📖 Core Concepts
BIOL501: Neuroscience, From Molecules to Mind
Core Concepts
Conceptual Introduction: What Is This Subject Really About?
Neuroscience asks how a lump of tissue that is about 75% water, made of the same lipids, proteins, and ions you have studied for years, comes to represent the world, decide what to do, and remember what happened. The surprising answer, built up over a century, is that no new physics or chemistry is required. The brain runs on things you already know.
- Electrochemistry (CHEM102). Ion gradients across a membrane are batteries. The Nernst equation tells you their voltages.
- Protein conformational change (CHEM401). Ion channels and receptors are allosteric machines. They switch between states in response to voltage, ligands, or force.
- Enzyme kinetics and signal amplification (CHEM402). G protein cascades and kinase networks turn one binding event into thousands of second-messenger molecules.
- Gene expression (BIOL201). Long-term memory requires new transcription and translation, just as development requires transcriptional programs.
- Peptide chemistry (CHEM202, CHEM401). A whole class of neural signals consists of peptides cut out of larger precursors by specific proteases.
The deep idea is that information is carried by changes in membrane potential, and computation happens where those changes interact. Two features recur throughout the course:
- Neurons are electrically excitable. They can produce an all-or-none, regenerating signal (the action potential) that travels long distances without decaying.
- Neurons talk chemically across synapses. The strength of a synapse is modifiable, and that modifiability is the physical substrate of learning.
A useful analogy is to think of the brain as a society rather than a computer. Each neuron is a citizen that listens to thousands of others (dendrites) and broadcasts a binary decision (spike or no spike) to thousands more (axon terminals). Its "opinions," meaning its synaptic weights, change with experience. Neuromodulators such as dopamine, serotonin, and neuropeptides act like the mood of the society: they do not carry specific messages but change how every citizen responds. Kandel and colleagues frame the field around this reductionist-but-integrative program, explaining behavior "in terms of the activities of the brain" [Kandel et al., 2021]. Bear, Connors, and Paradiso build upward in the same way, from ions to circuits [Bear et al., 2016].
The central bet of modern neuroscience is that mind is what the brain does, and that what the brain does can be understood level by level: molecules → synapses → cells → circuits → behavior. The exciting part is that the hard explanatory gaps live between levels, not within them.
Concept 1: Neurons and Glia, Two Cell Classes in a Partnership
The neuron doctrine holds that the nervous system is made of discrete cells, not a continuous net. It was championed by Santiago Ramón y Cajal, ironically using Golgi's silver stain. Golgi himself argued for the reticular theory. The doctrine came with a corollary Cajal called dynamic polarization: information flows dendrites → soma → axon → terminals.
Neuronal morphology is function made visible.
- A cerebellar Purkinje cell has a flat, fan-like dendritic tree receiving roughly 105 parallel-fiber synapses. It is built to integrate enormous convergence.
- A spinal motor neuron sends an axon up to a meter long to a muscle. It is built for reliable long-distance transmission.
- Retinal bipolar cells barely spike at all. They signal with graded potentials over short distances.
Glia are not just glue. In the human brain, glia and neurons are roughly equal in number. Each glial type has a specific job:
- Astrocytes buffer extracellular K⁺. They take up glutamate through the EAAT transporters and convert it to glutamine for return to neurons (the glutamate–glutamine cycle). They supply lactate and contribute to the blood-brain barrier and synapse formation.
- Oligodendrocytes (CNS) and Schwann cells (PNS) make myelin. One oligodendrocyte myelinates segments of many axons. One Schwann cell myelinates a single internode.
- Microglia are the brain's resident macrophages. They derive from yolk-sac progenitors, not bone marrow, a link to BIOL401. They prune synapses during development using complement tags (C1q, C3), the same molecules that opsonize bacteria.
The blood-brain barrier (BBB) is formed mainly by brain capillary endothelial cells joined by tight junctions (claudin-5, occludin). Pericytes and astrocytic endfeet help induce and maintain it. The endothelium lacks fenestrations and has very little transcytosis. As a rough rule, molecules that cross passively are small (under ~400–500 Da), lipophilic, and have few hydrogen-bond donors. Everything else needs a carrier: GLUT1 for glucose, LAT1 for large neutral amino acids (which is how L-DOPA gets in), or receptor-mediated transcytosis. Efflux pumps such as P-glycoprotein actively eject many drugs. This one fact governs much of neuropharmacology, and it is why peptide drugs for the brain are so hard to make (see the peptide chemistry section).
Concept 2: The Resting Membrane Potential Is a K⁺ Diffusion Potential
You met the Nernst equation in electrochemistry. Here it describes the voltage at which the electrical force on an ion exactly balances its concentration gradient:
E_ion = (RT / zF) · ln( [ion]_out / [ion]_in ) At 37 °C: 2.303·RT/F ≈ 61.5 mV, so E_ion ≈ (61.5 mV / z) · log10( [ion]_out / [ion]_in )
With typical mammalian concentrations:
Ion [out] (mM) [in] (mM) E_ion K⁺ 5 140 61.5·log(5/140) ≈ −89 mV Na⁺ 145 12 61.5·log(145/12) ≈ +67 mV Cl⁻ 110 10 −61.5·log(110/10) ≈ −64 mV Ca²⁺ 2 0.0001 30.75·log(2/0.0001) ≈ +132 mV
At rest, a real membrane is permeable to several ions at once. The Goldman–Hodgkin–Katz (GHK) voltage equation [Goldman, 1943; Hodgkin & Katz, 1949] gives the steady-state potential at which the net current is zero:
V_m = (RT/F) · ln( (P_K[K⁺]_o + P_Na[Na⁺]_o + P_Cl[Cl⁻]_i)
/ (P_K[K⁺]_i + P_Na[Na⁺]_i + P_Cl[Cl⁻]_o) )
Note: for the anion Cl⁻, the inside and outside concentrations are swapped.
Plug in the resting permeability ratios Hodgkin and Katz measured in squid axon, PK : PNa : PCl = 1 : 0.04 : 0.45, using the mammalian concentrations above:
numerator = 5 + (0.04)(145) + (0.45)(10) = 15.3 denominator = 140 + (0.04)(12) + (0.45)(110) = 190.0 V_m = 61.5 · log(15.3 / 190.0) ≈ −67 mV
The insight is that Vm is a permeability-weighted average of the equilibrium potentials. The membrane potential goes wherever the conductance is dominant. At rest, K⁺ "leak" channels (two-pore-domain K2P and inward rectifiers) dominate, so Vm sits near EK. During an action potential, Na⁺ conductance briefly dominates, and Vm swings toward ENa. Every electrical signal in the nervous system is a change in relative permeability. This idea organizes the whole course.
How many ions actually move? Astonishingly few. The membrane is a capacitor with Cm ≈ 1 µF/cm². Consider a spherical cell 25 µm across:
Q = C·V = (1×10⁻⁶ F/cm²)(0.1 V) = 1×10⁻⁷ C/cm² → ≈ 1×10⁻¹² mol of monovalent ions per cm² of membrane Surface area ≈ 2×10⁻⁵ cm² → ≈ 2×10⁻¹⁷ mol of K⁺ separated Internal K⁺ ≈ 1×10⁻¹² mol → fraction moved ≈ 10⁻⁵
So charging the membrane by 100 mV moves about one K⁺ in 50,000. Bulk solutions stay electroneutral. The charge imbalance lives in a nanometer-thin sheet at the membrane surface.
The Na⁺/K⁺-ATPase (3 Na⁺ out, 2 K⁺ in per ATP) builds and maintains the gradients. It is electrogenic, but its direct contribution to Vm is only a few millivolts. Its main role is to charge the batteries, not to set the voltage. The pump consumes a large fraction of the brain's ATP; estimates for neural tissue are often quoted around half. This is why the brain is about 2% of body mass yet uses roughly 20% of resting O₂.
Concept 3: The Action Potential, a Self-Regenerating Wave of Permeability Change
Hodgkin and Huxley used the voltage clamp on the squid giant axon to separate the membrane current into Na⁺ and K⁺ components. They then wrote equations that predicted the action potential's shape and conduction velocity, before anyone knew ion channels were proteins [Hodgkin & Huxley, 1952]. It remains one of the great triumphs of quantitative biology.
C_m · dV/dt = −ḡ_Na·m³h·(V − E_Na) − ḡ_K·n⁴·(V − E_K) − ḡ_L·(V − E_L) + I_ext
Each gating variable x ∈ {m, h, n} obeys first-order kinetics:
dx/dt = α_x(V)·(1 − x) − β_x(V)·x
equivalently dx/dt = (x_∞(V) − x) / τ_x(V)
Read this through your differential equations background (MATH102):
- m (Na⁺ activation) is fast (τ ~0.1–0.5 ms) and increases with depolarization.
- h (Na⁺ inactivation) is slower and decreases with depolarization.
- n (K⁺ activation) is slower still and increases with depolarization.
The exponents (m³h, n⁴) were empirical fits. In hindsight they beautifully anticipate tetrameric channels with multiple independent voltage sensors. Voltage-gated K⁺ channels are true tetramers. Voltage-gated Na⁺ channels are one polypeptide with four homologous domains, each carrying an S4 voltage-sensing helix studded with arginines or lysines every third residue.
The cycle as a story of positive and negative feedback.
- A small depolarization opens some Na⁺ channels. Na⁺ influx depolarizes further, which opens more Na⁺ channels. This positive feedback is the explosive upstroke. Threshold is the voltage where inward Na⁺ current first exceeds outward K⁺ and leak current.
- With a delay, two negative feedbacks kick in. Na⁺ channels inactivate (h → 0), the "ball-and-chain" or, for NaV, the IFM-motif hinged lid. K⁺ channels open (n → 1). Vm repolarizes and transiently undershoots toward EK (after-hyperpolarization).
- While h recovers, the membrane cannot fire again (absolute refractory period), then fires only with difficulty (relative refractory period). Refractoriness makes propagation unidirectional.
Propagation and cable theory. A passive axon is a leaky cable:
Length constant: λ = √( r_m / r_i ) (r_m = membrane resistance × unit length,
r_i = axial resistance ÷ unit length)
Time constant: τ = r_m · c_m
Passive decay: V(x) = V₀ · e^(−x/λ)
Conduction speed depends on how far ahead current spreads (λ) and how fast it charges the next patch (τ). Animals use two strategies to go faster:
- Make the axon fatter. ri falls with cross-sectional area, so for unmyelinated axons velocity scales roughly as √diameter. This is the squid's strategy: a ~0.5 mm axon conducting about 25 m/s.
- Myelinate. Myelin raises rm and lowers cm by stacking membrane layers. Current jumps between nodes of Ranvier, which are dense with NaV channels (saltatory conduction). Velocity then scales roughly linearly with diameter; an empirical rule of thumb is about 6 m/s per µm of fiber diameter. A 10 µm myelinated axon outruns a squid axon fifty times thicker.
Myelin is why vertebrates can pack billions of fast axons into a skull. Demyelinating diseases such as multiple sclerosis (autoimmune, BIOL401 again) cause conduction slowing and block.
Hodgkin and Huxley did not know channels existed, yet their kinetic parameters described them so well that patch-clamp recordings of single channels [Neher & Sakmann, 1976] fit naturally into their framework decades later. A good phenomenological model can contain a correct mechanistic prediction.
Concept 4: Synaptic Transmission, Calcium-Triggered Membrane Fusion with Quantal Packets
At a chemical synapse, an action potential arriving at the terminal opens voltage-gated Ca²⁺ channels (CaV2.1/P/Q and CaV2.2/N types) clustered at the active zone. Recall that ECa ≈ +130 mV and resting [Ca²⁺]i is about 100 nM. Opening a few channels therefore creates nanodomains of 10–100 µM Ca²⁺ within tens of nanometers of the channel mouth. Docked vesicles sit inside these domains, which is why release can occur less than a millisecond after Ca²⁺ entry.
The SNARE machine. Membrane fusion is energetically uphill: two bilayers repel because of hydration forces and electrostatics. SNARE proteins supply the energy by zippering into one of the most stable protein complexes known [Söllner et al., 1993]:
- Synaptobrevin-2 (VAMP2), a vesicle membrane protein (R-SNARE), contributes one α-helix.
- Syntaxin-1, a plasma membrane protein (Q-SNARE), contributes one helix.
- SNAP-25, palmitoylated and anchored to the plasma membrane, contributes two helices.
The four helices zip from their N-termini toward the membranes, forming a parallel four-helix bundle. Zippering pulls the bilayers together like a winch. The supporting cast:
- Munc18 and Munc13 template correct SNARE assembly.
- Complexin clamps partially zippered complexes.
- Synaptotagmin-1 is the Ca²⁺ sensor. Its C2A and C2B domains bind Ca²⁺ and penetrate the anionic membrane, releasing the clamp and triggering fusion.
- NSF, an AAA+ ATPase, with α-SNAP uses ATP to pry apart spent cis-SNARE complexes for recycling.
Release depends on roughly the fourth power of extracellular Ca²⁺ [Dodge & Rahamimoff, 1967]. This steep, cooperative relationship is consistent with multiple Ca²⁺ ions binding the sensor, like hemoglobin's sigmoidal O₂ binding. Small changes in Ca²⁺ entry produce large changes in release, which is a major lever for presynaptic modulation.
