌───────────────────────────┴────────────────────────────┐
        │  ANATOMICAL DIVISION — where is the tissue?            │
        └───────────────────────────┬────────────────────────────┘
              ┌────────────────────┴─────────────────────┐
   ╔══════════╧═══════════╗                   ╔══════════╧════════════╗
   ║ CENTRAL NS  (CNS)    ║                   ║ PERIPHERAL NS  (PNS)  ║
   ║ brain + spinal cord  ║ ◄───────────────► ║ 12 pairs cranial n.   ║
   ║ INTEGRATION,         ║                   ║ 31 pairs spinal n.    ║
   ║ DECISION, MEMORY     ║                   ║ ganglia · receptors   ║
   ║ (Chapter 12)         ║                   ║ (Chapter 13)          ║
   ╚══════════════════════╝                   ╚═══════════╤═══════════╝
                                                          │
        ┌─────────────────────────────────────────────────┘
        │  FUNCTIONAL DIVISION — which way is traffic going?
        │
        ├────────────────────────────┬──────────────────────────────┐
   ╔════╧═══════════════╗       ╔════╧═════════════════════════════╗│
   ║ SENSORY (AFFERENT) ║       ║ MOTOR (EFFERENT)                 ║│
   ║ receptor ──► CNS   ║       ║ CNS ──► effector                 ║│
   ║ "ad-fero, I carry  ║       ║ "ex-fero, I carry away"          ║│
   ║  toward"           ║       ╚════╤═════════════════════════════╝│
   ║                    ║            │                              │
   ║ somatic sensory:   ║   ┌────────┴─────────┐                    │
   ║  skin, muscle,     ║   │                  │                    │
   ║  joints, special   ║ ╔═╧══════════╗  ╔════╧═══════════════════╗│
   ║  senses            ║ ║ SOMATIC    ║  ║ AUTONOMIC (ANS)        ║│
   ║                    ║ ║ MOTOR      ║  ║ "visceral motor"       ║│
   ║ visceral sensory:  ║ ║            ║  ║                        ║│
   ║  organs, vessels   ║ ║ VOLUNTARY  ║  ║ INVOLUNTARY            ║│
   ║  ── the pathway    ║ ║ skeletal   ║  ║ cardiac muscle ·       ║│
   ║     Amara's pain   ║ ║ muscle     ║  ║ smooth muscle · glands ║│
   ║     travels        ║ ║ ONE neuron ║  ║ TWO neurons in series, ║│
   ╚════════════════════╝ ║ CNS→muscle ║  ║ with a ganglion between║│
                          ║ always     ║  ╚════╤═══════════════════╝│
                          ║ EXCITATORY ║       │                    │
                          ╚════════════╝       │                    │
                    ┌──────────────────────────┴────────┐           │
              ╔═════╧═════════════════╗   ╔═════════════╧════════╗  │
              ║ SYMPATHETIC           ║   ║ PARASYMPATHETIC      ║  │
              ║ thoracolumbar T1–L2   ║   ║ craniosacral         ║  │
              ║ "fight or flight"     ║   ║ III,VII,IX,X + S2–S4 ║  │
              ║ SHORT pre-, LONG      ║   ║ "rest and digest"    ║  │
              ║ postganglionic fibre  ║   ║ LONG pre-, SHORT     ║  │
              ║ DIVERGES widely       ║   ║ postganglionic fibre ║  │
              ║ ↑HR ↑BP ↑glucose      ║   ║ DISCRETE targeting   ║  │
              ║ pupils dilate         ║   ║ ↓HR ↑digestion       ║  │
              ║ Amara's triage vitals ║   ║ pupils constrict     ║  │
              ╚═══════════════════════╝   ╚══════════════════════╝  │
                                                                    │
   NOTE: the ENTERIC nervous system — ~500 million neurons in the ───┘
   gut wall — is a semi-autonomous third division of the ANS (Ch. 23).

Three observations about that tree are worth more than the tree itself.

Afferent and efferent are relative terms, not absolute ones. They describe direction with respect to a named reference point. A fiber running toward the spinal cord is afferent to the cord; a fiber running from the thalamus to the cortex is afferent to the cortex even though it lies entirely within the CNS. When you meet the words in a new context, always ask "afferent to what?"

The somatic and autonomic motor divisions are built differently, and the difference is mechanically consequential. A somatic motor neuron has its cell body in the CNS and its axon runs unbroken to skeletal muscle, where it is always excitatory — there is no such thing as an inhibitory motor neuron acting on skeletal muscle, which is why relaxing a muscle requires that you stop commanding it rather than command it to relax. The autonomic pathway is a two-neuron chain with a synapse in a peripheral ganglion, and its second neuron may excite or inhibit its target depending on the receptor present. That architecture, developed fully in Chapter 13, is why the same transmitter can speed one organ and slow another.

Sympathetic divergence is why Amara's whole body responded to a problem in one artery. A single sympathetic preganglionic neuron may synapse with 20 or more postganglionic neurons distributed across several ganglia. The sympathetic division is built to fire as a broad pattern, not as a precise command. When her myocardium became ischemic, the resulting sympathetic discharge simultaneously raised her heart rate to 104, constricted her cutaneous arterioles to make her pale, activated her sweat glands to make her diaphoretic, and raised her blood pressure to 168/98 (Chapter 1). Four findings, one anatomical fact about branching.

11.2 The Neuron

A neuron is a cell that has been rebuilt around a single problem: how do you get a signal from one end of a structure to the other, quickly, without it fading? Every feature of its anatomy is an answer to some part of that problem.

Neurons share three properties that define them.

• Extreme longevity. Most of the neurons you have now, you were born with, and if you live to 90 many of them will be 90 years old. There is limited neurogenesis in the adult hippocampus and olfactory system, and essentially none elsewhere.
• Amitotic. With those small exceptions, mature neurons have permanently exited the cell cycle (Chapter 3). They cannot divide. This is the reason that destroyed brain and spinal cord tissue is not replaced with brain and spinal cord tissue — it is replaced by a glial scar, exactly as destroyed cardiac muscle is replaced by fibrous scar (Chapter 4). Two different organs, the same principle: post-mitotic tissue heals by substitution, not by regeneration.
• Exceptionally high metabolic rate. Neurons have almost no capacity for anaerobic metabolism and store no meaningful fuel. The brain is 2% of body weight and consumes about 20% of resting oxygen. This is why four to six minutes without perfusion produces irreversible damage (Chapter 1) while a skeletal muscle can survive an hour.

        ═════════════ A MULTIPOLAR NEURON — THE STANDARD PLAN ═════════════

         DENDRITES                     ┌── NUCLEOLUS (large, prominent:
      receptive surface;               │   this cell makes enormous
      many, short, tapered,            │   quantities of protein)
      branched. INPUT.                 │
      Graded potentials only.     ┌────┴────┐
              ╲   │   ╱           │ NUCLEUS │
               ╲  │  ╱            └────┬────┘
          ╲     ╲ │ ╱     ╱          ╱
           ╲     ╲│╱     ╱      ╱───╯   NISSL BODIES
            ╲──┐  │  ┌──╱      ╱        = rough ER + free ribosomes.
               │╭─┴─╮│    ╱───╯         Stain deep violet. Protein
        ───────┤│ ● ││───╯              factory; extends into dendrites
               │╰───╯│                  but NOT into the axon hillock
            ╱──┘  │  └──╲               ── which is how you find the
           ╱      │      ╲              hillock down a microscope.
                  │
              SOMA (perikaryon, cell body)      NEUROFIBRILS = bundled
              biosynthetic centre. In the CNS   neurofilaments; tensile
              a cluster of somata = NUCLEUS.    scaffolding + the rails
              In the PNS = GANGLION.            for axonal transport.
                  │
                  ▼
          ╭───────────────╮
          │ AXON HILLOCK  │  ◄── THE DECISION POINT. Highest density of
          │ + INITIAL     │      voltage-gated Na+ channels in the cell
          │   SEGMENT     │      (~50× the soma). Every graded potential
          ╰───────┬───────╯      arriving anywhere on the dendrites or
                  │              soma is summed HERE against threshold.
                  │
     ═════════════╪══════════════════════════════════════════════════
        ┌───┐  ┌──┴┐  ┌───┐  ┌───┐   MYELIN SHEATH
        │███│  │███│  │███│  │███│   (Schwann cell in PNS / oligodendro-
     ───┤███├──┤███├──┤███├──┤███├── cyte in CNS). Up to 100 wraps of
        └───┘  └─┬─┘  └───┘  └───┘   plasma membrane; ~80% lipid.
                 │  ▲       ▲
                 │  │       └── NODE OF RANVIER (gap, 1–2 µm)
                 │  └────────── INTERNODE (0.2–2 mm)
                 │
              AXON — ONE per neuron. Uniform diameter. OUTPUT.
              Carries ACTION POTENTIALS, which do not decay.
              Contains NO rough ER: every protein it needs must be
              manufactured in the soma and shipped.
                 │
                 │   ◄── AXON COLLATERAL (branches at right angles)
                 ▼
        ╱────────┴────────╲
       ╱    TELODENDRIA    ╲   terminal branching — one axon may end on
      ╱   ╱    │    ╲   ╲   ╲  thousands of targets (DIVERGENCE)
     ○   ○     ○     ○   ○   ○
     └───┴─────┴─────┴───┴───┘
        AXON TERMINALS (synaptic knobs / boutons)
        Packed with mitochondria and synaptic vesicles.
        Convert the electrical signal back into a CHEMICAL one.

     ─────────────────────────────────────────────────────────────────
     SCALE: a motor neuron in the spinal cord commanding a muscle in
     the great toe has a soma ~100 µm across and an axon ~1 m long.
     If the soma were the size of a tennis ball, the axon would be
     about 1.2 km long and the width of a garden hose.

The soma and its contents

The soma (cell body, perikaryon) contains the nucleus and is the cell's factory. Two structures visible under a light microscope deserve their own names because they behave diagnostically.

Nissl bodies (chromatophilic substance) are aggregates of rough endoplasmic reticulum and free ribosomes, so densely packed that they stain deep violet with basic dyes. They are the physical evidence of the neuron's protein appetite: an axon a metre long, whose membrane must be continuously renewed and whose synaptic vesicles are constantly consumed, is supplied entirely from here. Nissl substance extends into the dendrites but stops abruptly at the axon hillock, which is exactly how a histologist tells the axon from a dendrite in a section.

Neurofibrils are bundled neurofilaments — the intermediate filaments of the neuronal cytoskeleton (Chapter 3). They give the soma and its processes tensile strength, and, together with microtubules, they form the track system along which cargo is moved.

Dendrites versus axon: the asymmetry that makes a signal directional

The asymmetry is the point. Information enters over a huge, high-resistance, branched receptive surface where signals can be weighed against each other; it is summed at a single narrow neck; and if it passes threshold there, it leaves down a single low-loss cable. Structure determines function almost embarrassingly directly here.

Classification by structure

Neurons are classified by the number of processes emerging from the soma.

Unipolar neurons carry Amara's pain, and their odd geometry is worth a paragraph. The soma sits in a dorsal root ganglion just outside the spinal cord, off to the side of the signal path rather than in it. A single process leaves the soma and immediately forks: a peripheral process runs out to the tissue and ends as a receptor, and a central process runs into the spinal cord. Functionally, the entire length from receptor to spinal cord behaves as one axon — the impulse never enters the soma at all, it simply shoots past the junction. That is why a signal from your toe reaches your spinal cord without being slowed by a detour through a cell body, and it is why the sensory neuron carrying pain from Amara's heart can be a single uninterrupted cell more than half a metre long.

Classification by function

That last row is the one to sit with. Approximately ninety-nine percent of your neurons talk only to other neurons. The nervous system is overwhelmingly occupied with talking to itself, and the input and output layers are a thin skin over an enormous internal computation. Every capacity you value — deciding, remembering, imagining, choosing not to say the thing you were about to say — lives in that 99%.

Axonal transport: the supply line

An axon has no ribosomes. Every enzyme, every channel protein, every vesicle membrane, every mitochondrion in an axon terminal a metre away was built in the soma and physically carried there. The cell solves this with motor proteins walking along microtubules, burning ATP.

Run the numbers on the slow component. Rebuilding the cytoskeleton of a one-metre axon at 1 mm/day takes about three years. This is not a trivia item. It is why peripheral nerve regeneration after an injury proceeds at roughly 1 mm per day — about an inch a month — so that a nerve severed at the elbow takes the better part of a year to reinnervate the hand, and why recovery from a severe nerve injury is measured in seasons.

11.3 Neuroglia: The Other Half of the Nervous System

For most of the twentieth century the glia were described as "supporting cells," a phrase that implied scaffolding and packing material. That description is now known to be wrong in an interesting way. Glial cells regulate the composition of the fluid around neurons, control which molecules reach the brain at all, form and maintain the insulation on which fast conduction depends, prune synapses, and — in the case of astrocytes — signal to each other and to neurons using calcium waves. Neurons are the wiring. Glia are the entire infrastructure that makes the wiring possible, and roughly half of all primary brain tumours arise from them, because unlike neurons, glia can still divide.

There are six types: four in the CNS, two in the PNS.

   ══════════════════ CENTRAL NERVOUS SYSTEM ══════════════════════════════

   1 · ASTROCYTE  ("star cell")            MOST ABUNDANT glial cell; ~20–40%
       ╲   │   ╱                            of CNS glia.
        ╲  │  ╱   ╭─── endfoot on capillary
     ────╳─●─╳────┤    JOBS
        ╱  │  ╲   ╰─── endfoot on synapse   • induces + maintains the
       ╱   │   ╲                              BLOOD–BRAIN BARRIER by
      ╱    │    ╲                             signalling to endothelium
   ══╡ CAPILLARY ╞══                        • K+ spatial buffering: soaks
      ╲    │    ╱                             up K+ released by firing
       ▪ ── ── ▪  ← synapse it wraps          neurons, disperses via gap
                                              junctions to distant vessels
   CLINICAL: astrocytes form the GLIAL      • recycles glutamate (EAAT
   SCAR after CNS injury — sealing the        transporters) — prevents
   damage but physically and chemically       EXCITOTOXICITY
   BLOCKING axon regrowth. This is a        • supplies neurons with lactate
   principal reason CNS axons do not          from stored glycogen
   regenerate and PNS axons do.             • glioma = commonest primary
                                              malignant brain tumour

   2 · MICROGLIA                            SMALLEST + FEWEST (~5–10%).
        ○──┬──○   spiky, mobile             NOT ectodermal — derived from
       ╱   │   ╲  processes constantly      yolk-sac macrophages; the
      ○    ●    ○ sampling their territory  brain's resident IMMUNE cell.
       ╲   │   ╱
        ○──┴──○                             JOBS
                                            • phagocytose debris, microbes,
   CLINICAL: chronic microglial activation    dying cells
   ("neuroinflammation") is implicated in   • antigen presentation
   Alzheimer, Parkinson, and the cognitive  • SYNAPTIC PRUNING in
   effects of chronic systemic illness.       development and learning
                                            • release cytokines → the
                                              behavioural "sickness
                                              response" during infection

   3 · EPENDYMAL CELL                       Cuboidal-to-columnar EPITHELIUM
     ╭─┬─┬─┬─┬─┬─┬─╮  ← cilia               lining the ventricles and the
     │▪│▪│▪│▪│▪│▪│▪│                        central canal of the cord.
     └─┴─┴─┴─┴─┴─┴─┘  ← microvilli
     ═══ ventricle ═══                      JOBS
                                            • cilia circulate CSF
   CLINICAL: ependymal cells at the         • specialized ependyma forms
   aqueduct can be obstructed (tumour,        the CHOROID PLEXUS, which
   haemorrhage, congenital stenosis) →        MAKES the CSF
   HYDROCEPHALUS (Chapter 12).              • a selectively permeable
                                              barrier CSF ↔ brain

   4 · OLIGODENDROCYTE                      MYELINATES the CNS.
        ╱▬▬▬▬  axon 1                       ONE oligodendrocyte extends
       ╱─▬▬▬▬  axon 2                       processes to as many as
      ●──▬▬▬▬  axon 3    ...up to 60        50–60 SEPARATE axons, wrapping
       ╲─▬▬▬▬  axon 4    axon segments      ONE internode of each.
        ╲▬▬▬▬  axon 5
                                            NO neurilemma → NO regeneration
   CLINICAL: the target of MULTIPLE           tube. CNS axons do not regrow.
   SCLEROSIS. Killing ONE oligodendrocyte
   demyelinates up to 60 axons at once.

   ══════════════════ PERIPHERAL NERVOUS SYSTEM ═══════════════════════════

   5 · SCHWANN CELL (neurolemmocyte)        MYELINATES the PNS.
                                            ONE Schwann cell = ONE internode
     ╭──────────────╮                       of ONE axon. Its nucleus and
     │ ●  nucleus   │ ← NEURILEMMA          cytoplasm are squeezed to the
     │ ╭──────────╮ │   (outer cytoplasm    OUTSIDE, forming the NEURILEMMA.
     │ │ ((( ● ))) │ │    + nucleus, kept
     │ │  axon     │ │    outside the wrap) After injury the neurilemma +
     │ ╰──────────╯ │                       basal lamina survive as a TUBE
     ╰──────────────╯                       that guides the regrowing axon
                                            at ~1 mm/day. THIS is why a cut
   Unmyelinated PNS axons are still         nerve in the arm can recover and
   enfolded — up to 15 axons sitting in     a cut spinal cord cannot.
   grooves of a single Schwann cell,
   with no spiral wrapping.

   6 · SATELLITE CELL                       Flattened cells surrounding
       ○ ○ ○ ○ ○                            neuronal SOMATA in PNS ganglia.
      ○ ╭─────╮ ○      The PNS equivalent   Regulate the chemical environment
      ○ │ ● ▪ │ ○      of the astrocyte.    of the cell body; exchange
      ○ ╰─────╯ ○                           signals with it; become
       ○ ○ ○ ○ ○                            activated in chronic pain states
                                            and amplify nociceptive firing.

Myelination: the same trick, built two different ways

Myelin is not a secretion and not a special substance. It is ordinary plasma membrane, wrapped many times around an axon, with the cytoplasm squeezed out from between the layers so that what remains is a compact spiral of lipid bilayers — up to 100 wraps, roughly 80% lipid by dry weight, compared with about 40% for a typical plasma membrane. Lipid is an electrical insulator. Wrapping an axon in a hundred layers of it does two things simultaneously: it multiplies the transverse resistance of the membrane, so current leaks out far more slowly, and it divides the membrane capacitance, so far less charge is needed to change the voltage. Both effects speed conduction, and §11.7 makes that quantitative.

The CNS and PNS build the same insulation with different cells, and the differences have direct clinical consequences.

The regeneration asymmetry is one of the most consequential facts in clinical neurology, and it has three causes, all visible in the table. Peripheral axons regrow because the Schwann cell survives the injury, because the neurilemma and basal lamina persist as a physical tube pointing the growth cone at its original target, and because Schwann cells actively secrete growth factors and clear myelin debris. Central axons do not regrow because oligodendrocytes leave no tube, because the astrocytic glial scar is a physical and chemical barrier, and because CNS myelin itself contains growth-inhibitory proteins such as Nogo. Note that the axon is not the problem. Transplant experiments have shown that a central axon placed in a peripheral nerve graft will grow enthusiastically. The neuron is willing; the environment forbids it.

11.4 The Resting Membrane Potential

Everything in the next three sections rests on one idea, so it is worth building slowly and completely. If you understand the resting membrane potential properly, action potentials, synapses, cardiac conduction (Chapter 18), the effect of potassium on the heart, the mechanism of local anesthetics, and the pancreatic beta cell's response to glucose (Chapter 16) all follow with very little extra work. If you do not, you will be memorizing all of them separately for the rest of the course.

Step 1 — There are unequal concentrations of ions across the membrane

Two features of that table do most of the work. Potassium is concentrated inside; sodium is concentrated outside. And the cell contains a large population of negatively charged proteins and organic phosphates that are physically too big to leave — a fixed, trapped negative charge.

Step 2 — The membrane is selectively permeable, and at rest it is a potassium membrane

A resting neuron's membrane is not equally leaky to everything. It contains leak channels (also called two-pore-domain or "background" channels) that are open all the time, and there are far more potassium leak channels than sodium leak channels. At rest the membrane is roughly 25 to 40 times more permeable to K⁺ than to Na⁺.

This asymmetry — not the concentration gradient by itself — is what sets the voltage. Hold both facts together: the gradients determine which way each ion wants to move; the permeabilities determine how much say each ion gets.

Step 3 — The equilibrium potential, in plain language

Imagine, for a moment, a membrane permeable to potassium and to nothing else.

Potassium is concentrated inside, so it diffuses out down its concentration gradient. But potassium carries a positive charge, and the large anions it leaves behind cannot follow. So every K⁺ that leaves makes the inside slightly more negative — and a negative interior attracts K⁺ back in.

Two forces now act on potassium in opposite directions: the chemical gradient pushing it out, and the growing electrical gradient pulling it back. As more K⁺ leaves, the electrical pull strengthens until it exactly balances the chemical push. At that voltage, K⁺ still moves in both directions, but at equal rates, so there is no net movement. That voltage is the equilibrium potential for potassium, written E_K.