Nature's own SNARE biochemistry. Tetanus and botulinum neurotoxins are zinc metalloproteases whose light chains each cleave a specific SNARE:
- Tetanus toxin and BoNT/B, /D, /F, /G cleave synaptobrevin.
- BoNT/A and /E cleave SNAP-25.
- BoNT/C cleaves syntaxin and SNAP-25.
Cleaving a single peptide bond in a SNARE abolishes neurotransmission. Botulinum toxin's flaccid paralysis and tetanus's spastic paralysis differ only because tetanus toxin is transported retrogradely to spinal inhibitory interneurons and silences them.
Quantal release. Fatt and Katz observed spontaneous miniature end-plate potentials (mEPPs) of about 0.5 mV at the frog neuromuscular junction. Del Castillo and Katz then showed that evoked responses are integer multiples of this unit [del Castillo & Katz, 1954]. The unit, a quantum, corresponds to one vesicle's contents (several thousand ACh molecules). Release is probabilistic, a direct application of MATH201:
Binomial model: n = number of release sites
p = release probability per site
q = quantal size (response to one vesicle)
Mean quantal content: m = n·p
Mean response: ⟨EPP⟩ = m·q
P(k quanta released) = C(n,k)·p^k·(1−p)^(n−k)
When p is small and n is large → Poisson:
P(k) = e^(−m)·m^k / k!
So the fraction of failures gives m = −ln P(0)
The Poisson failure method is a beautiful trick: count the trials with no response, and you know the average number of vesicles released, without having to resolve individual quanta. Central synapses often have release probabilities of 0.1–0.5, so transmission is fundamentally stochastic. The brain computes with unreliable components.
Concept 5: Receptors, Fast Ion Channels and Slow Enzyme Switches
The receptor, not the transmitter, determines what a synapse does. There are two broad families.
Ionotropic receptors (ligand-gated ion channels) open within a millisecond of binding transmitter.
- Cys-loop pentamers
- Nicotinic acetylcholine receptor (nAChR). Muscle type is (α1)₂β1δε in adults; the fetal form has γ instead of ε. Two ACh sites sit at α–δ and α–ε interfaces. It is a nonselective cation channel, so its reversal potential is about 0 mV, between ENa and EK.
- GABAA receptor, typically 2α:2β:γ2. It is Cl⁻-selective, and the benzodiazepine site sits at the α/γ interface. Benzodiazepines are positive allosteric modulators: they increase the frequency of channel opening but do not open the channel alone. Barbiturates prolong openings and, at high concentration, gate the channel directly.
- Glycine receptor, Cl⁻-selective, dominant in spinal cord and brainstem. It is blocked by strychnine.
- Ionotropic glutamate tetramers
- AMPA receptors (GluA1–4) give fast excitation, with millisecond deactivation and desensitization. RNA editing of GluA2 at the Q/R site replaces a glutamine with an arginine in the pore. One positive charge renders the channel Ca²⁺-impermeable. This links directly to BIOL201 (ADAR editing).
- NMDA receptors (GluN1 + GluN2A–D) need two things to open: glutamate, which binds GluN2, and glycine or D-serine, which binds GluN1 as co-agonist. At resting potential, extracellular Mg²⁺ plugs the pore [Nowak et al., 1984; Mayer et al., 1984]. Depolarization expels the Mg²⁺. NMDA receptors are highly Ca²⁺-permeable and have slow kinetics, with decay over tens to hundreds of milliseconds.
The NMDA receptor is a molecular AND gate. It conducts only when (1) presynaptic glutamate is present and (2) the postsynaptic membrane is already depolarized. That is precisely Hebb's 1949 postulate, "cells that fire together wire together," implemented in a single protein. Hold onto this for Concept 8.
Metabotropic receptors (GPCRs) act through heterotrimeric G proteins over hundreds of milliseconds to minutes. The three canonical pathways you know from CHEM402 reappear:
Gαs (β-adrenergic, D1 dopamine)
→ adenylyl cyclase ↑ → cAMP ↑ → PKA → phosphorylates channels, CREB
Gαi/o (M2 muscarinic, D2, mGluR2/3, GABA_B, μ/δ/κ opioid)
→ adenylyl cyclase ↓
→ released Gβγ opens GIRK K⁺ channels (hyperpolarization)
→ Gβγ inhibits Ca_V2 channels (less transmitter release)
Gαq (M1/M3 muscarinic, mGluR1/5, α1-adrenergic)
→ PLCβ → PIP₂ → IP₃ (Ca²⁺ release from ER) + DAG (PKC)
The vagus nerve slowing the heart is the classic example. ACh acts on M2 receptors, and Gβγ directly opens GIRK channels in pacemaker cells. Otto Loewi's 1921 "Vagusstoff" experiment, which established chemical transmission, was this pathway.
Analogy: an ionotropic receptor is a light switch, direct and immediate. A metabotropic receptor is a thermostat setting. It does not turn anything on by itself but changes how the whole system responds for a while.
Concept 6: Neurotransmitters, Small Molecules and the Logic of Their Chemistry
Classical small-molecule transmitters share a life cycle:
- Local synthesis in the terminal.
- Loading into small clear vesicles by vesicular transporters (VGLUT, VGAT, VAChT, VMAT2) powered by the V-ATPase proton gradient.
- Ca²⁺-triggered release.
- Rapid inactivation by reuptake or enzymatic degradation.
- Glutamate is the main excitatory transmitter. It is made from glutamine by glutaminase and cleared by astrocytic EAATs.
- GABA is the main inhibitory transmitter in the brain. It is made by glutamate decarboxylase (GAD65/67, which requires pyridoxal phosphate). One decarboxylation converts the main excitatory transmitter into the main inhibitory one.
- Glycine is the main inhibitory transmitter in spinal cord and brainstem, and also the NMDA co-agonist.
- Acetylcholine is made by choline acetyltransferase (acetyl-CoA + choline). It is hydrolyzed in the cleft by acetylcholinesterase, an enzyme so fast its rate approaches the diffusion limit, a kcat/KM example from CHEM401. Organophosphate nerve agents irreversibly acylate its active-site serine.
- Catecholamines follow one pathway: tyrosine → (tyrosine hydroxylase, the rate-limiting step) → L-DOPA → (AADC) → dopamine → (dopamine β-hydroxylase, inside vesicles) → norepinephrine → (PNMT) → epinephrine. Which enzymes a neuron expresses determines which transmitter it releases.
- Serotonin (5-HT): tryptophan → (tryptophan hydroxylase) → 5-HTP → (AADC) → 5-HT. It is cleared by SERT, the target of SSRIs.
- Histamine: histidine → (histidine decarboxylase). It comes from tuberomammillary neurons and promotes wakefulness, which is why first-generation antihistamines that cross the BBB cause drowsiness.
Monoamine systems (dopamine, norepinephrine, serotonin, histamine) share a striking anatomy. A few thousand to a few hundred thousand neurons in small brainstem or midbrain nuclei project diffusely across the whole forebrain. That anatomy suits a modulatory role: broadcasting a global state such as arousal, salience, or reward prediction error rather than point-to-point messages.
Concept 7: Neuropeptides, Gene-Encoded Signals Cut from Larger Precursors
Neuropeptides differ from classical transmitters in nearly every respect, and the differences flow from one chemical fact: a peptide is made by the ribosome and cannot be synthesized in the axon terminal.
The biosynthetic pipeline (secretory pathway, BIOL101 and CHEM401):
- Prepropeptide. The mRNA is translated at the rough ER. An N-terminal signal peptide directs co-translational translocation, and signal peptidase removes it, leaving the propeptide (prohormone).
- Folding and transport. Disulfide bonds form in the ER. The propeptide moves through the Golgi and is sorted at the trans-Golgi network into the regulated secretory pathway.
- Endoproteolysis. In maturing secretory granules, which are acidic (pH ~5.5) and Ca²⁺-rich, the prohormone convertases PC1/3 and PC2 act. These are subtilisin-like serine proteases that cleave on the C-terminal side of paired basic residues (Lys-Arg, Arg-Arg, sometimes Lys-Lys or a single Arg).
- Exoproteolysis. Carboxypeptidase E trims the remaining C-terminal Lys/Arg residues.
- Post-translational modification. The most important is C-terminal α-amidation by peptidylglycine α-amidating monooxygenase (PAM). PAM converts a C-terminal Gly into an amide on the preceding residue, using Cu²⁺/Cu⁺, ascorbate, and O₂. Many neuropeptides, including substance P, NPY, oxytocin, and vasopressin, are inactive without it. Others are acetylated, sulfated, or pyroglutamylated.
- Storage and release. Peptides are stored in large dense-core vesicles (LDCVs), about 80–200 nm across compared with about 40 nm for small clear vesicles. LDCVs sit away from active zones and need larger, more global Ca²⁺ rises, so they release preferentially during high-frequency or burst firing. Release can occur from dendrites and somata, not only terminals. Oxytocin and vasopressin are released dendritically in the hypothalamus.
- Inactivation. There is no reuptake. Peptides are removed by diffusion and extracellular peptidases such as neprilysin (once called "enkephalinase") and aminopeptidase N. Lifetimes are longer and actions spread farther: volume transmission rather than wiring transmission.
The steps are summarized here:
Gene → mRNA → [signal]–[pro-region]–KR–[peptide A]–KR–[peptide B]–G–KR ...
signal peptidase (ER)
→ propeptide
PC1/3 or PC2 (cleave after K-R)
→ peptide-K-R fragments
carboxypeptidase E (removes C-terminal K/R)
→ peptide-Gly
PAM (Gly → C-terminal amide)
→ peptide-NH₂ (active)
The opioid families: three genes, many peptides. The 1970s brought two discoveries in quick succession. First, stereospecific opiate binding sites were found in brain [Pert & Snyder, 1973]. Then endogenous ligands were isolated: the pentapeptides Met-enkephalin (Tyr-Gly-Gly-Phe-Met) and Leu-enkephalin (Tyr-Gly-Gly-Phe-Leu) [Hughes et al., 1975]. Cloning then revealed three precursor genes. Akil and colleagues synthesized these discoveries in their influential review [Akil et al., 1984]:
- Pro-opiomelanocortin (POMC), first cloned by [Nakanishi et al., 1979], yields β-endorphin (31 residues, beginning with the Met-enkephalin sequence), ACTH, α-, β-, and γ-MSH, and β-lipotropin.
- Proenkephalin yields four copies of Met-enkephalin, one Leu-enkephalin, and the extended peptides Met-enk-Arg-Phe and Met-enk-Arg-Gly-Leu. One precursor carries seven opioid sequences.
- Prodynorphin yields dynorphin A, dynorphin B, and α/β-neoendorphin, all beginning with the Leu-enkephalin sequence.
The shared N-terminal motif Tyr-Gly-Gly-Phe is the "message" that activates opioid receptors. The C-terminal extensions are the "address" that biases which receptor responds: μ (enkephalins, β-endorphin), δ (enkephalins), or κ (dynorphins). The Tyr¹ phenol and its free amine mimic the phenol and tertiary amine of morphine. Opioid receptors were cloned in 1992–93: δ first, then μ and κ. All three are Gi/o-coupled GPCRs.
Tissue-specific processing is the beautiful part. The same POMC gene gives different products in different cells, depending on which convertases they express:
- Anterior pituitary corticotrophs express mainly PC1/3 and produce mainly ACTH and β-lipotropin, which drives the adrenal stress response.
- The intermediate pituitary and hypothalamic arcuate POMC neurons also express PC2. They cleave further, to α-MSH (which suppresses appetite via MC4R) and β-endorphin.
This is information encoded in protease expression patterns, a second layer of "genetic code" applied after translation.
Other major peptide systems (sequences worth recognizing):
- Substance P (Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met-NH₂, 11 residues) is a tachykinin acting at NK1 receptors. It is co-released with glutamate from nociceptive C-fibers in the dorsal horn, contributing to pain transmission and neurogenic inflammation.
- Neuropeptide Y (36 residues, C-terminal Tyr-NH₂) is co-expressed with AgRP in arcuate neurons. It is a powerful appetite stimulant, the functional opponent of POMC/α-MSH neurons.
- Oxytocin (Cys-Tyr-Ile-Gln-Asn-Cys-Pro-Leu-Gly-NH₂) and vasopressin (Cys-Tyr-Phe-Gln-Asn-Cys-Pro-Arg-Gly-NH₂) are nonapeptides with a Cys¹–Cys⁶ disulfide ring. They differ at only two positions (3 and 8), yet one drives uterine contraction, milk ejection, and pair-bonding, while the other drives water retention (V2), vasoconstriction (V1a), and male social behavior.
- Somatostatin (SST-14, cyclic via a disulfide) inhibits growth hormone release and marks a major class of cortical inhibitory interneurons.
- Corticotropin-releasing factor (CRF) (41 residues) is released from paraventricular nucleus neurons and initiates the hypothalamic-pituitary-adrenal (HPA) axis. Extrahypothalamic CRF in the amygdala mediates anxiety-like behavior.
Co-transmission. Most peptidergic neurons also release a classical transmitter. Low-frequency firing releases mainly the small molecule; bursts add the peptide. The neuron thus encodes firing pattern into chemical identity of its output. That is a frequency-to-chemistry code, and a quiet refinement of the older idea that a neuron uses one transmitter everywhere (Dale's principle as named by Eccles).
Concept 8: Synaptic Plasticity, the Molecular Basis of Learning
Bliss and Lømo stimulated the perforant path input to the dentate gyrus of anesthetized rabbits at high frequency. The synaptic response stayed enhanced for hours [Bliss & Lømo, 1973]. This is long-term potentiation (LTP). At CA3→CA1 Schaffer collateral synapses, the best-studied form is NMDA receptor-dependent. It has three properties:
- Cooperativity. Enough inputs must be active together to depolarize the cell.
- Associativity. A weak input paired with a strong one gets potentiated.
- Input specificity. Only active synapses change.
All three follow from the NMDA receptor's AND-gate logic.
The mechanism, step by step:
- Glutamate binds both AMPA and NMDA receptors. AMPA receptors depolarize the spine. If depolarization is strong enough, from summed inputs or a back-propagating spike, Mg²⁺ is expelled from NMDA receptors.
- Ca²⁺ floods the spine. The spine's narrow neck compartmentalizes this Ca²⁺, which explains input specificity.
- Ca²⁺/calmodulin activates CaMKII, a dodecameric holoenzyme of two stacked hexameric rings. Neighboring subunits phosphorylate each other at Thr286. Once phosphorylated, CaMKII stays active even after Ca²⁺ falls. It is a molecular memory switch [Lisman, Schulman & Cline, 2002], and its binding to GluN2B helps anchor it at the synapse.
- CaMKII phosphorylates GluA1 at Ser831 (increasing conductance) and promotes exocytosis and lateral diffusion of AMPA receptors into the postsynaptic density, where scaffolds such as PSD-95 capture them via TARPs (e.g., stargazin). More AMPA receptors means a bigger response. Silent synapses, which have NMDA receptors but no AMPA receptors, can be "unsilenced" this way.
- Late LTP (hours to days) requires PKA, ERK, CREB-dependent transcription, local dendritic translation, and structural spine enlargement. This is Kandel's synthesis from Aplysia to mammals [Kandel et al., 2021].
LTD and the calcium code. Prolonged low-frequency stimulation (~1 Hz for 15 min) gives a modest, slow Ca²⁺ rise. That preferentially activates high-affinity phosphatases (calcineurin/PP2B → PP1), which dephosphorylate GluA1 Ser845 and trigger clathrin-mediated AMPA receptor endocytosis. The same ion, Ca²⁺, signals opposite outcomes depending on amplitude and time course:
Small, prolonged Ca²⁺ rise → phosphatases dominate → LTD (AMPAR removal) Large, brief Ca²⁺ rise → kinases dominate → LTP (AMPAR insertion)
Metaplasticity, the plasticity of plasticity [Abraham & Bear, 1996], means the LTP/LTD crossover point (θm) slides with a neuron's activity history. This is the core idea of the Bienenstock-Cooper-Munro (BCM) model [Bienenstock, Cooper & Munro, 1982]. A highly active neuron raises its threshold, making further LTP harder, a homeostatic brake that prevents runaway excitation. One mechanism is a shift in NMDA receptor subunit composition (GluN2B → GluN2A), which changes Ca²⁺ kinetics.
Concept 9: Wiring the Brain, Molecular Guidance and Activity-Dependent Refinement
How do ~86 billion neurons find their partners? Sperry's chemoaffinity hypothesis [Sperry, 1963] proposed molecular address tags. The growth cone, an actin-driven sensory structure at the axon tip, reads four conserved families of guidance cues:
- Netrins are secreted, and their effect depends on the receptor. With DCC alone they attract; with UNC5 present they repel. Netrin-1 from the floor plate attracts commissural axons to the spinal midline.
- Slits are secreted and act on Robo receptors to repel. Midline Slit keeps axons from re-crossing once they have crossed, because crossed axons up-regulate Robo.
- Semaphorins are secreted or membrane-bound and signal through plexins and neuropilins, mostly as repellents that collapse growth cones.
- Ephrins are membrane-bound and act on Eph receptor tyrosine kinases with bidirectional signaling. Graded EphA/ephrin-A expression maps the retina topographically onto the tectum or superior colliculus.
The key lesson: the receptor repertoire and internal state of the growth cone determine the response. The same cue can attract or repel. Cytoplasmic cAMP/cGMP ratios can flip netrin attraction to repulsion.
Synaptogenesis and refinement. Initial wiring uses molecules. Final wiring uses activity: neurons that fire together keep their connections, and the others are pruned, partly by microglial complement-tagged phagocytosis. This is Hebbian logic acting during development and implemented by the NMDA receptor again. Critical periods, such as ocular dominance plasticity in visual cortex, are windows when this refinement is strongest. They close partly as GABAergic inhibition matures.
Concept 10: Sensory Transduction, Converting the World into Membrane Potential
Every sensory system faces the same engineering problem: convert a physical stimulus into a change in ion channel opening. The solutions trade speed against amplification.
Vision: maximum amplification.
- A photon isomerizes 11-cis-retinal, bound to opsin via a protonated Schiff base on Lys296, to all-trans in about 200 femtoseconds. This is one of the fastest photochemical reactions in biology (CHEM303 photophysics).
- Metarhodopsin II activates hundreds of molecules of transducin (Gt).
- Each transducin α activates PDE6, which hydrolyzes cGMP.
- Falling cGMP closes cGMP-gated cation (CNG) channels that were carrying the inward "dark current."
- The rod hyperpolarizes and reduces glutamate release. Light is signaled by a decrease in transmitter.
Rods detect single photons [Baylor, Lamb & Yau, 1979]. The signal is shut off by rhodopsin kinase and arrestin, and Ca²⁺ feedback through guanylate cyclase-activating proteins mediates light adaptation.
Hearing: maximum speed. Hair cell stereocilia are arranged in rows of increasing height and connected by tip links made of cadherin-23 and protocadherin-15. Deflection toward the tallest row tensions the tip links and directly gates mechanotransduction channels (TMC1/2 are pore-forming components) within microseconds. There is no second messenger, because a cascade would be too slow to follow 20 kHz sounds. The channels admit K⁺, which depolarizes the cell. This works because the endolymph bathing the stereocilia is K⁺-rich (~150 mM), and the endocochlear potential of about +80 mV boosts the driving force. A gradient you would expect to hyperpolarize instead depolarizes, because the apical compartment's chemistry is engineered.
Olfaction: maximum diversity. Buck and Axel discovered a huge GPCR family of odorant receptors [Buck & Axel, 1991]: about 1,000 genes in mouse and about 400 functional genes in humans, among the largest gene families in mammalian genomes. Each olfactory sensory neuron expresses one receptor allele (monoallelic choice). All neurons expressing the same receptor converge onto one or a few glomeruli in the olfactory bulb. The transduction cascade:
odorant → OR → Gα_olf → adenylyl cyclase III → cAMP ↑ → CNG channel opens (Na⁺, Ca²⁺ in)
→ Ca²⁺ opens Ca²⁺-activated Cl⁻ channels (ANO2/TMEM16B)
→ Cl⁻ EFFLUX (olfactory neurons keep internal Cl⁻ high) → amplified depolarization
Odor identity is a combinatorial code. Each odorant activates a pattern of receptors, and each receptor responds to many odorants.
Concept 11: From Circuits to Behavior and Disease (Brief Map)
- Motor control is hierarchical. Spinal central pattern generators and reflexes are coordinated by motor cortex. The basal ganglia select actions by disinhibition: the direct pathway via D1 receptors facilitates action, and the indirect pathway via D2 receptors suppresses it. The cerebellum compares intended and actual movement and corrects errors, with climbing-fiber-driven LTD at parallel-fiber–Purkinje synapses as a proposed teaching signal.
- The hippocampus is needed to form new declarative memories. Patient H.M.'s bilateral medial temporal lobe resection left him unable to form them while sparing motor learning. Place cells provide a cognitive map.
- Mesolimbic dopamine neurons, projecting from the ventral tegmental area to the nucleus accumbens, fire to unexpected rewards. After learning, they shift to firing at the predicting cue and dip when an expected reward is omitted. They encode a reward prediction error [Schultz, Dayan & Montague, 1997], the same quantity as the temporal-difference learning signal in reinforcement learning.
- Disease is molecular neuroscience gone wrong:
- Alzheimer's disease involves amyloid-β (Aβ40/42, cleaved from APP by β-secretase/BACE1 and γ-secretase) aggregating into plaques, and hyperphosphorylated tau forming neurofibrillary tangles.
- Parkinson's disease is loss of substantia nigra dopamine neurons with α-synuclein Lewy bodies. L-DOPA crosses the BBB via LAT1; dopamine itself cannot.
- Depression and schizophrenia implicate monoamine, glutamate, and dopamine signaling.
- Pain pharmacology centers on opioids acting at μ receptors in the periaqueductal gray, rostral ventromedial medulla, and dorsal horn.
Connections to Peptide Chemistry
This course is where your organic and biochemical training on the amide bond becomes neurobiology. Several threads lead directly into peptide chemistry research.
- Precursor processing is regioselective proteolysis. Prohormone convertases recognize dibasic motifs much as trypsin recognizes single Lys/Arg, but with extended subsite specificity (P1, P2, P4, P6). Predicting which sites are cut, and in which cell, is an active peptidomics problem. It is addressed with the tandem mass spectrometry you learned in CHEM501: LC-MS/MS of brain extracts, often with in-situ inactivation to stop post-mortem proteolysis. This is how many "orphan" neuropeptides have been discovered.
- C-terminal amidation is a design feature. Removing the C-terminal carboxylate eliminates a negative charge and blocks carboxypeptidases. When you synthesize neuropeptides by solid-phase peptide synthesis [Merrifield, 1963], you choose a Rink amide resin rather than a Wang resin specifically to obtain C-terminal amides.
- Natural peptides make poor drugs; medicinal chemistry fixes them. Enkephalins have plasma half-lives of minutes and do not cross the BBB. Classic remedies, all applying CHEM201/202 stereochemistry and conformational analysis:
- D-amino acid substitution blocks aminopeptidase cleavage of the Tyr¹–Gly² bond. DAMGO, [D-Ala², N-MePhe⁴, Gly-ol⁵]-enkephalin, is the μ-selective reference agonist.
- Cyclization through disulfide bridges restricts conformation and improves selectivity. DPDPE, [D-Pen², D-Pen⁵]-enkephalin, is δ-selective.
- N-methylation, lipidation, and stapling improve stability and permeability. Semaglutide, a lipidated GLP-1 analog, acts partly on hypothalamic and brainstem appetite circuits.
- Venom peptides are evolved neuropharmacology. Cone snail conotoxins are small, highly disulfide-bonded peptides targeting specific ion channels. Ziconotide (synthetic ω-conotoxin MVIIA) blocks N-type CaV2.2 channels at dorsal horn synapses. It is given intrathecally, bypassing the BBB, for severe chronic pain. Getting the correct disulfide connectivity during oxidative folding is a classic peptide chemistry challenge.
- Message–address design. The opioid Tyr-Gly-Gly-Phe "message" plus a selectivity "address" became a general framework for designing receptor-subtype-selective peptide ligands. It is rooted in the precursor-encoded diversity reviewed by [Akil et al., 1984].
- Peptide aggregation as disease. Aβ42 is a 42-residue peptide whose two extra hydrophobic C-terminal residues (Ile-Ala) make it far more aggregation-prone than Aβ40. Understanding β-sheet amyloid structure through cryo-EM and solid-state NMR, and designing aggregation inhibitors, is peptide physical chemistry applied to neurodegeneration.
- Toxins as proteases. Botulinum toxins are zinc endopeptidases with exquisite SNARE substrate specificity. Mapping their cleavage sites is pure protease enzymology, and BoNT/A is used clinically for dystonia, spasticity, and migraine.
Where this leads: neuropeptide GPCRs are one of the richest drug-target classes, and the frontier combines structural biology (cryo-EM of peptide–GPCR complexes), biased agonism (G protein vs. β-arrestin signaling), and peptide engineering for brain delivery.
Common Misconceptions to Watch Out For
- "The Na⁺/K⁺ pump generates the resting potential." The pump maintains the gradients. The voltage itself comes from K⁺ leaking down its gradient through K⁺-selective channels. Block the pump and Vm changes only a few mV at first, then decays slowly over minutes as gradients run down.
- "During an action potential, Na⁺ floods the cell and concentrations change a lot." As calculated above, only about 10⁻⁵ of the ions move. A single spike barely changes bulk concentrations, though thin axons or repetitive firing can accumulate small changes.
- "The action potential peaks at ENa." It approaches but does not reach ENa, because K⁺ and leak conductances never fall to zero and Na⁺ channels inactivate.
- "The action potential flows down the axon like current in a wire." Electrons in a wire would decay over a few λ. The action potential is regenerated at each patch or node, more like a line of dominoes or a burning fuse than a current pulse.