Three things about equilibrium potentials that are worth stating explicitly, because they are where intuition usually fails:

• An equilibrium potential is a property of one ion, not of the cell. Each ion has its own, set entirely by its own concentration ratio. They exist whether or not the membrane is permeable to that ion.
• It is the voltage at which that ion would stop having any net effect. Equivalently: it is the voltage the membrane would go to if it became permeable only to that ion.
• Astonishingly few ions actually move. The membrane is a capacitor with a very high capacitance-to-volume ratio, so changing the voltage by 100 mV requires the net transfer of fewer than one in 100,000 of the cell's potassium ions. Measurable voltages are produced by chemically undetectable movements of charge. This is why a cell can fire a thousand action potentials without meaningfully changing its internal ion concentrations.

Step 4 — Putting it together: why −70 mV and not −90 mV

The resting membrane potential of a typical neuron is about −70 mV (the range across excitable cells runs from about −40 mV in some smooth muscle to −90 mV in skeletal muscle and cardiac ventricular myocytes). The convention is that the voltage is stated for the inside relative to the outside, so −70 mV means the inside is 70 millivolts negative with respect to the outside.

Why not −90 mV, the potassium equilibrium potential? Because the membrane is not purely permeable to K⁺. That small sodium leak — one-twenty-fifth to one-fortieth of the potassium permeability — is continuously letting Na⁺ trickle in, and Na⁺ pulls the membrane toward its own equilibrium potential of +60 mV. The resting potential settles at a weighted average of the equilibrium potentials, weighted by each ion's permeability. Potassium dominates because it is 25–40 times more permeable, so the answer lands near E_K — but the sodium leak drags it about 20 mV positive of E_K.

That gap is not an imperfection. It is the entire operating margin of the nervous system. A cell sitting exactly at E_K would have no room to hyperpolarize; a cell sitting at 0 mV would have no stored energy. Resting at −70 mV, a neuron sits about 15 mV below threshold and about 130 mV below the sodium equilibrium potential — poised, with an enormous inward driving force on Na⁺ held back by nothing but closed gates.

    ══════════ THE RESTING MEMBRANE POTENTIAL AS A BALANCE OF FORCES ═══════

  EXTRACELLULAR FLUID                                         + + + + + + +
  Na+ 145 mM ●●●●●●●●●●●●●●●   K+ 4 mM ○      Cl- 110 mM ▪▪▪▪▪▪▪▪
  ─────┬──────────┬─────────────┬──────────────┬──────────┬──────────────
       │ Na+ LEAK │  Na+/K+     │   K+ LEAK    │ K+ LEAK  │  VOLTAGE-GATED
       │ CHANNEL  │  ATPase     │   CHANNEL    │ CHANNEL  │  Na+ CHANNEL
       │  (few)   │  (PUMP)     │   (MANY)     │ (MANY)   │  ── CLOSED ──
   ════╪══════════╪═════════════╪══════════════╪══════════╪══════════════
    ●  │          │  ●●●        │              │          │
    ●  ▼ trickle  │   ▲ 3 Na+   │      ▲       │    ▲     │   ✗ no flow
    ●  │   IN     │   │ OUT     │      │ K+    │    │ K+  │     at rest
       │          │   │         │      │ OUT   │    │ OUT │
       │          │   ▼ 2 K+    │      │       │    │     │
       │          │  ○○  IN     │      ○       │    ○     │
   ════╪══════════╪══════ ATP ══╪══════════════╪══════════╪══════════════
       ▼          │      ↓ADP   │              │          │
  INTRACELLULAR FLUID                                     − − − − − − −
  Na+ 12 mM ●     K+ 140 mM ○○○○○○○○○○○○○○○   Cl- 5–10 mM ▪
  A− (proteins, phosphates) ▬▬▬▬▬▬▬▬▬▬▬  ◄── TRAPPED. Cannot leave.
                                              The fixed negative charge.

    ─────────────────────────────────────────────────────────────────────
    THE THREE CONTRIBUTIONS, IN ORDER OF SIZE

    1 · K+ DIFFUSES OUT through many open leak channels, leaving
        trapped anions behind.  Drives V toward  E_K = −90 mV   ●●●●●●●●●
    2 · Na+ TRICKLES IN through few open leak channels.
        Drives V toward         E_Na = +60 mV                   ●●
    3 · Na+/K+ ATPase pumps 3 Na+ OUT for every 2 K+ IN.
        Net loss of ONE positive charge per cycle = ELECTROGENIC.
        Direct contribution   ≈ −3 mV                           ●
        INDIRECT contribution = everything: it maintains the
        gradients that 1 and 2 depend on.

              WEIGHTED AVERAGE, P_K : P_Na ≈ 25–40 : 1
                              ▼
                     V_rest ≈ −70 mV
                              ▼
    ─────────────────────────────────────────────────────────────────────
    THE MEMBRANE AS A CHARGED CAPACITOR

      OUTSIDE  + + + + + + + + + + + + + + + + + + + + + + + + +
      ═══════════════════ lipid bilayer ═══════════════════════════
      INSIDE   − − − − − − − − − − − − − − − − − − − − − − − − −

      The charge separation exists ONLY in a nanometre-thin layer
      hugging each face of the membrane. The bulk cytoplasm is
      electrically NEUTRAL. Fewer than 1 in 100,000 K+ ions has to
      move to generate the entire −70 mV.

Step 5 — The Na⁺/K⁺ ATPase and what it actually contributes

The sodium-potassium pump is a transmembrane enzyme that hydrolyses one ATP to move three Na⁺ out and two K⁺ in against both ions' gradients. It runs continuously in every cell you have, and it consumes roughly 20–30% of a resting cell's ATP — considerably more in a neuron, and a substantial fraction of the entire brain's enormous energy budget.

Its contribution is of two very different kinds, and students routinely conflate them.

Directly, the pump is electrogenic: three positive charges leave for every two that enter, a net export of one positive charge per cycle, which makes the inside slightly more negative. This accounts for only about 3 mV of the −70 mV. If you poisoned the pump with ouabain and measured the voltage a second later, it would have changed by about that much.

Indirectly, the pump is responsible for essentially all of it, because it is what maintains the concentration gradients on which steps 1 through 4 depend. Poison it and wait: Na⁺ leaks inward and K⁺ leaks outward with nothing to restore them, the gradients run down over minutes to tens of minutes, E_K and E_Na collapse toward zero, and the membrane potential decays with them. The cell dies electrically not because it lost 3 mV of pump current but because it lost its battery charger.

This is exactly what happens in ischemia. When Amara's coronary artery narrows and a region of myocardium is deprived of oxygen, ATP production falls, the Na⁺/K⁺ ATPase slows, K⁺ leaks out of the affected cells into the extracellular space, and those cells depolarize. The electrical consequences of that local depolarization are what the ECG detects (Chapter 18) — the ST-segment changes on her tracing are, at bottom, a Figure 11.4 problem.

11.5 Graded Potentials: The Currency of Decision

A neuron does two electrically distinct things. On its dendrites and soma it listens, adding up thousands of small, variable inputs. At its axon hillock it decides, and on its axon it broadcasts an all-or-nothing verdict. The listening signal is the graded potential; the verdict is the action potential. Confusing the two is the single most common error in this material, so hold the differences in view from the start.

EPSPs and IPSPs

A postsynaptic potential is a graded potential produced in the receiving cell when a neurotransmitter binds its receptor.

An excitatory postsynaptic potential (EPSP) moves the membrane toward threshold. The classic mechanism is a ligand-gated cation channel — the glutamate-activated AMPA receptor is the standard example — that is permeable to both Na⁺ and K⁺. Because the driving force on Na⁺ at rest (−70 mV against E_Na of +60 mV, a gap of 130 mV) vastly exceeds the driving force on K⁺ (−70 mV against E_K of −90 mV, a gap of 20 mV), far more Na⁺ enters than K⁺ leaves, and the net effect is depolarization toward a reversal potential near 0 mV. A single EPSP at a central synapse is small — typically 0.5 to 1 mV.

An inhibitory postsynaptic potential (IPSP) moves the membrane away from threshold, or holds it where it is. Two mechanisms:

• Cl⁻ influx through GABA-A or glycine receptors. Since E_Cl (~−70 mV) sits at or just below rest, this either hyperpolarizes slightly or — more often — produces the shunting inhibition described in Check 10.4: it clamps the membrane at −70 mV so that excitatory currents are drained away before they can reach the hillock.
• K⁺ efflux through channels opened by metabotropic receptors (GABA-B, many others), which drives the membrane toward E_K at −90 mV — a true hyperpolarization.

Decremental conduction, and why it forces summation

A graded potential begins where the channels opened and then spreads passively through the cytoplasm, getting smaller as it goes. Two properties govern the decay:

• The length constant (λ) is the distance over which the signal falls to 37% of its original amplitude — typically 0.1 to 1 mm in a neuron. Fat processes with well-insulated membranes have long length constants; thin, leaky ones have short ones.
• The time constant (τ) is how long the membrane takes to charge or discharge, typically 1 to 20 ms. A long time constant means an EPSP lingers, and a lingering EPSP can meet the next one.

The consequence is decisive. A 1 mV EPSP generated on a distal dendrite arrives at the axon hillock as a fraction of a millivolt, against a threshold sitting 15 mV away. No single synapse in the central nervous system can fire a neuron. Reaching threshold requires the cooperation of somewhere between fifteen and fifty simultaneous or near-simultaneous EPSPs. This is not a design flaw. It is the mechanism by which a neuron is made to be a device that weighs evidence rather than a relay that repeats whatever it is told.

  ═══ SUMMATION: HOW A NEURON ADDS UP ITS INPUTS AT THE AXON HILLOCK ═══

  A · NO SUMMATION — single EPSPs, far apart in time
   mV
  -55┈┈┈┈┈┈┈┈┈┈┈┈┈ THRESHOLD ┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
  -70──╱╲──────────────╱╲──────────────╱╲───────────  no firing
       ▲               ▲               ▲
     stim A          stim A          stim A     each decays fully

  B · TEMPORAL SUMMATION — ONE synapse firing RAPIDLY
   mV                                        ┌─ AP! ─┐
  -55┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈╱╲┈┈┈┈┈┈┈┈┈┈
  -70──╱╲─╱─╲╱──╲╱────╱─────────────────────────╱  ╲
       ▲  ▲  ▲  ▲  ▲                        each new EPSP arrives
      A  A  A  A  A   ← <15 ms apart        before the last decays;
                                            they ADD.

  C · SPATIAL SUMMATION — MANY synapses firing AT ONCE
   mV                    ┌── AP! ──┐
  -55┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈╱╲┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
  -70─────────────────────╱──╱  ╲
                          ▲
                    A+B+C+D fire simultaneously on
                    DIFFERENT parts of the dendritic tree

  D · EPSP + IPSP — the actual situation, every millisecond
   mV
  -55┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
  -70───╱╲──╲╱──╱╲╱╲──╲──╱────────────  net = below threshold
        ▲   ▼   ▲ ▲   ▼  ▲             NO firing despite excitation
       EPSP IPSP        IPSP           INHIBITION IS AN ANSWER,
                                       NOT AN ABSENCE.