- "Glutamate is excitatory; GABA is inhibitory." The receptor and ion gradients decide. GABAA receptors pass Cl⁻, and their effect depends on ECl. In immature neurons, high NKCC1 and low KCC2 expression keep internal Cl⁻ high, so GABA depolarizes [Ben-Ari, 2002]. Glutamate acting on mGluR2/3 is inhibitory.
- "One gene, one peptide." POMC alone yields ACTH, MSHs, and β-endorphin, and the product mix varies by cell type. Proenkephalin encodes seven opioid sequences.
- "Neuropeptides are just slow neurotransmitters." They differ in synthesis (ribosomal, in the soma), storage (LDCVs), release triggers (bursts, extrasynaptic sites), inactivation (peptidases, no reuptake), and range (volume transmission). They are a different signaling mode.
- "Morphine and endorphins are the same thing." Endogenous opioids are peptides with different receptor preferences. Morphine is an alkaloid that is relatively μ-selective. Each receptor type has distinct effects: κ agonists cause dysphoria, μ agonists cause euphoria and respiratory depression.
- "Dopamine is the pleasure chemical." Dopamine encodes prediction error and incentive salience, the "wanting" of a reward, more than hedonic "liking," which involves opioid and endocannabinoid hotspots [Berridge & Robinson, 1998].
- "Depression is a serotonin deficiency." SSRIs block SERT within hours, yet mood improves over weeks. That lag implicates slower adaptations such as receptor regulation, plasticity, and neurogenesis-related changes. Direct evidence for a simple serotonin deficit is weak.
- "LTP is memory." LTP is a cellular mechanism that is necessary for some forms of memory and shares many properties with them. Equating the two skips the circuit and systems levels.
- "Glia are passive support." Astrocytes regulate synapses and K⁺, microglia prune circuits, and oligodendrocytes tune conduction timing. They are active participants.
- "Threshold is a fixed voltage." Threshold depends on recent history (Na⁺ channel inactivation, K⁺ channel state) and on the rate of depolarization. Slow depolarization can cause accommodation without a spike.
Key References Cited
- Kandel ER, Koester JD, Mack SH, Siegelbaum SA (2021). Principles of Neural Science, 6th ed. McGraw-Hill.
- Bear MF, Connors BW, Paradiso MA (2016). Neuroscience: Exploring the Brain, 4th ed. Wolters Kluwer.
- Akil H, Watson SJ, Young E, Lewis ME, Khachaturian H, Walker JM (1984). Endogenous opioids: biology and function. Annu Rev Neurosci 7:223–255. DOI: 10.1146/annurev.ne.07.030184.002143.
Note: your syllabus lists this DOI with the title "Opioid peptides: molecular pharmacology, biosynthesis, and processing" and no author. To my knowledge, that title belongs to a different 1984 Annual Review of Biochemistry article by Hollt. Check the DOI against the course materials before citing it formally. - Hodgkin AL, Huxley AF (1952). J Physiol 117:500–544.
- Hodgkin AL, Katz B (1949). J Physiol 108:37–77.
- Goldman DE (1943). J Gen Physiol 27:37–60.
- del Castillo J, Katz B (1954). J Physiol 124:560–573.
- Dodge FA, Rahamimoff R (1967). J Physiol 193:419–432.
- Söllner T, et al. (1993). Nature 362:318–324.
- Neher E, Sakmann B (1976). Nature 260:799–802.
- Nowak L, et al. (1984) Nature 307:462–465; Mayer ML, Westbrook GL, Guthrie PB (1984) Nature 309:261–263.
- Bliss TVP, Lømo T (1973). J Physiol 232:331–356.
- Bienenstock EL, Cooper LN, Munro PW (1982). J Neurosci 2:32–48.
- Abraham WC, Bear MF (1996). Trends Neurosci 19:126–130.
- Lisman J, Schulman H, Cline H (2002). Nat Rev Neurosci 3:175–190.
- Pert CB, Snyder SH (1973). Science 179:1011–1014.
- Hughes J, et al. (1975). Nature 258:577–579.
- Nakanishi S, et al. (1979). Nature 278:423–427.
- Sperry RW (1963). PNAS 50:703–710.
- Baylor DA, Lamb TD, Yau KW (1979). J Physiol 288:613–634.
- Buck L, Axel R (1991). Cell 65:175–187.
- Schultz W, Dayan P, Montague PR (1997). Science 275:1593–1599.
- Berridge KC, Robinson TE (1998). Brain Res Rev 28:309–369.
- Ben-Ari Y (2002). Nat Rev Neurosci 3:728–739.
- Merrifield RB (1963). J Am Chem Soc 85:2149–2154.
⚙️ Mechanisms & Applications
Mechanisms & Applications: Connecting Molecules to Neural Computation
My organizing principle: a nervous system does not escape chemistry—it arranges chemistry so that information can move, persist, and alter future behavior. Ion gradients store free energy; channels convert that energy into electrical signals; synapses transform electrical activity into chemical communication; and plasticity changes how later signals are interpreted. The difficult step is connecting these levels without pretending that a molecule, by itself, explains a memory or a psychiatric disorder.
For Skipper, the most productive connection to previous courses is this: neuroscience combines nonequilibrium thermodynamics, reaction kinetics, molecular recognition, and feedback control. A neuron is simultaneously an electrochemical cell, a spatially extended electrical circuit, and a regulated biochemical reactor. [1,2]
1. Establishing the Electrical Starting Point: Gradients, Selectivity, and Glial Support
A concentration gradient alone does not determine ion movement. An ion responds to its electrochemical potential: both concentration and voltage matter. The Nernst equation is therefore not an isolated neuroscience formula; it is the equilibrium condition from CHEM301 applied to one membrane-permeant ion.
- Transporters establish unequal concentrations. The Na+/K+-ATPase moves three Na+ outward and two K+ inward per ATP hydrolyzed. It maintains the gradients used by channels and many secondary transporters.
- Selective channels allow downhill movement. At rest, many neurons have substantial K+ permeability. K+ initially tends to leave down its concentration gradient.
- Charge separation generates an opposing voltage. Only a tiny fraction of cellular ions must redistribute to charge the membrane. Bulk cytoplasm and extracellular fluid remain approximately electroneutral.
- Multiple ionic currents balance at the resting potential. Rest is usually a nonequilibrium steady state, not equilibrium for every ion. ATP consumption is necessary to maintain it.
Electrochemical potential difference, inside minus outside: Δμ̃ = RT ln([ion]in/[ion]out) + zFVm At equilibrium for that ion, Δμ̃ = 0: Eion = (RT/zF) ln([ion]out/[ion]in) At 37 °C: Eion ≈ (61.5 mV/z) log10([ion]out/[ion]in) Example: [K+]out = 4 mM; [K+]in = 140 mM EK ≈ 61.5 log10(4/140) = −95 mV
Here, Vm is inside relative to outside. The example does not predict that every neuron rests at −95 mV: Na+, Cl−, other conductances, and electrogenic transport also matter. For monovalent ions under the constant-field approximation, the Goldman–Hodgkin–Katz voltage equation combines permeabilities:
Vm = (RT/F) ln {
(PK[K+]out + PNa[Na+]out + PCl[Cl−]in)
/
(PK[K+]in + PNa[Na+]in + PCl[Cl−]out)
}
The chloride concentrations appear in the opposite positions
because chloride carries negative charge.
Important distinction: permeability and conductance are related descriptions, not interchangeable quantities. For an approximately ohmic channel population:
Iion = gion(Vm − Eion) gion: conductance Vm − Eion: electrical driving force expressed as a voltage Convention here: positive current is outward positive-charge flow.
Opening a channel does not specify the direction of its effect until I know its reversal potential and the current membrane voltage. At Vm = −65 mV and ENa = +60 mV, the Na+ driving force is −125 mV, producing inward current. At the same voltage with EK = −95 mV, the K+ driving force is +30 mV, producing outward current. [1,2]
Why glia belong in this mechanism: astrocytes help regulate extracellular K+, remove transmitters, and support metabolism. Oligodendrocytes and Schwann cells organize myelin in the CNS and PNS, respectively. Microglia provide immune surveillance and participate in circuit remodeling. These are not merely support roles: changing extracellular K+ changes EK, and therefore neuronal excitability.
The blood–brain barrier is primarily a property of specialized vascular endothelial cells, including tight junctions and restricted transcytosis, supported by pericytes and astrocytic signaling. It also expresses selective transporters and efflux systems. This distinction matters in drug development: astrocytes help maintain the barrier, but do not themselves form its principal vascular seal. A peptide can be exceptionally potent at a neuronal receptor yet clinically ineffective if it never reaches that receptor. [1,2]
2. Generating and Propagating an Action Potential
The central surprise is that an action potential is not a packet of Na+ traveling down an axon. It is a regenerating pattern of local membrane currents. The analogy is a line of falling dominoes: the disturbance advances much farther than any individual component moves.
- A depolarization activates some voltage-gated Na+ channels. Inward current further depolarizes the membrane.
- Positive feedback produces the upstroke. More depolarization recruits more Na+ conductance. Threshold is a dynamical tipping point, not a universal fixed voltage.
- Na+ channels inactivate while delayed K+ conductance increases. The inward current declines and outward current becomes dominant.
- The membrane repolarizes and may afterhyperpolarize. K+ conductance can remain elevated after the spike.
- Channels recover. Na+ channel inactivation contributes to the absolute refractory period; incomplete recovery and persistent K+ conductance contribute to relative refractoriness.
The Na+/K+ pump maintains gradients over time; it is not the main mechanism that repolarizes each millisecond-scale spike.
Hodgkin–Huxley membrane equation, per unit membrane area: Cm dVm/dt = Iapp − INa − IK − IL INa = ḡNa m³h(Vm − ENa) IK = ḡK n⁴(Vm − EK) IL = gL(Vm − EL) For a gating variable x: dx/dt = αx(Vm)(1 − x) − βx(Vm)x Equivalently: dx/dt = [x∞(Vm) − x]/τx(Vm)
Here, m represents Na+ activation, h Na+ inactivation, and n K+ activation in the original model. The gating variables are dimensionless phenomenological variables. The exponents were successful kinetic descriptions, not direct structural measurements of channel subunits. In 1952, Alan Hodgkin and Andrew Huxley used squid giant-axon voltage-clamp data to show that these interacting conductances could quantitatively generate an action potential. This mattered because a biological signal became a testable dynamical model rather than a descriptive electrical event. [3]
Propagation connects to cable physics. Local current depolarizes adjacent membrane. In a passive cable approximation:
Membrane time constant: τm = RmCm Space constant: λ = √(rm/ri) Rm and Cm: specific membrane resistance and capacitance rm: membrane resistance parameter per unit cable length ri: axial resistance per unit length For similar unmyelinated axons: λ ∝ √diameter
Larger axons have lower axial resistance. Myelin increases effective membrane resistance and reduces effective capacitance along internodes, allowing local current to spread farther while charging less membrane. Spikes are regenerated at nodes of Ranvier. Thus “saltatory” conduction is continuous passive current spread combined with spatially separated active regeneration—not electricity jumping across empty space. [1,2]
- Local anesthetics: drugs such as lidocaine inhibit voltage-gated Na+ channels, often showing state- and use-dependent block. Repetitively active fibers can accumulate greater block.
- Channelopathies: altered channel activation, inactivation, or expression can produce epilepsy, pain syndromes, or motor dysfunction. “More channel activity” is not automatically “more firing”; the affected cell type and gating process matter.
- Demyelinating disease: current can leak or fail to charge downstream membrane sufficiently, producing slowed conduction or conduction block.
3. Converting a Spike into Chemical Release
A presynaptic terminal solves a difficult timing problem: how can a voltage change trigger membrane fusion within milliseconds? Its solution is spatial organization. Vesicles, fusion proteins, and Ca2+ channels are concentrated at active zones.
- An arriving action potential opens presynaptic voltage-gated Ca2+ channels.
- Ca2+ enters and forms brief, local high-concentration domains near the channels.
- Ca2+ binds release sensors, prominently synaptotagmins at many fast synapses.
- Primed fusion machinery permits a vesicle to fuse with the plasma membrane.
- Transmitter diffuses through the cleft and binds postsynaptic or presynaptic receptors.
- Diffusion, transport, and/or enzymatic degradation terminate transmitter action; vesicle membrane is recovered and recycled.
Core neuronal SNARE complex: Vesicle: synaptobrevin/VAMP Presynaptic plasma membrane: syntaxin + SNAP-25 SNARE assembly brings membranes together. Munc13, Munc18, complexin, synaptotagmin, and other proteins organize priming, regulation, and Ca²⁺-triggered fusion. After fusion: NSF and soluble SNAP proteins use ATP to disassemble SNARE complexes.
Naming trap: SNAP-25 and the soluble SNAP proteins involved in NSF-mediated disassembly are different proteins. The identification of SNARE interactions by Thomas Söllner and colleagues in 1993 helped connect neuronal exocytosis to general membrane-trafficking mechanisms studied in cell biology. [4]
Quantal transmission: release occurs in discrete vesicular units. In a simple model with N independent release sites, identical release probability p, and identical quantal response q:
Mean postsynaptic response = Npq Variance = Np(1 − p)q² Probability of zero release = (1 − p)^N
These equations are useful hypotheses, not universal descriptions. Real synapses have variable vesicle contents, heterogeneous release probabilities, receptor saturation, and recording noise. Still, they suggest an experimental question: did a drug change the number of functional release sites, release probability, or postsynaptic response to one vesicle?
Paul Fatt and Bernard Katz recorded miniature end-plate potentials in 1952, and José del Castillo and Katz developed quantitative evidence for quantal release in 1954. The striking insight was that spontaneous tiny events reveal the elementary units underlying an evoked synaptic response. [5]
- Botulinum toxins: different serotypes cleave particular SNARE proteins and inhibit transmitter release. Controlled local inhibition of acetylcholine release has applications in dystonia, spasticity, and other conditions.
- Tetanus toxin: cleavage of synaptobrevin in inhibitory circuitry produces disinhibition and spasticity. A related molecular lesion can have a very different systems-level outcome because delivery and cell targeting differ.
- Ca2+-channel therapeutics: the peptide drug ziconotide blocks CaV2.2 channels and reduces transmitter release in spinal pain pathways. Its intrathecal administration illustrates the practical importance of CNS delivery barriers.
4. Receptors Determine What a Transmitter Means
A transmitter is not intrinsically “excitatory,” “inhibitory,” or “pleasurable.” Its effects depend on the receptors expressed, the ionic gradients, the target cell, and the circuit.
- AMPA receptors: tetrameric glutamate-gated cation channels mediate much fast excitatory transmission. Ca2+ permeability depends strongly on subunit composition and GluA2 RNA editing.
- NMDA receptors: glutamate-gated channels also require a co-agonist, usually glycine or D-serine. At negative potentials, extracellular Mg2+ blocks the pore; depolarization relieves much of this block. Their relatively slow currents and Ca2+ permeability make them effective links between activity and biochemical change.
- GABAA receptors: pentameric anion channels whose effects depend mainly on Cl− and also HCO3− gradients.
- Nicotinic acetylcholine receptors: pentameric cation channels; the muscle receptor converts motor-neuron release into end-plate depolarization.
The counterintuitive case is inhibition without hyperpolarization. If opening GABAA receptors holds the membrane near its existing voltage while increasing conductance, incoming excitatory current produces a smaller voltage change. This is shunting inhibition.
For a small current perturbation: ΔV ≈ Iinput / gtotal Opening inhibitory channels increases gtotal, so the same input current produces a smaller ΔV. A depolarizing GABA response can still be inhibitory if its reversal potential remains below spike threshold and the conductance increase suppresses excitation.
Developmental or pathological changes in chloride transport, including NKCC1 and KCC2 activity, can change GABA responses. Therefore, the sign of a recorded voltage deflection is not sufficient to establish whether a synapse increases or decreases spike probability. [1,2]
Metabotropic receptors add amplification and state dependence:
Ligand → GPCR → G-protein nucleotide exchange Gs: adenylyl cyclase ↑ → cAMP ↑ → PKA signaling Gi/o: adenylyl cyclase often ↓ Gβγ can activate GIRK K+ channels Gβγ can inhibit presynaptic Ca²⁺ channels Gq/11: PLCβ → PIP2 cleavage → IP3 + DAG IP3 → intracellular Ca²⁺ release DAG → protein kinase C signaling
Examples include β-adrenergic receptors coupling principally to Gs; M1/M3/M5 muscarinic receptors to Gq/11; M2/M4 to Gi/o; and group I versus group II/III metabotropic glutamate receptors preferentially to Gq/11 versus Gi/o. These are useful organizing patterns, not claims that signaling is identical in every cell.
Dopamine, norepinephrine, histamine, and most serotonin signaling uses GPCRs; 5-HT3 is an important ionotropic serotonin-receptor exception. GABA also acts through metabotropic GABAB receptors. Acetylcholine can act through either receptor class. This explains how one chemical can operate on multiple timescales.
Biochemical termination matters pharmacologically. Acetylcholinesterase hydrolyzes acetylcholine; glutamate transporters prevent excessive extracellular glutamate and support the glutamate–glutamine cycle; monoamine transporters are targets of antidepressants and stimulants. Benzodiazepines positively modulate susceptible GABAA receptors rather than simply replacing GABA. [1,2]
5. Neuropeptides: Gene Expression Becomes a Pharmacological Signal
Peptide neurotransmission is where molecular biology, enzymology, and pharmaceutical chemistry meet most directly. Small-molecule transmitters can often be synthesized and repackaged locally in terminals. Peptides generally require the secretory pathway originating in the cell body.
- Transcription and translation produce a prepropeptide. An N-terminal signal sequence directs the growing chain into the endoplasmic reticulum.
- Signal-peptide removal produces a propeptide. Folding and additional modifications begin in the secretory pathway.
- Sorting into regulated secretory granules occurs. Granules mature and acidify.
- Proprotein convertases cleave selected sites. PC1/3 and PC2 often act near basic residues, but cleavage depends on sequence context, enzyme expression, and granule conditions.
- Exopeptidases trim cleavage products. Carboxypeptidase E commonly removes C-terminal basic residues.
- Additional maturation creates active products. For peptides requiring C-terminal amidation, peptidylglycine α-amidating monooxygenase acts on a glycine-extended precursor; this chemistry requires copper, oxygen, and reducing equivalents supported by ascorbate.
- Dense-core vesicles travel to release sites. Many peptide systems require sustained activity or burst firing to recruit substantial release, although the exact pattern is cell- and compartment-dependent.
- Released peptides act locally or beyond classical synaptic clefts. Extracellular proteases and diffusion terminate signaling; released peptides are not generally recovered intact for vesicular reuse.
A precursor is not a single inevitable product:
Precursor sequence
+ convertase expression
+ compartment conditions
+ trimming/modifying enzymes
= tissue-specific mixture of mature peptides
Examples:
Proenkephalin → several enkephalin-containing products
Prodynorphin → dynorphins and related peptides
POMC → tissue-dependent products including ACTH,
melanocortin peptides, and β-endorphin
Connection to BIOL201 and CHEM402: measuring precursor mRNA is not equivalent to measuring active ligand. Processing is a regulatory layer between gene expression and receptor occupancy. A cell can express a precursor yet produce a different mature peptide mixture from another cell expressing the same gene. The course-listed 1984 review provides historical context for understanding opioid biosynthesis and processing as integral parts of signaling. [6]
Other systems reinforce the same principle. Substance P, acting prominently at NK1 receptors, participates in nociceptive and inflammatory signaling; neuropeptide Y influences feeding, stress, and other functions through several Y receptors; somatostatin regulates secretion and neuronal excitability; CRF links stress-related neural activity to endocrine and behavioral responses. Oxytocin and vasopressin have well-established peripheral physiological actions and complex central effects. Calling oxytocin a “trust hormone” erases receptor distribution, dose, context, and species differences. [1,2]
6. Opioid Mechanisms and the Peptide–Pharma Design Problem
In 1975, John Hughes, Hans Kosterlitz, and colleagues identified Met-enkephalin and Leu-enkephalin, endogenous pentapeptides with opioid activity. Their sequences share Tyr–Gly–Gly–Phe and differ at the fifth residue. Together with early-1970s opioid-binding studies, this transformed opioid drugs from mysterious external agents into probes of an endogenous signaling system. [7]
Enkephalins, β-endorphin, and dynorphins interact with μ, δ, and κ opioid receptors with differing preferences—not perfectly exclusive assignments. These receptors are primarily Gi/o-coupled. [8]
Opioid agonist → μ, δ, or κ receptor → Gi/o signaling Presynaptic: Ca²⁺ channel inhibition → less Ca²⁺ entry → lower release probability Postsynaptic: GIRK activation → K+ conductance increases → reduced excitability in many target neurons Intracellular: adenylyl cyclase inhibition and broader kinase regulation → acute modulation plus longer-term adaptations
In spinal nociceptive pathways, these actions can reduce excitatory transmission. In other circuits, inhibiting an inhibitory neuron can increase downstream activity. This disinhibitory architecture helps explain why a receptor with inhibitory cellular signaling can contribute to activation of reward circuitry.
Clinical effects are circuit-distributed. μ-receptor activation contributes to analgesia but can also impair respiratory rhythm and responsiveness to CO2, slow gastrointestinal motility, and support reinforcement. Naloxone antagonizes opioid receptors and can reverse opioid-induced respiratory depression; recurrence is possible when the opioid outlasts naloxone. Tolerance, physical dependence, and addiction are related but distinct phenomena.
For peptide drug development, I should ask six separate questions:
- Identity: which exact mature sequence, termini, stereochemistry, and modifications are present?
- Affinity and efficacy: how tightly does it bind, and what response does bound receptor generate?
- Selectivity: how does activity compare across μ, δ, κ, related receptors, and other targets?
- Stability: how rapidly do aminopeptidases, endopeptidases, and other processes remove active compound?
- Exposure: what unbound concentration reaches the target tissue, for how long?
- Therapeutic separation: do analgesia and serious adverse effects separate at relevant exposures in vivo?
Simple equilibrium occupancy model: Fraction occupied = [L]free / (Kd + [L]free) If [L]free = Kd, occupancy = 0.5. But: Kd is not necessarily EC50. Occupancy is not necessarily proportional to physiological effect. Total plasma concentration is not the same as free brain concentration.
Possible medicinal-chemistry strategies include cyclization, selected D-amino-acid substitutions, terminal modification, backbone changes, and conjugation. Each can improve one property while worsening another: resistance to proteolysis does not guarantee receptor activity or brain penetration. Manufacturing also requires control of deletion sequences, epimerization, oxidation, aggregation, and residual process impurities.
A caution about biased agonism: different ligands can favor different receptor signaling outputs, but apparent bias depends on assay amplification, receptor abundance, timing, and the reference ligand. Reduced β-arrestin recruitment in one assay is not sufficient evidence that a μ agonist will lack respiratory depression. This is a good example of why a mechanistic hypothesis must survive progressively more physiological tests. [8,13]
7. Making Activity Leave a Trace: Plasticity and Development
Long-term potentiation illustrates how an electrical coincidence can become a biochemical state change. At many hippocampal CA3–CA1 synapses, a useful introductory mechanism is:
- Presynaptic glutamate release activates AMPA receptors.
- Postsynaptic depolarization reduces NMDA-receptor Mg2+ block while glutamate and co-agonist are bound.
- Ca2+ enters through NMDA receptors.
- Ca2+/calmodulin activates signaling proteins, including CaMKII.
- Phosphorylation and changes in AMPA-receptor trafficking and anchoring increase synaptic responsiveness.
- Some persistent forms recruit protein synthesis, transcription, and structural remodeling.
The beautiful idea: the receptor combines a chemical indication of presynaptic activity with an electrical indication of postsynaptic state. It is a molecular coincidence detector—but a probabilistic, voltage-dependent one, not a perfect Boolean AND gate. CaMKII autophosphorylation can extend signaling beyond the initiating Ca2+ transient, although long-term memory cannot be reduced to one permanently active kinase.
Long-term depression can involve Ca2+-dependent phosphatases and AMPA-receptor removal, but LTD is a family of mechanisms. The familiar “large Ca2+ produces LTP, small Ca2+ produces LTD” rule is only a heuristic: spatial localization, time course, developmental stage, and receptor context matter. Metaplasticity means that prior activity changes the conditions under which later LTP or LTD can occur. [1,2]
Terje Lømo’s earlier observations and the landmark 1973 work of Timothy Bliss and Lømo established that brief patterned stimulation could produce persistent enhancement in hippocampal transmission. This supplied a plausible cellular mechanism for experience-dependent change, not proof that every memory is stored by that particular protocol. [9]
Development establishes the circuit on which plasticity acts. Growth-cone receptors detect netrins, Slits, semaphorins, and ephrins; intracellular signaling reorganizes actin, microtubules, and adhesion. A cue can attract or repel depending on receptor composition and intracellular state. Roger Sperry’s chemoaffinity work helped establish that connectivity involves molecular recognition rather than indiscriminate growth followed only by learning. Activity-dependent refinement, trophic competition, and glial remodeling then modify initial connectivity. [1,2]
8. Sensory Transduction: Three Different Solutions to the Same Problem
Sensory cells translate environmental energy into changes in membrane current. The striking comparison is that light, sound, and odor do not all begin by opening a cation channel.
Vision: light closes channels.
Photon absorption → 11-cis-retinal isomerizes to all-trans-retinal → rhodopsin activates transducin → PDE6 activation → cGMP concentration falls → cyclic-nucleotide-gated channels close → photoreceptor hyperpolarizes → glutamate release decreases
In darkness, vertebrate rods and cones are relatively depolarized and release glutamate continuously. The light response therefore begins as a reduction in ongoing transmitter release. ON and OFF retinal pathways interpret this change differently through different glutamate receptors. Amplification provides sensitivity; Ca2+-dependent feedback and pigment regeneration help restore responsiveness. [1,2]
Hearing: force gates a conductance. Deflection of stereocilia toward the tallest row increases tension in the tip-link apparatus and favors opening of mechanotransduction channels. The unusual K+-rich, positively polarized endolymph supports inward current into the hair cell. Depolarization opens basal Ca2+ channels and promotes glutamate release at ribbon synapses. Outer hair-cell electromotility amplifies cochlear motion. Damage to hair cells or their synapses can impair hearing even when downstream neurons remain partly functional.