  ══════════════════════════════════════════════════════════════════════
                    THE NEURON AS A SPATIAL INTEGRATOR

        distal dendrite  ──► EPSP decays to ~0.1 mV by the hillock
         (weak vote)          λ = 0.1–1 mm; τ = 1–20 ms
                        ╲
        proximal dendrite ──► EPSP arrives ~0.6 mV
         (stronger vote)  ╲
                           ╲    ┌──────────────────────────┐
        soma synapse    ─────►  │      AXON HILLOCK        │
         (strong vote)   ╱      │  Na+ channel density     │
                        ╱       │  ~50× that of the soma   │
        AXO-AXONIC     ╱        │  ⇒ LOWEST THRESHOLD IN   │
        synapse on the           │    THE WHOLE CELL       │
        initial segment ──────► │  ⇒ THE DECISION IS MADE  │
         (VETO — it can          │    HERE, AND ONLY HERE  │
          block output           └───────────┬──────────────┘
          whatever the                       │
          dendrites say)                     ▼
                                    all-or-none ACTION POTENTIAL
                                    (or nothing at all)

Why the axon hillock, and only the axon hillock

Every synapse on the cell votes, but the votes are not equal, and the count happens in one place. The axon hillock and initial segment carry a density of voltage-gated Na⁺ channels roughly fifty times higher than the soma membrane. A higher channel density means a smaller depolarization suffices to trigger regenerative opening, which means the hillock has the lowest threshold in the entire cell. Whatever voltage arrives there, having been summed across the whole dendritic tree and soma, is what determines whether the neuron speaks.

Three consequences follow immediately, and all three are exploited by real circuits:

1. Position is weight. A synapse on the soma has far more influence than an equally strong synapse on a distal dendritic tip, because less of its signal is lost on the way. Inhibitory synapses cluster on the soma and proximal dendrites — precisely where they can cancel the most.
2. Timing is weight. Inputs arriving within one time constant of each other sum; inputs arriving 50 ms apart do not. A neuron is therefore a coincidence detector, and the width of the coincidence window is a membrane property that can itself be modulated.
3. A single well-placed synapse can veto everything. An axo-axonic synapse onto the initial segment — the arrangement used by cortical chandelier cells — sits downstream of all summation and can prevent output no matter what the dendrites concluded.

11.6 The Action Potential

An action potential is a brief, stereotyped, self-propagating reversal of membrane polarity — about 100 mV in amplitude, about 1 ms in duration, identical every time in a given axon. It is the only signal that can travel a metre without weakening, and it is therefore the only signal the nervous system can use to communicate over distance.

The molecular machine: one channel, two gates

Everything about the action potential's shape follows from an unusual property of the voltage-gated Na⁺ channel: it has two independent gates that respond to the same voltage change with different speeds and in opposite directions.

The channel therefore has three states, not two: closed but available (activation gate shut, inactivation gate open — the resting state), open (both gates open — the brief conducting window), and inactivated (activation gate open, inactivation gate shut — cannot conduct and cannot be made to conduct by any stimulus, no matter how strong).

That third state is the whole secret. A closed channel can be opened by depolarizing it. An inactivated channel cannot. The only thing that restores it is repolarization, which takes time. Two of the most important properties in this chapter — the refractory period and the one-way propagation of the impulse — are consequences of this one asymmetry.

The voltage-gated K⁺ channel, by contrast, is simple: one gate, no inactivation, and it opens slowly, taking about a millisecond to respond to the same depolarization that opens the Na⁺ activation gate in a tenth of that. That delay is why depolarization gets to happen at all.

   ══════════════ THE ACTION POTENTIAL, PHASE BY PHASE ══════════════════

  mV
 +40┤                    ╭─╮  ③ ← PEAK. P_Na has risen ~600-fold.
    │                   ╱   ╲       V_m heads for E_Na (+60) but never
 +20┤                  ╱     ╲      arrives: Na+ channels are already
    │                 ╱       ╲     inactivating and K+ channels opening
   0┤                ╱         ╲
    │               ╱           ╲   ④ REPOLARIZATION
 -20┤              ╱             ╲     Na+ inactivation gates SHUT
    │   ②         ╱               ╲    Voltage-gated K+ channels OPEN
 -40┤  DEPOL-    ╱                 ╲   K+ rushes OUT
    │  ARIZATION╱                   ╲
 -55┼┈┈┈┈┈┈┈┈┈┈╱┈THRESHOLD┈┈┈┈┈┈┈┈┈┈┈╲┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
    │        ╱ ╱                       ╲
 -70┼───────╱─╯                         ╲          ╭──────────  ⑥ REST
    │  ①  ╱  graded potentials           ╲        ╱   restored
 -80┤ STIMULUS  summating                 ╲______╱
    │                                      ⑤ HYPERPOLARIZATION
 -90┤                                        (undershoot). Slow K+
    └──┬─────┬─────┬─────┬─────┬─────┬─────┬─┬───┬─────┬───► ms
       0    0.5   1.0   1.5   2.0   2.5   3.0   3.5   4.0

  ═══ CHANNEL STATES, ALIGNED TO THE TRACE ABOVE ═══════════════════════
  Na+ ACTIVATION gate  │ shut │OPEN OPEN│ open  open  open │ shut  shut │
  Na+ INACTIVATION gate│ OPEN │OPEN shut│ shut  shut  shut │ OPEN  OPEN │
  Na+ CHANNEL CONDUCTS?│  no  │YES  no  │  no    no    no  │  no    no  │
  K+ voltage-gated     │ shut │shut open│ OPEN  OPEN  OPEN │ open  shut │
                       └──────┴─────────┴──────────────────┴────────────┘

  ═══ REFRACTORY PERIODS ═══════════════════════════════════════════════
       ├──── ABSOLUTE (0.4–1.0 ms) ────┤├──── RELATIVE (2–5 ms) ───┤
       │ NO stimulus of ANY strength   ││ A STRONGER-than-normal   │
       │ can fire a second AP.         ││ stimulus CAN fire an AP. │
       │ WHY: Na+ channels are         ││ WHY: some Na+ channels   │
       │ INACTIVATED — the gate is     ││ have reset, but K+       │
       │ shut and only repolarization  ││ channels are still open  │
       │ reopens it.                   ││ and the cell is hyper-   │
       │ ACCOMPLISHES:                 ││ polarized, so more       │
       │ • sets MAXIMUM FIRING RATE    ││ depolarization is needed.│
       │   (~1 ms floor ⇒ ≤1000 Hz)    ││ ACCOMPLISHES:            │
       │ • forces ONE-WAY propagation  ││ • graded frequency coding│
       │   (the membrane behind is     ││   — stimulus strength is │
       │   inexcitable, so the impulse ││   translated into RATE   │
       │   cannot turn round)          ││ • prevents runaway firing│
       │ • keeps each AP DISCRETE:     │└──────────────────────────┘
       │   no summation ⇒ ALL-OR-NONE  │
       └───────────────────────────────┘

  ═══ ALL-OR-NONE ══════════════════════════════════════════════════════
   Stimulus at 60% of threshold  ──► nothing
   Stimulus at 99% of threshold  ──► nothing
   Stimulus at 100% of threshold ──► FULL 100 mV spike
   Stimulus at 500% of threshold ──► FULL 100 mV spike, identical
                                     (but the NEXT one comes sooner)
   ⇒ AMPLITUDE carries no information. FREQUENCY carries all of it.

The phases, causally

① Resting state. V_m = −70 mV. Na⁺ activation gates closed, inactivation gates open, voltage-gated K⁺ channels closed. Nothing is happening except leak and pump.

② Depolarization to threshold. Summed graded potentials (§11.5) push V_m upward. Below −55 mV, nothing regenerative occurs; the voltage decays back if the input stops. At threshold (≈ −55 mV), enough voltage-gated Na⁺ channels open that the inward Na⁺ current exceeds the outward K⁺ leak current, and the process becomes self-sustaining.

③ Rising phase — a positive feedback loop with a built-in stop. Na⁺ entry depolarizes the membrane; depolarization opens more Na⁺ channels; more Na⁺ enters. Sodium permeability rises roughly 600-fold in under half a millisecond. This is the third example of physiological positive feedback promised in Chapter 1, and, as promised there, it terminates itself: the inactivation gates are closing on the same channels the whole time. The peak, around +30 mV, falls short of E_Na (+60 mV) precisely because inactivation and K⁺ opening arrive before equilibrium can be approached.

④ Repolarization. Two events, both essential. Na⁺ inactivation gates shut, cutting the inward current. Voltage-gated K⁺ channels, opening on their slow schedule, now let K⁺ pour out down a driving force that the depolarization has hugely increased. The membrane falls rapidly back toward E_K.

⑤ Hyperpolarization (undershoot). The K⁺ channels close slowly and are still open after V_m has passed −70 mV, so the membrane briefly overshoots toward E_K, reaching −80 to −90 mV. This is not an error; it is a functional refractory brake, and it is the physical basis of the relative refractory period.

⑥ Return to rest. K⁺ channels close, Na⁺ inactivation gates reopen, and the leak conductances restore −70 mV. The Na⁺/K⁺ ATPase restores the trivial quantities of Na⁺ and K⁺ that actually moved — a housekeeping task that is not part of the action potential and does not need to be complete before the next one can fire.

What each refractory period accomplishes

The absolute refractory period (0.4–1.0 ms) exists because Na⁺ channels are inactivated and no stimulus can reopen them. It does three things:

1. Sets a ceiling on firing rate. A hard floor of about 1 ms between spikes means a theoretical maximum near 1,000 Hz; real neurons top out at 100–300 Hz for sustained firing.
2. Forces one-way propagation. The patch of membrane the impulse just left is inexcitable, so the impulse cannot turn around. Direction is guaranteed not by any structural valve but by the timing of a molecular gate.
3. Keeps action potentials discrete. Because no second spike can start during the first, spikes cannot summate — which is what makes the output all-or-none and therefore uncorruptible over distance.

The relative refractory period (a further 2–5 ms) exists because some Na⁺ channels have reset while K⁺ channels remain open and the membrane is hyperpolarized. A stronger-than-normal stimulus can fire the cell during it. This is what converts stimulus intensity into firing frequency: a mild stimulus fires the neuron only once the cell has fully recovered, giving a low rate; an intense stimulus fires it during the relative refractory period, giving a high rate. The relative refractory period is where intensity coding physically happens.

11.7 Conduction Velocity: Why Some Signals Arrive Late

An action potential does not travel. Strictly, it is regenerated at every point along the axon: current from an active patch of membrane flows forward inside the axon, depolarizes the next patch to threshold, and that patch fires its own full-size spike. What propagates is a sequence of local events, which is why the signal never weakens. It also means conduction velocity is set by one question: how fast can the region ahead be brought to threshold?