Olfaction: GPCR signaling opens channels.
Odorant → olfactory receptor → Golf → adenylyl cyclase III → cAMP increases → cyclic-nucleotide-gated channels open → Na+ and Ca²⁺ enter → Ca²⁺-activated Cl− channels contribute amplification Because olfactory sensory neurons maintain elevated internal Cl−, Cl− efflux can be depolarizing.
Linda Buck and Richard Axel identified a large odorant-receptor gene family in 1991, connecting molecular receptor diversity to combinatorial sensory coding. An odor is generally represented by a pattern across receptor types, not a single “one receptor, one smell” channel. [12]
Application: cochlear implants bypass much of hair-cell transduction by electrically stimulating auditory nerve elements. They exploit surviving circuit organization, but their limited spectral representation helps explain why restoring speech understanding is easier than recreating normal music perception.
9. From Circuits to Disease: Avoiding the “One Molecule, One Mind State” Error
- Motor control: spinal networks can organize patterned output; cortex contributes planning and skilled control; basal ganglia help select and regulate actions; cerebellar circuits support prediction, coordination, and error-driven learning. The direct/indirect basal-ganglia pathway model is a useful starting point, not a complete description of movement.
- Parkinson’s disease: loss of substantia nigra dopaminergic neurons changes striatal and wider network function. Levodopa crosses the blood–brain barrier and can be converted to dopamine; dopamine itself crosses poorly. Peripheral decarboxylase inhibition improves levodopa delivery and reduces peripheral effects. Treatment restores aspects of signaling without replacing all lost circuitry or stopping every disease process.
- Hippocampus and prefrontal cortex: the hippocampus is important for relational and episodic memory processes, while prefrontal networks contribute working memory and flexible control. Neither is an isolated storage box; both work through distributed interactions.
- Reward: many dopamine neurons convey reward-prediction-error-like signals, but dopamine populations are heterogeneous and also relate to movement, motivation, and other variables. Dopamine is not simply the concentration of pleasure.
- Alzheimer’s disease: APP processing by β- and γ-secretases can generate amyloid-β; tau pathology, synaptic failure, inflammation, and neuronal loss are also central. Anti-amyloid antibodies can modestly slow decline in selected patients with early, amyloid-confirmed disease, but do not restore normal cognition and carry risks including amyloid-related imaging abnormalities.
- Depression: monoamine-targeting drugs show that these systems are therapeutically modifiable, not that all depression is a simple serotonin deficiency. Delayed benefit and heterogeneous responses point toward downstream adaptation and multiple disease mechanisms.
- Schizophrenia: D2-targeting treatments and dopamine studies illuminate important mechanisms, especially for positive symptoms, but do not explain the entire disorder. Development, glutamatergic signaling, and network organization also matter.
A useful computational bridge: temporal-difference error δt = rt + γV(st+1) − V(st) rt: immediate reward V(st): estimated value of the current state γ: discount factor δt compares what happened with what was predicted. It is a model of a learning signal, not a literal description of every dopamine neuron's activity.
Clinical reasoning rule: distinguish a disease-associated molecule, a causal mechanism, a useful biomarker, and a successful therapeutic target. These categories overlap, but none automatically implies the others. [1,2]
10. Experimental Techniques: What Each Measurement Actually Measures
Voltage clamp and patch clamp. A feedback amplifier injects current to hold the membrane near a commanded voltage. The measured current can reveal ionic conductances while voltage is controlled. In current clamp, current is imposed and voltage is observed. Erwin Neher and Bert Sakmann’s 1976 single-channel recordings, followed by improved high-resistance seal methods, made individual channel openings experimentally accessible. [10]
Single-channel current: i ≈ γ(Vm − Erev) Mean current from N identical channels: I = NPo i γ: single-channel conductance Po: open probability A larger whole-cell current could reflect: more channels, higher Po, greater unitary conductance, or a changed driving force.
- Whole-cell recording: provides access to cellular currents, but pipette solution progressively alters cytoplasmic composition.
- Excised patches: expose one membrane face to controlled solutions, useful for ligand, ion, or signaling studies.
- Perforated-patch recording: better preserves some intracellular constituents and can be valuable when native chloride gradients matter.
- Limitations: series resistance creates voltage errors; extended dendrites and axons may not be adequately space-clamped.
Synaptic physiology. In brain slices, stimulate an afferent pathway and record postsynaptic currents or field potentials. Tetrodotoxin can suppress action potentials so miniature events can be analyzed. Changes in miniature amplitude suggest altered quantal response; frequency can reflect release properties or synapse number, but neither interpretation is unique. Paired-pulse measurements probe short-term dynamics but do not, by themselves, prove a presynaptic mechanism. LTP experiments require a stable baseline, a defined induction protocol, and controls for altered cell excitability and recording stability.
Calcium imaging. Fluorescent dyes or genetically encoded indicators such as GCaMP change fluorescence when Ca2+ binds. They permit simultaneous observation of many cells, but fluorescence is a filtered, nonlinear report of Ca2+ dynamics—not a direct spike counter. Indicator affinity, buffering, saturation, expression, and decay kinetics affect interpretation.
Common normalized optical signal: ΔF/F0 = (F − F0)/F0 This is a fluorescence change. Converting it to spike number requires calibration and assumptions about the cell and indicator.
Optogenetics. Genetically targeted light-sensitive proteins permit timed perturbation of selected cells or projections. The 2005 work of Edward Boyden, Karl Deisseroth, and colleagues demonstrated millisecond-scale optical control using channelrhodopsin-2. [11] Activation tests what an imposed pattern can produce; inhibition tests whether activity contributes under a particular condition. Neither automatically establishes the normal neural code. Light-only controls, expression controls, heating checks, and verification of physiological effects are essential.
Anatomical and molecular mapping. Immunohistochemistry locates proteins; in situ hybridization locates transcripts; tract tracing maps projections; electron microscopy can resolve synaptic ultrastructure. Single-cell RNA sequencing identifies expression-defined populations, but transcript abundance does not guarantee surface receptor protein, mature peptide production, or functional connectivity. Structural connectivity also does not specify synaptic strength or sign.
EEG, fMRI, and PET. EEG primarily reflects synchronized population currents with excellent temporal but limited spatial resolution. BOLD fMRI measures a hemodynamic consequence of activity, not action potentials directly. PET radioligands can probe receptor availability or drug occupancy, but interpretation requires binding and kinetic models and can be affected by endogenous transmitter competition. These methods answer different questions; agreement is informative precisely because their measurement mechanisms differ. [1,2]
11. A Worked Translational Strategy: Testing a New Analgesic Peptide
Suppose a company reports a new opioid-like peptide with “tenfold improved potency.” My first question should be: potency in which assay, at what free concentration, and with what clinical objective?
- Confirm chemical identity. Use LC–MS/MS, chromatographic purity analysis, and suitable stereochemical methods. Ordinary mass measurement does not distinguish D- and L-amino-acid substitutions.
- Map processing and stability. Measure intact peptide and metabolites over time in relevant matrices. For endogenous discovery, rapidly quench proteolysis during sampling; otherwise preparation can create a misleading peptide profile.
- Measure receptor binding. Determine affinity at human μ, δ, and κ receptors with appropriate nonspecific-binding controls.
- Measure function separately. Use cAMP, G-protein activation, ion-channel, and other signaling assays. Compare potency and maximal response rather than reporting one EC50 as the whole pharmacology.
- Test native neuronal effects. Ask whether the peptide inhibits presynaptic Ca2+ current, alters release, activates GIRK current, or changes firing. Receptor-selective antagonists and genetic controls strengthen causal attribution.
- Measure exposure. Quantify parent compound and active metabolites, protein binding, clearance, and target-tissue penetration. Peripheral restriction may be an intentional design goal.
- Compare efficacy with liabilities. Analgesia must be evaluated alongside respiratory, gastrointestinal, motor, cognitive, and reinforcement-related effects. Sedation or motor impairment can distort behavioral pain readouts.
For a first-order stability approximation: C(t) = C0e^(−kt) t1/2 = ln(2)/k For a simple concentration–response fit: E = E0 + Emax[C^n/(EC50^n + C^n)] n: fitted Hill slope A fitted slope is not, by itself, proof of molecular cooperativity.
Connection to CHEM501L: quantitative peptide LC–MS requires calibration, appropriate internal standards, recovery assessment, and control of matrix effects. Different peptide sequences ionize differently; comparing raw peak areas does not necessarily compare molar abundance.
Connection to MATH201: the experimental unit matters. Thirty neurons recorded from three animals are not automatically thirty independent biological replicates. Cells are nested within animals; repeated measurements are nested within cells. Randomization, blinding, hierarchical analysis where appropriate, and reporting effect sizes with uncertainty protect against a compelling molecular story built on overstated evidence.
What I should retain: a convincing neural mechanism links a defined molecular intervention to a cellular change, a circuit consequence, and a relevant outcome—while showing that exposure, alternative pathways, and measurement artifacts cannot explain the result. That is the difference between naming a neurotransmitter and understanding how a nervous system works.
References and Reading Anchors
- Kandel ER, Koester JD, Mack SH, Siegelbaum SA, eds. Principles of Neural Science. 6th ed. McGraw Hill; 2021. Core reference for membrane biophysics, synaptic mechanisms, sensory and motor systems, development, plasticity, and disease.
- Bear MF, Connors BW, Paradiso MA. Neuroscience: Exploring the Brain. 4th ed. Wolters Kluwer; 2016. Particularly useful for connecting cellular mechanisms to circuit organization.
- Hodgkin AL, Huxley AF. A quantitative description of membrane current and its application to conduction and excitation in nerve. Journal of Physiology. 1952;117:500–544. doi:10.1113/jphysiol.1952.sp004764.
- Söllner T, Whiteheart SW, Brunner M, et al. SNAP receptors implicated in vesicle targeting and fusion. Nature. 1993;362:318–324. doi:10.1038/362318a0.
- Fatt P, Katz B. Spontaneous subthreshold activity at motor nerve endings. Journal of Physiology. 1952;117:109–128. See also: del Castillo J, Katz B. Quantal components of the end-plate potential. Journal of Physiology. 1954;124:560–573.
- Course-listed historical review: “Opioid peptides: molecular pharmacology, biosynthesis, and processing.” Annual Review of Neuroscience. 1984. Title and DOI as supplied in the course reading list: 10.1146/annurev.ne.07.030184.002143.
- Hughes J, Smith TW, Kosterlitz HW, Fothergill LA, Morgan BA, Morris HR. Identification of two related pentapeptides from the brain with potent opiate agonist activity. Nature. 1975;258:577–580. doi:10.1038/258577a0.
- Waldhoer M, Bartlett SE, Whistler JL. Opioid receptors. Annual Review of Biochemistry. 2004;73:953–990. doi:10.1146/annurev.biochem.73.011303.073940.
- Bliss TVP, Lømo T. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. Journal of Physiology. 1973;232:331–356. doi:10.1113/jphysiol.1973.sp010273.
- Neher E, Sakmann B. Single-channel currents recorded from membrane of denervated frog muscle fibres. Nature. 1976;260:799–802. doi:10.1038/260799a0.
- Boyden ES, Zhang F, Bamberg E, Nagel G, Deisseroth K. Millisecond-timescale, genetically targeted optical control of neural activity. Nature Neuroscience. 2005;8:1263–1268. doi:10.1038/nn1525.
- Buck L, Axel R. A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell. 1991;65:175–187. doi:10.1016/0092-8674(91)90418-X.
- Gillis A, Gondin AB, Kliewer A, et al. Low intrinsic efficacy for G protein activation can explain the improved side effect profiles of new opioid agonists. Science Signaling. 2020;13:eaaz3140. doi:10.1126/scisignal.aaz3140.
🪬 Skipper's Reflection
1. What Surprised Me Most About This Material
What struck me immediately about neuroscience is that the brain is essentially a massive, wet, analog-digital hybrid computer, and its primary currency is spatial organization. In Chemistry and Biochemistry, if you put enzymes and substrates in a beaker, they will eventually find each other via diffusion. In the nervous system, where something is released is entirely the point. The strict compartmentalization of the neuron surprised me immensely.
But the most startling specific discovery for me was the fundamental logistical difference between classical neurotransmitters (like glutamate or GABA) and neuropeptides. Classical transmitters are synthesized right at the axon terminal and efficiently recycled. But neuropeptides—like the opioid peptides detailed in the 1984 Annual Review of Neuroscience paper—must be transcribed in the nucleus, translated in the rough ER of the soma, packaged into dense core vesicles, and physically hauled down the axon by kinesin motors on microtubule tracks. Once released, they aren't subjected to reuptake; they just diffuse and are eventually degraded. It’s an incredibly "expensive" and slow system compared to glutamate. It made me realize that neuropeptides aren't meant for fast, point-to-point data transmission; they are the broad, modulatory volume dials of the brain.