Two variables answer it.

Axon diameter

A wider axon offers lower internal (axial) resistance to the local current, exactly as a wider pipe offers less resistance to water. Current therefore spreads further ahead before it dissipates, bringing more distant membrane to threshold sooner. In unmyelinated axons velocity scales with roughly the square root of diameter — an expensive relationship, since doubling velocity requires quadrupling cross-sectional area, and space inside a nerve is finite. The squid giant axon, which conducts an escape reflex at about 25 m/s, is 1 mm across; a human nerve trunk of that composition, carrying the same information, would be as thick as a wrist.

Myelination and saltatory conduction

Vertebrates solved the problem differently, and better. Wrapping an axon in myelin does two things at once (§11.3): it raises transverse membrane resistance, so current leaks out far less, and it lowers membrane capacitance, so far less charge is needed to swing the voltage. Current injected at one node therefore reaches the next node with enough amplitude to fire it.

Voltage-gated Na⁺ channels are then concentrated almost entirely at the nodes of Ranvier — densities of roughly 1,000–2,000 per µm² at a node, against fewer than 25 per µm² under the myelin. The action potential is regenerated only at nodes, and the impulse appears to jump from node to node: saltatory conduction, from saltare, to leap.

In myelinated fibers velocity scales linearly with diameter — approximately velocity (m/s) ≈ 6 × diameter (µm) — which is a far better return on space. A 20 µm myelinated fiber conducts at about 120 m/s; matching that with an unmyelinated axon would require a diameter measured in millimetres.

Saltatory conduction is also cheaper. Only the nodal membrane depolarizes, so only nodal ion movement occurs, so the Na⁺/K⁺ ATPase has far less to restore. A myelinated axon conducts roughly thirty times faster than an unmyelinated axon of the same diameter while consuming substantially less ATP per impulse. Faster and cheaper is an unusual combination in biology and it explains why myelin is universal in vertebrates.

  ══════ UNMYELINATED (C fiber, 0.5 µm, 1 m/s) ═══════════════════════════
   Every patch of membrane must depolarize in sequence. Continuous, slow.

   ─▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓─
     ●●●●▶
     ↑ Na+ channels spread EVENLY along the whole membrane
       t=0        t=1ms       t=2ms       t=3ms       t=4ms
       │──────────│───────────│───────────│───────────│
       0        1 mm        2 mm        3 mm        4 mm
   ATP cost: HIGH — the entire membrane surface must be repumped.

  ══════ MYELINATED (A-alpha fiber, 15 µm, 90 m/s) ═══════════════════════
   Only the NODES depolarize. Current LEAPS the insulated internodes.

        ┌──────┐    ┌──────┐    ┌──────┐    ┌──────┐
        │██████│    │██████│    │██████│    │██████│   ◄ myelin: high R,
   ─────┤██████├─╳──┤██████├─╳──┤██████├─╳──┤██████├──   low capacitance
        │██████│ ▲  │██████│ ▲  │██████│ ▲  │██████│
        └──────┘ │  └──────┘ │  └──────┘ │  └──────┘
                 │           │           │
              NODE OF     NODE        NODE      Na+ channel density:
              RANVIER                            at node  ~1000–2000/µm²
              (1–2 µm)                           under myelin  <25/µm²
        ├─ internode ─┤
          0.2–2.0 mm                    ≈ 100 × the axon diameter

        ╭─── local current flows INSIDE the axon, node → node ───╮
        │  ═══════════════════════════════════════════════════►  │
        ╰──────────────────────────────────────────────────────── ╯
        t=0        t=0.01ms   t=0.02ms   t=0.03ms
   ATP cost: LOW — only ~1% of the membrane area ever depolarizes.
   NET: ~30× faster than an unmyelinated axon of the SAME diameter.

  ══════ DEMYELINATED (the same fiber, in MS or Guillain-Barré) ══════════
        ┌──────┐              ┌──────┐
        │██████│  ✗ stripped  │██████│    Current now LEAKS out across
   ─────┤██████├──────────────┤██████├──  bare internodal membrane that
        │██████│   ~~~leak~~~ │██████│    has almost NO Na+ channels.
        └──────┘   ~~~leak~~~ └──────┘
                    ▼    ▼
        Arriving current at the next node falls below threshold
        ⇒ first SLOWING (velocity 90 → 20 m/s), then CONDUCTION BLOCK.
        Warming shortens Na+ channel opening → tips slowing into block
        = UHTHOFF PHENOMENON.

The fiber classification — and the key to the Case File

Peripheral nerve fibers are grouped by diameter, myelination, and velocity. This table is the single most useful object in the chapter, because it converts an anatomical measurement into a prediction about what a person will feel and when.

Read the two bolded rows against each other, because the entire Case File turns on the contrast.

A-delta fibers are thin but myelinated, so they conduct at 5–30 m/s. They innervate the skin densely, with small, well-defined receptive fields — each fiber reports on a few square millimetres. They carry the sharp, bright, instantly localizable pain of a paper cut.

C fibers are the thinnest axons in the body and have no myelin at all, so they conduct at 0.5–2 m/s — walking pace. They have large, overlapping receptive fields, they diverge widely in the spinal cord, and essentially all visceral afferents are C fibers. They carry dull, aching, burning, poorly localized pain, and it arrives a second or more after the A-delta signal.

Do the arithmetic for a stubbed toe. The distance from toe to spinal cord is about 1 metre. The A-delta signal, at 15 m/s, arrives in roughly 70 ms. The C-fiber signal, at 1 m/s, arrives in roughly 1,000 ms. The same injury generates two sensations nearly a second apart, on two different cables. You have felt this your whole life without naming it, and §11.10 names it.

11.8 The Synapse

A synapse is the junction at which one neuron communicates with another cell. It is the place where the nervous system stops being a wiring diagram and starts being a computer, because a synapse can be strengthened, weakened, blocked, modulated, or eliminated — and an axon cannot.

Electrical synapses: fast, simple, rare

At an electrical synapse the two cells are joined by gap junctions — paired connexon channels forming continuous aqueous pores between cytoplasms. Ions flow directly from one cell into the next.

• Essentially zero delay and usually bidirectional.
• Cannot invert the signal: excitation cannot be turned into inhibition.
• Cannot be finely modulated, and cannot amplify.
• Used where speed and synchrony matter more than computation: cardiac muscle (Chapter 18), single-unit smooth muscle in the gut and uterus (Chapters 23, 28), some brainstem escape circuits, and astrocytic networks.

Amara's heart depends absolutely on this. Cardiac myocytes are electrically coupled through gap junctions in their intercalated discs, which is what allows an entire ventricle to contract as one synchronized unit rather than as a bag of independent twitching cells.

Chemical synapses: slower, and worth every microsecond

At a chemical synapse the cells are physically separated by a synaptic cleft of 20–40 nm. The electrical signal must be converted into a chemical one, released, diffused across the gap, and converted back. This costs a synaptic delay of 0.3–0.5 ms — trivial for one synapse, substantial across the dozens in a decision path, as the reaction-time sidebar showed.

What the delay buys is everything the nervous system is for: the ability to make an excitatory signal inhibitory, to amplify a small input into a large one, to sum inputs from thousands of sources, to change the strength of a connection with use, and to be targeted by drugs. More than 90% of the pharmacology in this book acts at a chemical synapse.

  ══════════ CHEMICAL SYNAPTIC TRANSMISSION — SEVEN STEPS ═══════════════

          AXON TERMINAL (presynaptic)
   ┌──────────────────────────────────────────────────────────┐
   │  ① AP ARRIVES                                            │
   │  ═════════════════►  depolarizes the terminal            │
   │                                                          │
   │  ② VOLTAGE-GATED Ca²⁺ CHANNELS OPEN                      │
   │     ╔═══╗  Ca²⁺ ●  driving force is ENORMOUS:            │
   │     ║ ↓ ║      ●   1.2 mM outside vs 0.0001 mM inside    │
   │     ╚═╤═╝     ●    (a 10,000-fold gradient)              │
   │       ▼                                                  │
   │  ③ Ca²⁺ TRIGGERS VESICLE FUSION                          │
   │     Ca²⁺ binds SYNAPTOTAGMIN on the vesicle;             │
   │     SNARE proteins (synaptobrevin + syntaxin +           │
   │     SNAP-25) zipper the membranes together.              │
   │     ◯ ◯ ◯  docked vesicles ──► EXOCYTOSIS                │
   │     (~5,000–10,000 transmitter molecules per vesicle)    │
   │      ╲│╱                                                 │
   └───────┼──────────────────────────────────────────────────┘
           ▼   ④ DIFFUSION across the cleft
     ░░░░░░░░░░░░░░  20–40 nm — crossed in ~0.1 ms
     ░░ ▪ ▪ ▪ ▪ ░░  SYNAPTIC CLEFT
     ░░░░░░░░░░░░░░
           ▼
   ┌──────────────────────────────────────────────────────────┐
   │  ⑤ BINDING to receptors on the POSTSYNAPTIC membrane     │
   │     ╭────────╮        ╭────────╮                         │
   │     │IONOTROPIC│      │METABOTROPIC│                     │
   │     │ = the    │      │ = GPCR;    │                     │
   │     │ receptor │      │ receptor and│                    │
   │     │ IS the   │      │ channel are │                    │
   │     │ channel  │      │ SEPARATE    │                    │
   │     ╰────┬─────╯      ╰──────┬─────╯                     │
   │          │                   │                           │
   │  ⑥ POSTSYNAPTIC POTENTIAL    │                           │
   │     fast: 0.1–2 ms onset,    │  slow: 100s ms – minutes; │
   │     ms duration              │  2nd messengers AMPLIFY   │
   │     ▲ EPSP (Na+ in)          │  one molecule ⇒ thousands │
   │     ▼ IPSP (Cl- in / K+ out) │  of downstream events     │
   │                                                          │
   │     ⇒ GRADED POTENTIAL ⇒ summed at the axon hillock      │
   │       (§11.5) ⇒ fire or not                              │
   └──────────────────────────────────────────────────────────┘

     ⑦ TERMINATION — the signal MUST be switched off, or the
        synapse cannot transmit a second message. Three routes:
        • REUPTAKE by transporters into the terminal or into
          astrocytes  ── serotonin, norepinephrine, dopamine,
          GABA, glutamate.   [TARGET: SSRIs, cocaine, amphetamine]
        • ENZYMATIC DEGRADATION in the cleft ── acetylcholinesterase
          hydrolyses ACh in <1 ms.  [TARGET: neostigmine, nerve agents]
        • DIFFUSION away from the cleft ── always occurring; the
          only mechanism for peptides and gases.