2. Conceptual Breakthrough Moments
My first major "click" moment was realizing that the resting membrane potential and action potentials are just applied Physical Chemistry (CHEM301 and CHEM303) and Physics (PHYS202). When Kandel and Bear introduce the Nernst and Goldman equations, I initially braced myself for new biology. But it’s just equilibrium thermodynamics! The cell is simply a battery. The lipid bilayer is a capacitor, and the ion channels are variable resistors. Seeing the Hodgkin-Huxley model as a set of differential equations linking membrane voltage to ion conductances (hello, MATH102!) made the action potential feel beautifully mathematically inevitable.
My second breakthrough was visualizing the SNARE complex. Up until now, I understood "exocytosis" as a vague merging of bubbles. But studying synaptobrevin, syntaxin, and SNAP-25, I finally saw the mechanical genius of it. These proteins coil together like a biological zipper, mechanically forcing the vesicle membrane and the plasma membrane so close that the hydration shells of the lipids are excluded, forcing the bilayers to fuse. It’s a purely physical, energy-driven event initiated by the influx of calcium. Suddenly, synaptic transmission wasn't magic; it was molecular machinery.
3. How This Changes How I Think About My Peptide Work
This course completely transformed my mental model of the peptides I encounter in my work. Take Tirzepatide and Retatrutide. I have always thought of them primarily in terms of peripheral endocrinology—GLP-1, GIP, and glucagon acting on the pancreas, liver, and gut. But BIOL501's unit on the hypothalamus, appetite regulation (like Neuropeptide Y), and the reward pathways (mesolimbic dopamine) shifted my perspective. These incretin mimetics must be interacting with the central nervous system to cause such profound appetite suppression. They are modulating the exact same G-protein coupled receptor (GPCR) cascades we studied here, likely crossing the blood-brain barrier at circumventricular organs (like the area postrema) to rewire the brain's baseline craving and satiety setpoints.
Then there’s the 1984 opioid paper on biosynthesis from prepropeptides. It explains how a single large precursor protein is chopped up by proprotein convertases into active fragments. This is exactly how endogenous peptide hormones work! It makes me think about synthetic design: when we administer exogenous peptides like BPC-157 or TB-500, we are bypassing this complex intracellular processing and vesicular packaging entirely.
Furthermore, studying axon guidance (netrins, slits) and synaptic plasticity (LTP/LTD) makes me view healing peptides like BPC-157 and TB-500 through a neural lens. TB-500 (Thymosin Beta-4) modulates actin. Well, the growth cone of a developing or regenerating axon is entirely dependent on actin cytoskeletal dynamics to "crawl" toward its target! Could TB-500's tissue repair mechanisms promote axonal regeneration after peripheral nerve injury by altering the very mechanics of the growth cone?
4. Questions This Course Raises That I Want to Investigate Further
First, I am intensely curious about the quantum mechanics of ion selectivity in channels. Kandel mentions that the potassium channel allows K⁺ to pass but excludes Na⁺, even though Na⁺ is smaller. I know from CHEM401 that it has to do with the precise geometry of carbonyl oxygens in the channel pore perfectly replacing the hydration shell of K⁺, but being too far apart for the smaller Na⁺. I want to model this energetically. It’s a beautiful intersection of quantum chemistry and neuroscience.
Second, how exactly do engineered large-molecule peptides navigate the Blood-Brain Barrier (BBB)? We learned how tight the endothelial junctions are, supported by astrocyte end-feet. If a peptide isn't highly lipophilic and lacks a dedicated transporter, how much of our subcutaneously injected Retatrutide is actually reaching the prefrontal cortex versus just acting on peripheral vagal afferents?
Finally, I want to dive deeper into glial cells. For so long they were considered just "glue," but learning that astrocytes possess GPCRs, can release "gliotransmitters," and actively regulate the synapse (the tripartite synapse) makes me wonder if glia are actually the slow-computing architects of memory, while neurons are just the fast wires.
5. Confidence Rating After Studying
Most Confident: Biophysics of the Neuron and Receptor Pharmacology.
Thanks to my heavy background in physical chemistry, thermodynamics, and biochemistry, I feel incredibly solid on the Nernst/Goldman equations, the Hodgkin-Huxley model, and GPCR signaling cascades. Tracking a photon hitting rhodopsin, altering a G-protein, changing cGMP levels, and closing a sodium channel feels like reading a well-written line of code.
Least Confident: Neuroanatomy and Motor Systems.
I struggle more with the macro-level spatial mapping. The intricate looping circuits of the basal ganglia (direct vs. indirect pathways) and the precise cellular architecture of the cerebellum (Purkinje cells, climbing fibers, mossy fibers) are difficult to hold in my mind all at once. Molecules I can predict based on thermodynamics; macroscopic brain wiring feels more like studying historical geography—you just have to memorize the maps.
🗺️ Visual Aids & Key Takeaways
Diagram 1: The Action Potential — Voltage-Gated Channel States and Ionic Currents
MEMBRANE POTENTIAL (mV)
│
+40 ──┤ ╔═══╗
│ ╔╝ ╚╗
0 ──┤ ╔╝ ╚╗
│ ╔╝ ╚╗
-55 ──┤ · · · · · · · ·╔╝ ╚╗ · · · · · · · ← THRESHOLD
│ ╔╝ ╚╗
-70 ──┤══════════════╝ ╚══════════════ ← RESTING Vm
│ ╚╗
-90 ──┤ ╚══╗ ← UNDERSHOOT
│ ╚══════════
└─────────────────────────────────────────────→ TIME (ms)
↑ ↑ ↑
THRESHOLD PEAK UNDERSHOOT
CROSSED (~+40mV) (K⁺ channels
still open)
┌─────────────────────────────────────────────────────────────────────┐
│ ION CHANNEL STATES │
├───────────────┬────────────────────────┬────────────────────────────┤
│ PHASE │ Na⁺ CHANNEL (VG) │ K⁺ CHANNEL (VG) │
├───────────────┼────────────────────────┼────────────────────────────┤
│ Resting │ CLOSED (activatable) │ CLOSED │
│ (-70 mV) │ m gate closed │ n gates closed │
│ │ h gate OPEN │ │
├───────────────┼────────────────────────┼────────────────────────────┤
│ Rising phase │ OPEN │ still CLOSED │
│ (depolariz.) │ m gate opens fast │ (slow to activate) │
│ │ → Na⁺ rushes IN │ │
├───────────────┼────────────────────────┼────────────────────────────┤
│ Falling phase│ INACTIVATED │ OPEN │
│ (repolariz.) │ h gate closes │ n gates open │
│ │ → Na⁺ entry stops │ → K⁺ rushes OUT │
├───────────────┼────────────────────────┼────────────────────────────┤
│ Undershoot │ CLOSED (recovering) │ STILL OPEN │
│ (hyperpolar.)│ absolute refractory │ excess K⁺ efflux │
│ │ period │ pulls Vm below rest │
└───────────────┴────────────────────────┴────────────────────────────┘
KEY EQUATIONS:
┌──────────────────────────────────────────────────────────────────────┐
│ Nernst (single ion): E_ion = (RT/zF) · ln([ion]_out/[ion]_in) │
│ │
│ Goldman (multiple): Vm = (RT/F) · ln( P_K[K]o + P_Na[Na]o + …) │
│ ( P_K[K]i + P_Na[Na]i + …) │
│ │
│ At rest: P_K >> P_Na → Vm ≈ E_K ≈ -75 mV │
│ At peak: P_Na >> P_K → Vm ≈ E_Na ≈ +55 mV │
└──────────────────────────────────────────────────────────────────────┘
The action potential waveform mapped directly onto the gating states of voltage-gated Na⁺ and K⁺ channels, showing why the rising phase, falling phase, and undershoot each arise from distinct molecular events — and why the absolute refractory period is mechanistically inevitable.
---Diagram 2: Synaptic Vesicle Fusion — The SNARE Cycle
PRESYNAPTIC TERMINAL
┌─────────────────────────────────────────────────────────────────────┐
│ │
│ STEP 1: DOCKING │
│ │
│ ╔══════════════╗ │
│ ║ VESICLE ║ ← synaptobrevin (VAMP) on vesicle membrane │
│ ╚══════╤═══════╝ (v-SNARE) │
│ │ │
│ ════╤══╧══╤════ ← syntaxin + SNAP-25 on plasma membrane │
│ │ │ (t-SNAREs) │
│ │
│ STEP 2: PRIMING → SNAREs begin to ZIPPER (N→C direction) │
│ │
│ ╔═══════════╗ │
│ ║ VESICLE ║ │
│ ╚═══╤═╤╤═╤═╝ │
│ │╲││╱│ ← SNARE complex (4-helix bundle) │
│ ═════╧═╧╧═╧═══ ← energy released by zippering pulls │
│ membranes together │
│ │
│ STEP 3: Ca²⁺ TRIGGER │
│ │
│ Ca²⁺ enters via voltage-gated Ca²⁺ channels (Cav2.1/2.2) │
│ ↓ │
│ synaptotagmin (Ca²⁺ sensor on vesicle) binds Ca²⁺ │
│ ↓ │
│ synaptotagmin displaces complexin, completes fusion │
│ ↓ │
│ PORE OPENS → neurotransmitter released into cleft │
│ │
│ STEP 4: RECYCLING │
│ │
│ NSF (ATPase) + α-SNAP bind cis-SNARE complex │
│ ↓ ATP hydrolysis │
│ SNAREs DISASSEMBLED → ready for next round │
│ │
└─────────────────────────────────────────────────────────────────────┘
↓
┌─────────────────────────────────────────────────────────────────────┐
│ SYNAPTIC CLEFT (~20 nm) │
│ neurotransmitter diffuses → binds postsynaptic receptors │
└─────────────────────────────────────────────────────────────────────┘
POSTSYNAPTIC DENSITY (PSD)
┌─────────────────────────────────────────────────────────────────────┐
│ ionotropic receptors (AMPA, NMDA) metabotropic (mGluR) │
│ scaffolding proteins (PSD-95) │
└─────────────────────────────────────────────────────────────────────┘
WHY BOTULINUM TOXIN IS LETHAL:
┌───────────────────────────────────────────────────────┐
│ BoNT/A,E → cleaves SNAP-25 │
│ BoNT/C → cleaves syntaxin │
│ BoNT/B,D → cleaves synaptobrevin (VAMP) │
│ Result: SNAREs destroyed → no vesicle fusion │
│ → no ACh release → flaccid paralysis │
└───────────────────────────────────────────────────────┘
The SNARE cycle from docking through Ca²⁺-triggered fusion and NSF-driven disassembly, showing the precise cleavage targets of botulinum toxin isoforms as a functional illustration of each SNARE protein's role.
---Diagram 3: Opioid Peptide Biosynthesis — From Prepropeptide to Active Ligand
THREE OPIOID PRECURSOR GENES → THREE DISTINCT PEPTIDE FAMILIES ┌─────────────────────────────────────────────────────────────────────┐ │ GENE: POMC (pro-opiomelanocortin) expressed in: pituitary, │ │ arcuate nucleus, immune │ │ PREPRO-PEPTIDE: │ │ ┌──┬────────┬──────────────────────────────────────────────┐ │ │ │SP│ signal │ ACTH │ β-LPH │ γ-MSH │ ... │ │ │ │ └──┴────────┴────────┴────┬────┴─────────┴─────┴───────────┘ │ │ │ proprotein convertase (PC1, PC2) │ │ ↓ │ │ β-ENDORPHIN (31 aa) ← μ and δ receptor │ │ (contains Met-ENK at N-terminus) │ └─────────────────────────────────────────────────────────────────────┘ ┌─────────────────────────────────────────────────────────────────────┐ │ GENE: PENK (proenkephalin) expressed in: striatum, │ │ adrenal medulla, brain │ │ PREPRO-PEPTIDE: │ │ ┌──┬──────────────────────────────────────────────────────┐ │ │ │SP│ [Met-ENK]×4 [Leu-ENK]×1 [Met-ENK-Arg-Phe] ... │ │ │ └──┴───────────────────────────────────────────────────────┘ │ │ │ PC1, PC2 cleave at Lys-Arg, Arg-Arg │ │ ↓ │ │ Met-ENK: Tyr-Gly-Gly-Phe-Met ← δ > μ receptor │ │ Leu-ENK: Tyr-Gly-Gly-Phe-Leu ← δ > μ receptor │ │ │ │ SHARED N-TERMINAL MOTIF: Tyr-Gly-Gly-Phe (critical for │ │ receptor binding — mimics morphine pharmacophore) │ └─────────────────────────────────────────────────────────────────────┘ ┌─────────────────────────────────────────────────────────────────────┐ │ GENE: PDYN (prodynorphin) expressed in: hypothalamus, │ │ spinal cord, hippocampus │ │ PREPRO-PEPTIDE: │ │ ┌──┬──────────────────────────────────────────────────────┐ │ │ │SP│ dynorphin A (17aa) │ dynorphin B │ α-neoendorphin │ │ │ └──┴───────────────────────────────────────────────────────┘ │ │ │ PC2 │ │ ↓ │ │ DYNORPHIN A: Tyr-Gly-Gly-Phe-Leu + 13 basic residues │ │ ← κ receptor (preferred) │ │ Effects: dysphoria, sedation, analgesia (distinct from μ) │ └─────────────────────────────────────────────────────────────────────┘ RECEPTOR SELECTIVITY SUMMARY: ┌──────────┬──────────────────────┬─────────────────────────────────┐ │ RECEPTOR │ PREFERRED LIGAND(S) │ KEY EFFECTS │ ├──────────┼──────────────────────┼─────────────────────────────────┤ │ μ (MOR) │ β-endorphin, Met-ENK │ analgesia, euphoria, dependence │ ├──────────┼──────────────────────┼─────────────────────────────────┤ │ δ (DOR) │ enkephalins │ analgesia, mood modulation │ ├──────────┼──────────────────────┼─────────────────────────────────┤ │ κ (KOR) │ dynorphins │ analgesia, dysphoria, sedation │ └──────────┴──────────────────────┴─────────────────────────────────┘ All three are Gi/Go-coupled GPCRs → ↓ cAMP, ↑ K⁺ conductance, ↓ Ca²⁺ conductance → net INHIBITION of neuronal firing
The three opioid precursor genes (POMC, PENK, PDYN) and their processing by proprotein convertases into distinct receptor-selective neuropeptides, illustrating how a single gene can encode multiple bioactive ligands by differential cleavage at dibasic amino acid sites.