  ═══ WHERE DRUGS AND TOXINS ACT, BY STEP NUMBER ════════════════════════
   ① tetrodotoxin, local anesthetics    ⑤ curare, atropine, nicotine,
   ② ω-conotoxin, gabapentin (α2δ)         benzodiazepines, opioids,
   ③ BOTULINUM TOXIN (cleaves SNAREs),      beta-blockers, antipsychotics
     TETANUS TOXIN (same, inhibitory     ⑦ SSRIs, cocaine, amphetamine,
     neurons), α-latrotoxin                 MAO inhibitors, neostigmine,
                                            organophosphates

Notice what step ② implies. Calcium is the trigger that converts electricity into secretion, and it does so wherever this coupling appears: at the synapse, in the cardiac myocyte, in the pancreatic beta cell releasing insulin, in the mast cell degranulating. Learn it once here and you will recognize it six more times before Chapter 33.

Ionotropic versus metabotropic receptors

The functional division is worth stating plainly: ionotropic receptors carry the message; metabotropic receptors set the volume. Amara's beta-blocker, which Chapter 13 will start her on, works at a metabotropic receptor — it does not block any signal, it turns down the responsiveness of her heart to the sympathetic drive that is raising its oxygen demand.

The neurotransmitter classes

Two of these rows connect directly to Amara. Her nitroglycerin works by releasing nitric oxide, which diffuses into vascular smooth muscle, activates guanylyl cyclase, raises cyclic GMP, and relaxes the vessel — a gasotransmitter used as a drug (Chapter 19). And substance P and CGRP are the peptides her cardiac C fibers release in the dorsal horn, which is where §11.10 picks the story up.

11.9 Neuronal Integration: Circuits and Codes

Individual neurons are not intelligent. Circuits are. This section covers the small number of wiring patterns from which the nervous system builds everything, and the code it uses to carry quantity.

The four basic circuit patterns

Divergence — one input, many outputs. A single presynaptic neuron branches to contact many postsynaptic cells. Divergence amplifies and distributes. A single upper motor neuron in the motor cortex ultimately influences hundreds of muscle fibers; a single preganglionic sympathetic neuron drives twenty or more postganglionic cells across several ganglia, which is why one stimulus made Amara pale, tachycardic, hypertensive, and diaphoretic simultaneously (§11.1).

Convergence — many inputs, one output. Many presynaptic neurons contact one postsynaptic cell. Convergence integrates. The alpha motor neuron is the standard example: descending cortical commands, cerebellar corrections, spinal reflex arcs, and inhibitory interneurons all terminate on the same cell, whose firing is the arithmetic sum of everything the nervous system has to say about that muscle at that moment — which is why it is called the final common pathway. Convergence is also what produces referred pain, and §11.10 uses it directly.

Reverberating (oscillating) circuits feed a collateral branch back onto a neuron earlier in the chain, so the circuit re-excites itself and continues firing after the input has stopped. It runs until fatigue or inhibition stops it. This is how the brainstem generates the rhythm of breathing (Chapter 22), how sleep-wake cycling is sustained (Chapter 12), and how short-term working memory holds a phone number for the twenty seconds between hearing it and dialing it.

Parallel after-discharge circuits split an input among several chains of different lengths that reconverge on one output cell. Because the paths differ in synapse number, signals arrive staggered in time, so a single brief input produces a burst of output lasting far longer than the input did. There is no feedback — the difference from a reverberating circuit — so the output is finite and precisely bounded. This arrangement appears where an exact, terminating computation is needed.

Neural coding: how an all-or-none signal carries quantity

The all-or-none principle creates a genuine problem. Every action potential in a given axon is identical, so amplitude cannot encode anything. Yet you can plainly tell a whisper from a shout and a warm cup from a scalding one. Four solutions are used together.

1. Frequency (rate) coding. Stimulus intensity is encoded as spikes per second. A gentle touch might drive an afferent at 10 Hz; a firm press, 80 Hz. This works precisely because of the relative refractory period (§11.6): a stronger stimulus can fire the cell before it has fully recovered. The absolute refractory period sets the ceiling. Note that the relationship is logarithmic, not linear — a tenfold increase in stimulus produces roughly a doubling of firing rate, which is why perceived intensity compresses across an enormous physical range.
2. Population (recruitment) coding. Stronger stimuli recruit more afferents, because receptors differ in threshold and because a larger stimulus reaches more of them. The CNS reads intensity partly from how many fibers are reporting at all.
3. Labeled-line coding. Which fiber is firing determines what is felt, regardless of how it was stimulated. Pressing on the eyeball produces a flash of light because mechanical deformation fires optic nerve fibers, and any activity in that line is interpreted as light. This is the single most important idea in Chapter 15 and it is also the key to referred pain: the brain does not perceive the stimulus, it perceives which line reported.
4. Temporal patterning. Bursts, pauses, and precise inter-spike intervals carry information beyond mean rate.

Adaptation is the fifth variable. Phasic (rapidly adapting) receptors fire at the onset and offset of a stimulus and fall silent during it — which is why you stop feeling your wristwatch. Tonic (slowly adapting) receptors keep firing for as long as the stimulus lasts, because their information must not be allowed to disappear. Nociceptors are tonic and some actually sensitize — firing more over time to the same stimulus. A pain that faded would be useless, and Amara's did not fade in forty-three minutes.

11.10 Advanced Topic · Pain

Pain is the most instructive sensation in physiology, because it is the one where the gap between the stimulus and the experience is widest and most obviously constructed. Everything in this chapter converges here.

Nociceptors: what actually detects damage

A nociceptor is a free nerve ending — no capsule, no accessory structure, just the bare terminal branches of an A-delta or C fiber. There are millions of them, in skin, muscle, joints, periosteum, blood vessel walls, and viscera, and they are the most widely distributed receptor type in the body. They are not simple damage detectors; they are chemically sophisticated cells carrying a suite of specific channels.

That last row is what makes NSAIDs work. Prostaglandins lower the nociceptor's threshold rather than exciting it, so inflamed tissue hurts when touched lightly — primary hyperalgesia. Aspirin and ibuprofen inhibit cyclooxygenase, prostaglandin synthesis falls, and the threshold returns toward normal. The drug does not block pain; it removes an amplifier.

Why a 90% coronary stenosis hurts. Ischemic myocardium produces and releases adenosine, ATP, bradykinin, lactate and hydrogen ions, and serotonin. Hydrogen ions open ASIC and TRPV1 channels; ATP opens P2X3; adenosine — probably the principal culprit in angina — acts at A1 receptors on cardiac afferents. So the heart does not have a "pain nerve" that detects blockage. It has chemically tuned free nerve endings that detect the metabolic signature of tissue working without enough oxygen. That distinction explains why angina correlates with demand and not with anatomy: Amara's artery was 90% blocked while she slept and did not hurt, and hurt at the end of a twelve-hour shift, because pain tracks the supply-demand gap rather than the stenosis.

The two-pain phenomenon

Stub your toe and attend carefully. There are two sensations, and they are separated by about a second.

• First pain: sharp, bright, pricking, easy to localize to within millimetres, brief. Carried by A-delta fibers at 5–30 m/s.
• Second pain: dull, aching, burning, throbbing, hard to localize, and it lingers. Carried by C fibers at 0.5–2 m/s.

This is the fiber table of §11.7 made into an experience. The delay is not processing time; it is conduction time down two cables of very different speeds. And the qualities differ because the two systems terminate differently in the brain: A-delta signals project through the lateral spinothalamic tract to the ventral posterolateral thalamus and thence to somatosensory cortex, which contains a precise body map, while C-fiber signals project heavily to the reticular formation, the medial thalamus, the insula, and the anterior cingulate — regions concerned with arousal, autonomic response, and the unpleasantness of pain rather than its location. First pain tells you where. Second pain tells you how bad, and makes you care.

Gate control: why rubbing it helps

Melzack and Wall proposed in 1965 that the dorsal horn contains a "gate" that can be closed against nociceptive traffic before it ever reaches the brain. The essentials have held up.

The projection neuron in the dorsal horn receives excitatory input from C fibers. It also receives input from an inhibitory interneuron in the substantia gelatinosa. That interneuron is excited by large A-beta fibers — the fast touch, pressure, and vibration fibers — and inhibited by C fibers.

So the two inputs compete. C-fiber traffic opens the gate by silencing the inhibitory interneuron; A-beta traffic closes it by exciting the same interneuron. When you rub a banged elbow, you are flooding the segment with A-beta activity, driving the inhibitory interneuron, and reducing the nociceptive signal that reaches the brain. Rubbing works, and it works because of a wiring diagram.

Two clinical technologies are this mechanism sold: TENS (transcutaneous electrical nerve stimulation) deliberately stimulates A-beta fibers at intensities below the pain threshold, and dorsal column stimulators implant electrodes to do the same thing centrally. Both are the 1965 diagram with a battery.

A second, independent gate runs top-down. Neurons in the periaqueductal gray of the midbrain project to the raphe nuclei and thence to the dorsal horn, where they release serotonin and norepinephrine and drive enkephalin-containing interneurons that inhibit nociceptive transmission both pre- and postsynaptically. This descending analgesic system is the target of opioid drugs and of the body's own endogenous opioids, and it is why the same injury hurts differently depending on context, expectation, and attention.

Referred pain: the resolution

Everything now assembles.

Visceral afferents are sparse. The skin of a fingertip carries hundreds of nociceptors per square centimetre. The heart, gut, and other viscera carry orders of magnitude fewer, and visceral afferents make up only about 10% of all afferent fibers entering the spinal cord. A sparse array cannot support fine localization for the same reason a camera with a hundred pixels cannot resolve a face.

Visceral afferents diverge enormously. A single visceral afferent entering the cord branches over five or more spinal segments rostrally and caudally in the tract of Lissauer before synapsing. A somatic afferent stays essentially within its segment. So one cardiac afferent excites second-order neurons across a wide swath of cord.

Visceral and somatic afferents converge on the same second-order neurons. This is the mechanism, and it is worth stating carefully: there is no separate visceral pathway to the brain. Visceral nociceptive information ascends on the same spinothalamic neurons that carry somatic pain from the corresponding body wall. The second-order neuron receives both and cannot distinguish them.

The brain resolves the ambiguity using prior experience. As §11.9 established, sensation is determined by which line fires. The line in question has carried skin signals thousands of times and cardiac signals almost never. The brain's interpretation is a Bayesian bet on the common case — and it is wrong exactly when the rare case occurs.

There is no cortical map of the viscera. The somatosensory homunculus (Chapter 12) devotes enormous territory to the hand and lips and essentially none to the heart or gut. There is no place in cortex for the sensation to be put, so it is placed on the body wall belonging to the same spinal segment.

   ═══════ CONVERGENCE-PROJECTION: WHY THE HEART HURTS IN THE ARM ════════

   THE HEART                              THE BODY WALL
   ischemic myocardium                    skin, muscle of chest + medial arm
        │                                          │
        │ C fibres (+ some A-δ)                    │ A-δ and A-β fibres
        │ VISCERAL afferents travel                │ SOMATIC afferents
        │ WITH the cardiac sympathetic             │ enter their OWN segment
        │ nerves → cardiac plexus →                │ only
        │ middle + inferior cervical and           │
        │ upper thoracic sympathetic ganglia       │
        │ → white rami communicantes               │
        ▼                                          ▼
   ┌════════════════════════════════════════════════════════════┐
   │  DORSAL ROOT GANGLIA and SPINAL CORD SEGMENTS  T1 – T5     │
   │                                                            │
   │   T1 ──┐                    ╭──────────────────────╮       │
   │   T2 ──┤   visceral input   │  SECOND-ORDER        │       │
   │   T3 ──┼──── DIVERGES ─────►│  PROJECTION NEURON   │◄──────┤ somatic
   │   T4 ──┤   over ≥5 segments │  (dorsal horn)       │       │ input
   │   T5 ──┘                    │                      │       │ (segmental)
   │                             │  RECEIVES BOTH.      │       │
   │                             │  CANNOT TELL THEM    │       │
   │                             │  APART.              │       │
   │                             ╰──────────┬───────────╯       │
   └════════════════════════════════════════│═══════════════════┘
                                            │ ONE spinothalamic
                                            │ axon, crossing the
                                            │ midline, ascending
                                            ▼
                              ┌──────────────────────────┐
                              │  THALAMUS → SOMATOSENSORY │
                              │  CORTEX                   │
                              │  "This line means CHEST   │
                              │   WALL and MEDIAL ARM —   │
                              │   it always has."         │
                              └──────────────────────────┘
                                            │
                                            ▼
                        PAIN IS PERCEIVED AT THE DERMATOMES
                        SHARING THE SEGMENT, NOT AT THE ORGAN

   ══════ THE JAW: A SECOND, SEPARATE CONVERGENCE ════════════════════════
        VAGUS NERVE (CN X) cardiac branches ──► nucleus of the solitary
        tract and the SPINAL TRIGEMINAL NUCLEUS, whose caudal end
        extends down to C2–C3 and receives BOTH trigeminal (face, jaw)
        AND upper cervical (C1–C3) afferents.
        ⇒ the TRIGEMINOCERVICAL COMPLEX: one more convergence point,
          one more shared line, one more site of referral. → JAW, NECK

   ══════ THE SAME RULE, OTHER ORGANS ════════════════════════════════════
   ORGAN            SEGMENTS      PAIN IS FELT AT
   ──────────────────────────────────────────────────────────────────────
   Heart            T1–T5         substernal · medial L arm · jaw · neck
   Diaphragm        C3–C5         TIP OF THE SHOULDER (phrenic n., C3–5)
   Stomach          T6–T9         epigastrium
   Appendix/midgut  T10           periumbilical → then RLQ when the
                                  PARIETAL peritoneum is touched (somatic)
   Gallbladder      T5–T9         R hypochondrium + inferior angle of
                                  the R scapula
   Kidney/ureter    T10–L1        flank → groin ("loin to groin")
   ──────────────────────────────────────────────────────────────────────
   THE RULE: pain from a viscus is referred to the DERMATOMES supplied
   by the SAME SPINAL SEGMENTS the viscus sends its afferents to — which
   are the segments it lay next to as an EMBRYO, wherever it later moved.

Chapter Summary

§11.1 The nervous system is divided anatomically into CNS (brain, spinal cord) and PNS (cranial and spinal nerves, ganglia, receptors), and functionally into afferent (receptor → CNS) and efferent (CNS → effector) limbs. The efferent limb divides into the somatic division — one neuron, skeletal muscle, always excitatory — and the autonomic division — two neurons with a ganglionic synapse, cardiac and smooth muscle and glands, either excitatory or inhibitory. The autonomic division divides again into sympathetic (thoracolumbar, divergent, fight-or-flight) and parasympathetic (craniosacral, discrete, rest-and-digest).

§11.2 A neuron is amitotic, long-lived, and metabolically expensive. Input arrives on branched dendrites carrying graded potentials; the soma with its Nissl bodies and neurofibrils is the biosynthetic center; the axon hillock decides; the single axon transmits action potentials without decay. Neurons are multipolar (>99%), bipolar (special senses), or unipolar (sensory ganglia — the type carrying Amara's pain). Because axons contain no ribosomes, everything is manufactured in the soma and moved by kinesin anterogradely (200–400 mm/day fast) and dynein retrogradely; slow transport at ~1 mm/day sets the rate of nerve regeneration.

§11.3 Six glia. Astrocytes maintain the blood-brain barrier, buffer K⁺, recycle glutamate, and form the growth-blocking glial scar. Microglia are yolk-sac-derived immune cells that phagocytose and prune. Ependymal cells line the ventricles and make CSF. Oligodendrocytes myelinate up to 60 CNS axons each. Schwann cells myelinate one PNS internode each and leave a neurilemma that permits regeneration. Satellite cells surround PNS somata. The presence or absence of the neurilemma is why peripheral nerves regrow and central ones do not.

§11.4 The resting membrane potential of about −70 mV arises from three things: unequal ion distributions (K⁺ in, Na⁺ out, with trapped intracellular anions), a membrane 25–40× more permeable to K⁺ than Na⁺, and the Na⁺/K⁺ ATPase that maintains the gradients. The resting voltage is the permeability-weighted average of the equilibrium potentials — near E_K (−90 mV) but pulled about 20 mV positive by the Na⁺ leak. The pump contributes only ~3 mV directly but everything indirectly.

§11.5 Graded potentials are variable in size, decremental, and summable — the currency of computation. EPSPs move the membrane toward threshold; IPSPs away from it or clamp it in place by shunting. Because a single EPSP is 0.5–1 mV and threshold is 15 mV away, firing requires temporal or spatial summation of 15–50 inputs. The axon hillock, with ~50× the somatic density of voltage-gated Na⁺ channels, has the lowest threshold in the cell and is therefore the sole point of decision.

§11.6 The action potential is a ~100 mV, ~1 ms, all-or-none, self-propagating reversal of polarity. It exists because the voltage-gated Na⁺ channel has two gates — a fast activation gate and a slow inactivation gate — creating a third, non-conducting inactivated state that only repolarization can reset. This generates the absolute refractory period, which sets the maximum firing rate, forces one-way propagation, and keeps spikes discrete; and the relative refractory period, in which stimulus intensity is converted into firing frequency.

§11.7 Conduction velocity rises with axon diameter (as the square root, unmyelinated; linearly, myelinated) and with myelination, which concentrates Na⁺ channels at the nodes of Ranvier and produces saltatory conduction — about 30× faster and cheaper. The fiber classification ties structure to sensation: A-alpha (80–120 m/s, proprioception and motor), A-beta (35–75 m/s, touch), A-delta (5–30 m/s, sharp localized first pain), B (preganglionic autonomic), and C (0.5–2 m/s, unmyelinated, dull poorly localized second pain and all visceral afferents).

§11.8 Electrical synapses use gap junctions and give speed and synchrony without computation. Chemical synapses cost 0.3–0.5 ms and buy everything else, in seven steps: action potential arrival, voltage-gated Ca²⁺ entry, SNARE-mediated vesicle fusion, diffusion across a 20–40 nm cleft, receptor binding, postsynaptic potential, and termination by reuptake, enzymatic degradation, or diffusion. Ionotropic receptors are fast and non-amplifying and carry the message; metabotropic receptors are slow and enormously amplifying and set the gain.

§11.9 Circuits are built from divergence (amplify, distribute), convergence (integrate — the motor neuron as final common pathway), reverberation (sustain a rhythm), and parallel after-discharge (prolong a bounded output). Because spikes are identical, intensity is encoded as frequency, supported by population coding, labeled lines, and temporal patterning, and shaped by phasic versus tonic adaptation.

§11.10 Nociceptors are free nerve endings carrying TRPV1, TRPM8, ASIC, and P2X3 channels that detect the chemical signature of tissue damage and ischemia. The two-pain phenomenon is the fiber table made into experience. Gate control explains why A-beta input suppresses C-fiber transmission in the dorsal horn, and a descending system from the periaqueductal gray does the same from above. Referred pain arises because visceral afferents are sparse, diverge over ≥5 segments, and converge on the same second-order neurons as somatic afferents from the corresponding dermatomes — so the brain, using a labeled-line rule that is almost always right, assigns cardiac pain to the T1–T5 and trigeminocervical body wall.

Case File 11 · Resolution

Question 1 — How can an injury confined to the heart produce a sensation in the jaw and the medial arm, when both are structurally intact?

Through convergence-projection (§11.10), and through it twice, by two different routes.

The arm. Visceral afferents from the heart are predominantly C fibers. They do not run in a nerve of their own; they travel with the cardiac sympathetic nerves, through the cardiac plexus, up to the middle and inferior cervical and upper thoracic sympathetic ganglia, then through the white rami communicantes into the dorsal root ganglia of segments T1 through T5. On entering the cord each afferent diverges over five or more segments, then synapses on second-order spinothalamic projection neurons in the dorsal horn.

Those same second-order neurons also receive somatic afferents — from the skin and muscle of the chest wall and, through the T1–T2 contribution to the medial brachial cutaneous and intercostobrachial nerves, from the medial aspect of the left arm. One neuron, two sources, one axon to the thalamus. The brain receives a signal on a line and must decide what it means. It has received signals on that line from skin thousands of times and from the heart almost never, so it applies the interpretation that has been correct every previous time. The pain is projected to the body wall of the shared segments.

The jaw. This is a second, separate convergence, and it uses a different nerve. Cardiac afferents also travel in the vagus nerve, whose fibers terminate in the nucleus of the solitary tract and in the caudal spinal trigeminal nucleus. The caudal end of that nucleus descends to the C2–C3 level of the cord and receives both trigeminal afferents from the face and jaw and upper cervical afferents — the trigeminocervical complex. Convergence there produces referral to the mandible and neck by exactly the same logic operating one level up.

So nothing is wrong with her jaw or her arm, and nothing needs to be. The sensation is generated entirely in her heart and addressed by her brain to two regions that share a second-order neuron with it. The addressing rule is a good rule. It simply meets, perhaps twice in a lifetime, an input it was never calibrated for.

Question 2 — Why is visceral pain so poorly localized when a paper cut is pinpoint-precise?

Four independent reasons stack, and each is quantitative (§11.7, §11.9, §11.10).

1. Receptor density. Skin carries hundreds of nociceptors per square centimetre with small receptive fields; the viscera carry orders of magnitude fewer with large ones. Visceral afferents are only about 10% of all afferents entering the cord. Localization is a sampling problem, and the visceral array is under-sampled.
2. Divergence. A visceral afferent branches over five or more spinal segments before synapsing; a cutaneous afferent stays essentially within one. Spreading one signal across a third of the thoracic cord destroys segmental precision by construction.
3. Convergence with somatic input. There is no separate visceral pathway to the brain. The information ascends on neurons that also carry somatic pain, so at the very first synapse it has already been mixed with a signal from somewhere else.
4. No cortical map. The somatosensory homunculus (Chapter 12) allocates vast cortical territory to the fingertips and lips and essentially none to the heart or gut. There is no place in the cortex for a visceral sensation to be put, so it is put on the nearest mapped surface — the body wall of the same segments.

The paper cut has the opposite of all four: dense innervation, tiny receptive fields, segmental entry, a dedicated pathway, and an enormous cortical representation. Amara covering her sternum with a flat palm rather than pointing with a fingertip is not vagueness in her description. It is an accurate report of the resolution of the system doing the reporting.

Question 3 — Why "pressure" rather than sharp pain?

Because of which fibers were carrying the signal, and where those fibers terminate (§11.7, §11.10).

Sharp, bright, instantly localized first pain is carried by A-delta fibers, which are thin but myelinated, conduct at 5–30 m/s, and project through the lateral spinothalamic tract to the ventral posterolateral thalamus and on to the mapped somatosensory cortex. They densely innervate skin. They barely innervate the heart.

Cardiac nociception is carried almost entirely by C fibers: unmyelinated, 0.5–2 m/s, large overlapping receptive fields, and projecting heavily to the reticular formation, medial thalamus, insula, and anterior cingulate cortex — regions that encode arousal, autonomic response, and the unpleasantness of a sensation rather than its position. A signal arriving there, with no somatotopic map to land on, is experienced as a diffuse, heavy, oppressive quality. There is no cortical machinery available to render it as a point.

The onset supports this. Her pain built over two to three minutes, because it was generated not by a mechanical event but by the gradual accumulation of ischemic metabolites — hydrogen ions, adenosine, ATP, bradykinin — acting on chemically tuned free nerve endings. Chemistry accumulates; a blade does not.

This is why the phrasing matters clinically, and why it cost Amara an hour. Patients with myocardial ischemia very often deny "pain" when asked, because the word they attach to the sensation is pressure, heaviness, tightness, squeezing, or burning. Experienced clinicians ask "is there any discomfort in your chest?" for exactly this reason. The vocabulary problem is downstream of a fiber-type fact: you cannot feel visceral injury sharply, because you have no fibers that would let you.

Systems Integration Case File · Entry 11

Review

Level 1 · Recall

10.1 The resting membrane potential is closest to the equilibrium potential for:

a) Na⁺    b) K⁺    c) Cl⁻    d) Ca²⁺

10.2 The absolute refractory period exists because:

a) the Na⁺/K⁺ pump has not yet restored the gradients b) voltage-gated K⁺ channels are closed c) voltage-gated Na⁺ channels are inactivated d) the membrane is hyperpolarized

10.3 Which fiber type carries dull, poorly localized, aching visceral pain?

a) A-alpha    b) A-beta    c) A-delta    d) C

10.4 One oligodendrocyte myelinates:

a) one internode of one axon b) the entire length of one axon c) one internode each of up to 50–60 axons d) the cell bodies of neurons in a ganglion

10.5 Temporal summation is:

a) several synapses firing simultaneously at different sites b) one synapse firing rapidly enough that successive EPSPs overlap c) the addition of an EPSP and an IPSP d) the summation of two action potentials

10.6 Neurotransmitter release from the axon terminal is triggered by the entry of:

a) Na⁺    b) K⁺    c) Ca²⁺    d) Cl⁻

10.7 Which of the following is a property of graded potentials but not action potentials?

a) they are all-or-none b) they can be either depolarizing or hyperpolarizing c) they propagate without decrement d) they have a threshold

10.8 A patient's serum potassium is 7.4 mEq/L. Compared with normal, the resting membrane potential of her cardiac cells is:

a) more negative, and the cells are less excitable b) less negative, and the cells are eventually less excitable c) less negative, and the cells are permanently more excitable d) unchanged

Level 2 · Comprehension

10.9 Explain why the axon hillock, rather than the dendrite receiving the strongest input, is the site at which a neuron decides to fire.

10.10 Myelination makes conduction both faster and cheaper. Explain each half.

10.11 A neuron receives a constant excitatory drive. Explain how the nervous system encodes a doubling of stimulus intensity, given that every action potential is identical.

10.12 Both botulinum toxin and tetanus toxin cleave SNARE proteins, yet one produces flaccid paralysis and the other rigid spasm. Explain.

Level 3 · Clinical Application

10.13 A 32-year-old woman reports that vision in her right eye became blurred and painful over two days, then largely recovered over three weeks. Eighteen months later she develops numbness in both legs. She notes that her vision blurs again every time she takes a hot bath, recovering within twenty minutes of getting out. Explain each element mechanistically.

10.14 A 68-year-old man with 20 years of type 2 diabetes presents with three days of increasing breathlessness and fatigue. He denies chest pain entirely. His ECG shows changes of a completed myocardial infarction. Explain why he had no pain, and state what this predicts about sensation in his feet.

10.15 A patient is given a spinal anesthetic before surgery. Over ten minutes she loses, in order: the ability to feel a pinprick, then the ability to distinguish warm from cold, then light touch, then the ability to move her legs. Explain the sequence, and predict the order of recovery.

Level 4 · Integration and Synthesis

10.16 Amara's troponin is elevated (Chapter 3), her ECG shows ST-segment depression (Chapter 18), and she reports pressure in her jaw. Build a single causal chain that starts with one coronary artery and ends with each of those three findings, naming the level of organization (Chapter 1) at every step.

10.17 Argue for or against: "The action potential is the fundamental unit of nervous system function." Use at least four specific pieces of evidence from this chapter.

10.18 Design a nervous system for an animal that must react in under 5 ms to a threat and also perform complex learning. Specify the fiber types, myelination strategy, synapse types, and circuit patterns you would use for each requirement, and explain the trade-off that forces you to build two systems rather than one.

Concept Map to Complete

Copy this onto blank paper and fill in every bracket from memory before checking the chapter. Then, in a different color, add what you missed.

                    RESTING MEMBRANE POTENTIAL ≈ [ ___ ] mV
                                    │
              ┌─────────────────────┼─────────────────────┐
     [ ________ ]              [ __________ ]        Na+/K+ ATPase
     gradients                 permeability          [ __ ] Na+ out
     K+ in = [ ___ ] mM        P_K : P_Na            [ __ ] K+ in
     Na+ out= [ ___ ] mM       = [ ___ ] : 1         direct effect [ __ ] mV
                                    │
                          GRADED POTENTIAL
                    ┌───────────────┴───────────────┐
              [ _____ ]                        [ _____ ]
              toward threshold                 away / clamped
              ion: [ ___ ] in                  ions: [ ___ ] in or [ ___ ] out
                    └───────────────┬───────────────┘
                        summed at [ ______________ ]
                        by [ _______ ] + [ _______ ] summation
                                    │
                            threshold [ ___ ] mV
                                    │
                          ACTION POTENTIAL
              ┌─────────────────────┼──────────────────────┐
      depolarization          repolarization        hyperpolarization
      Na+ [ ___ ] gate        Na+ [ ___ ] gate      [ ___ ] channels
      OPENS (fast)            CLOSES (slow)         still open
              │                      │
      [ ________ ] refractory   [ ________ ] refractory
      accomplishes:             accomplishes:
      1 [ ______________ ]      1 [ ______________ ]
      2 [ ______________ ]
      3 [ ______________ ]
                                    │
                        CONDUCTION down the axon
                    ┌───────────────┴───────────────┐
              unmyelinated                    myelinated
              velocity ∝ [ ______ ]           velocity ∝ [ ______ ]
              [ __ ] m/s = C fibre            [ __ ] m/s = A-alpha
              carries [ ____________ ]        carries [ ____________ ]
                                    │
                              THE SYNAPSE
                    step 2 trigger ion = [ ___ ]
                    step 7 termination: [ ____ ] / [ ____ ] / [ ____ ]
                                    │
                            REFERRED PAIN
                    heart afferents enter [ __ ]–[ __ ]
                    converge with somatic afferents from [ __________ ]
                    jaw referral is via [ __________ ] nerve → [ _________ ]

Lab / Self-Exploration

1. Find your own two-pain phenomenon. Sit quietly with a bare foot. Have someone press a fingernail firmly into the pad of your great toe for one second, with your eyes closed. Attend carefully: there is a sharp sensation, then, roughly a second later, a duller, spreading ache. Now repeat on the pad of your thumb. The gap is shorter — because the conduction distance is shorter, and the difference between 15 m/s and 1 m/s over 25 cm is much less than over 100 cm. You have just measured two conduction velocities with a fingernail.
2. Demonstrate gate control on yourself. Next time you knock a knuckle, note the pain level, then rub the surrounding skin firmly for twenty seconds and note it again. You are driving A-beta afferents into the same spinal segment, exciting the inhibitory interneuron, and closing the gate on C-fiber transmission. Note that rubbing the injured spot itself helps less than rubbing around it — predict why before deciding.
3. Map your own two-point discrimination. Bend a paper clip into a U and measure the distance between the tips. With eyes closed, have a partner touch either one tip or two to your skin and report which. Find the smallest separation you can reliably distinguish on the fingertip (typically 2–4 mm), the lip (2–5 mm), the back of the hand (20–30 mm), the calf (40–50 mm), and the back (50–70 mm). Plot it. You have measured receptive field size, which is the peripheral half of the explanation for the cortical homunculus you will meet in Chapter 12 — and the same principle that makes visceral pain unlocalizable.
4. Watch adaptation happen. Place a coin on the back of your forearm and time how long until you stop feeling it. That is a phasic receptor adapting. Now hold an ice cube against the same spot: the sensation does not fade, because thermal and nociceptive afferents are tonic. Ask yourself which behaviour each design is for.
5. Estimate your reaction time. Have a partner hold a 30 cm ruler vertically with the zero mark between your open thumb and forefinger. They drop it without warning; you catch it. The distance fallen gives the time: 5 cm ≈ 100 ms, 10 cm ≈ 143 ms, 20 cm ≈ 202 ms, 30 cm ≈ 247 ms. Do ten trials and take the median. Then repeat with your eyes closed and your partner saying "now" as they release — auditory reaction time is typically 30–50 ms faster. Using the table in the §11.7 exercise sidebar, estimate what fraction of your result was conduction.
6. Find a referred-pain zone in a textbook body. Using Figure 11.9, predict where a person with an inflamed gallbladder, a kidney stone in the upper ureter, and an irritated diaphragm would each report pain. Then check yourself against the table. Do it from the segment numbers, not from memory of the answers — the point is to run the mechanism forwards.

Key Terms

Next: Chapter 12 · The Central Nervous System — where these neurons are assembled into a brain and a spinal cord, and where the question of what the brain does with Amara's twenty years of night shifts finally gets an answer.