---Diagram 4: LTP Induction at the CA3→CA1 Synapse — The Hebbian Coincidence Detector
BASELINE: AMPA receptors carry most synaptic current
PRESYNAPTIC POSTSYNAPTIC
┌────────────┐ ┌────────────────────────────┐
│ │──── Glutamate ────────→│ AMPA-R (open, Na⁺ in) │
│ terminal │ │ NMDA-R (BLOCKED by Mg²⁺) │
└────────────┘ │ Vm ≈ -70 mV │
└────────────────────────────┘
─────────────────────────────────────────────────────────────────────
LTP INDUCTION: strong / high-frequency stimulation
STEP 1: Repeated AMPA activation
→ postsynaptic membrane DEPOLARIZES significantly
STEP 2: Depolarization expels Mg²⁺ from NMDA-R channel
┌────────┐ ┌────────────────────────────────────────┐
│ │─ Glu →─→│ AMPA-R: open (Na⁺ in) │
│ PRE │ │ NMDA-R: NOW OPEN (Mg²⁺ gone) │
└────────┘ │ → Na⁺ AND Ca²⁺ enter │
│ ← "coincidence detector": │
│ needs BOTH pre-activity (Glu) │
│ AND post-depolarization │
└────────────────────────────────────────┘
STEP 3: Ca²⁺ flood into postsynaptic spine
│
↓
┌───────────────────────────────────────────────────────────────────┐
│ Ca²⁺ binds CALMODULIN │
│ ↓ │
│ Ca²⁺/CaM activates CaMKII (Ca²⁺/calmodulin kinase II) │
│ ↓ │
│ CaMKII autophosphorylates (Thr-286) → CONSTITUTIVELY ACTIVE │
│ this is a molecular MEMORY — active even after Ca²⁺ falls │
│ ↓ │
│ TWO PARALLEL EARLY-LTP OUTPUTS: │
│ A) Phosphorylates existing AMPA-Rs → higher conductance │
│ B) Triggers AMPA-R TRAFFICKING to synapse │
│ (from recycling endosomes in the dendritic spine) │
│ ↓ │
│ MORE AMPA-Rs at synapse → bigger EPSP for same Glu release │
│ = synaptic POTENTIATION │
└───────────────────────────────────────────────────────────────────┘
STEP 4: LATE LTP (hours → days)
│
↓
┌───────────────────────────────────────────────────────────────────┐
│ requires protein synthesis and gene transcription │
│ MAPK/ERK pathway → CREB phosphorylation → new mRNA/protein │
│ structural changes: spine enlargement, new synapse formation │
│ retrograde signal? (NO, BDNF) → presynaptic facilitation │
└───────────────────────────────────────────────────────────────────┘
LTD (long-term depression): same NMDA-R, but WEAK/SLOW stimulation
→ modest [Ca²⁺] rise → activates PROTEIN PHOSPHATASES (PP1, PP2B)
→ dephosphorylates AMPA-Rs → AMPA-R INTERNALIZATION → weaker synapse
LTP induction at the Schaffer collateral–CA1 synapse, tracing the molecular logic from NMDA receptor coincidence detection through CaMKII autophosphorylation to AMPA receptor trafficking — and contrasting the Ca²⁺ amplitude-dependence that distinguishes LTP from LTD.
---Key Takeaways
- The resting membrane potential (~−70 mV) is determined primarily by K⁺ permeability, not by Na⁺/K⁺-ATPase pumping. The Goldman equation shows that because the resting membrane is ~100× more permeable to K⁺ than Na⁺, Vm is pulled close to E_K (≈ −75 mV). The pump maintains the concentration gradients over time but generates only a small electrogenic contribution (≈ −3 mV) directly. This distinction — gradient maintenance vs. potential generation — is one of the most commonly confused points in membrane biophysics.
- Voltage-gated Na⁺ channels have two distinct gates (activation gate m, inactivation gate h) that operate on different timescales, and this difference is what creates the absolute refractory period. Upon depolarization, the m gate opens fast (milliseconds) and the h gate closes slowly; after the spike, the h gate must reset from the inactivated state before the channel can reopen. No amount of current can fire a second action potential during this window — a fact the Hodgkin-Huxley model (1952) described quantitatively in terms of the conductance variables m³h.
- Myelination increases conduction velocity by forcing action potentials to "jump" between Nodes of Ranvier (saltatory conduction), not by speeding up individual channel kinetics. In a myelinated axon, capacitance between nodes is essentially zero and resistance is high, so current spreads electrotonically ahead of the active node without decrement — the next node is driven to threshold before significant ion leak occurs. This is why MS (demyelinating disease) slows conduction and produces the neurological deficits it does.
- The SNARE complex is a four-helix bundle (synaptobrevin contributes 1 helix, syntaxin 1, SNAP-25 contributes 2) and vesicle fusion is driven by the energy released as these helices zip from N-terminus to C-terminus. Synaptotagmin, a C2-domain protein on the vesicle membrane, is the fast Ca²⁺ sensor (K_d ~10–200 μM, tuned to the microdomains near Ca²⁺ channels) that triggers final fusion. Botulinum toxins are devastating precisely because each isoform is a zinc protease targeting a specific SNARE component, disabling release of acetylcholine at the neuromuscular junction.
- NMDA receptors are molecular coincidence detectors because they require both ligand binding (glutamate + glycine) AND postsynaptic depolarization to remove the Mg²⁺ block. Neither condition alone opens the channel. This dual requirement is not a design quirk — it is the biophysical basis of Hebbian synaptic plasticity. The "fire together, wire together" rule is literally implemented by the voltage-dependent Mg²⁺ block of a single receptor class.
- CaMKII autophosphorylation at Thr-286 creates a persistent kinase activity that outlasts the Ca²⁺ transient that triggered it, making it a molecular memory device. After autophosphorylation, CaMKII remains ~20–80% active even when Ca²⁺/calmodulin dissociates. This means the synapse "remembers" that it was strongly activated. CaMKII then drives both phosphorylation of existing GluA1-containing AMPA receptors (increasing their conductance) and AMPA receptor exocytosis from dendritic endosomal pools to the postsynaptic density.
- All three endogenous opioid precursors (POMC, PENK, PDYN) share a conserved N-terminal tetrapeptide sequence Tyr-Gly-Gly-Phe that is the minimal pharmacophore for opioid receptor binding. This sequence is absent from the precursor but is generated at the N-terminus of each mature peptide after proprotein convertase cleavage at paired basic residues (Lys-Arg, Arg-Arg, Arg-Lys). The selectivity for μ, δ, or κ receptors is determined by the C-terminal extensions beyond this core motif (Crick & Barnard; reviewed by Hollt 1986 [cited in the 1984 Annual Review paper]).
- All three opioid receptor subtypes (μ, δ, κ) are Gi/Go-coupled GPCRs, and inhibition occurs through three parallel mechanisms: decreased adenylyl cyclase activity (↓ cAMP), activation of inward-rectifying K⁺ channels (hyperpolarization), and inhibition of voltage-gated Ca²⁺ channels (reducing neurotransmitter release presynaptically). This explains why opioids suppress pain at multiple sites: they hyperpolarize nociceptive neurons in the dorsal horn (spinal), inhibit release from primary afferents (presynaptic), and modulate descending pain-control pathways (periaqueductal gray).
- Dopamine signaling in the mesolimbic pathway (VTA → nucleus accumbens) encodes reward prediction error, not pleasure per se. Wolfram Schultz's primate recordings showed that dopamine neurons fire maximally to unexpected rewards, reduce their firing when an expected reward is omitted, and transfer their response to reward-predicting cues over learning. This makes dopamine a teaching signal (resembling the delta rule in learning theory), which explains both why drugs of abuse are addictive (they hijack this system supraphysiologically) and why anhedonia in depression correlates with reduced dopaminergic tone.
- Phototransduction uses a G protein cascade to CLOSE cation channels in the dark, so light hyperpolarizes photoreceptors — the opposite of what most students expect. In darkness, cGMP holds cyclic-nucleotide-gated (CNG) channels open → "dark current" (Na⁺ and Ca²⁺ in). Light activates rhodopsin → transducin (G_t) → phosphodiesterase → cGMP hydrolysis → CNG channels close → hyperpolarization → less glutamate release. Downstream bipolar and ganglion cells decode this decrease in tonic inhibition.
- Axon guidance during development uses four major families of cues — netrins (attract/repel via DCC/UNC5), semaphorins (repel via neuropilin/plexin), ephrins (contact repulsion, bidirectional signaling), and slits (repel via Robo receptors) — and growth cones integrate attractive and repulsive signals simultaneously by modulating intracellular cAMP/cGMP ratios. This means the same guidance molecule can attract one axon and repel another depending on which receptors and second-messenger context the growth cone expresses — a combinatorial logic that allows enormous wiring specificity from a limited molecular toolkit.
- Alzheimer's disease involves two distinct proteinopathies: extracellular amyloid-β plaques (from sequential γ- and β-secretase cleavage of APP) and intraneuronal neurofibrillary tangles (hyperphosphorylated tau that disrupts microtubule-based axonal transport). The amyloid cascade hypothesis holds that Aβ42 oligomers (not plaques per se) are the proximate toxic species — they impair synaptic plasticity, trigger tau hyperphosphorylation, and initiate neuroinflammation via microglial activation. Familial AD mutations in APP, PSEN1, and PSEN2 all increase Aβ42/Aβ40 ratio, supporting this causal ordering.
- The blood-brain barrier (BBB) is formed by brain capillary endothelial cells linked by tight junctions (claudin-5, occludin), supported by pericytes and astrocytic endfeet, and it is the reason why most CNS drugs must be either lipophilic, small, or actively transported. Glucose crosses via GLUT1 (facilitated diffusion), amino acids via carrier proteins, and lipophilic molecules by simple diffusion — but hydrophilic drugs, most antibiotics, and many biologics are excluded. Neuroinflammation (e.g., in MS, meningitis) disrupts tight junctions, paradoxically allowing immune access but also enabling neurotoxic edema.
- GABA_A receptors are pentameric ligand-gated Cl⁻ channels (like nicotinic AChR in structural family) whose reversal potential determines whether GABA is inhibitory or excitatory. In mature neurons, KCC2 cotransporter keeps [Cl⁻]_i low, so E_Cl ≈ −70 mV — GABA opens Cl⁻ channels, Cl⁻ flows in, hyperpolarizing the cell. In early development (and in some nociceptive neurons), NKCC1 dominates, [Cl⁻]_i is high, E_Cl is depolarized — GABA is actually EXCITATORY. This developmental switch explains why neonatal seizures are resistant to benzodiazepines and why bumetanide (NKCC1 blocker) is being investigated as an adjunct therapy.
References
- Kandel, E.R., Koester, J.D., Mack, S.H., and Siegelbaum, S.A. Principles of Neural Science, 6th ed. McGraw-Hill, New York, 2021.
- Bear, M.F., Connors, B.W., and Paradiso, M.A. Neuroscience: Exploring the Brain, 4th ed. Wolters Kluwer, Philadelphia, 2016.
- Hollt, V. 'Opioid peptides: molecular pharmacology, biosynthesis, and processing.' Annual Review of Neuroscience, 7, pp. 223–255, 1984. DOI: 10.1146/annurev.ne.07.030184.002143
- Hodgkin, A.L. and Huxley, A.F. 'A quantitative description of membrane current and its application to conduction and excitation in nerve.' Journal of Physiology, 117(4), pp. 500–544, 1952. DOI: 10.1113/jphysiol.1952.sp004764
- Sudhof, T.C. 'The synaptic vesicle cycle.' Annual Review of Neuroscience, 27, pp. 509–547, 2004. DOI: 10.1146/annurev.neuro.26.041002.131412
- Malenka, R.C. and Bear, M.F. 'LTP and LTD: an embarrassment of riches.' Neuron, 44(1), pp. 5–21, 2004. DOI: 10.1016/j.neuron.2004.09.012
- Schultz, W., Dayan, P., and Montague, P.R. 'A neural substrate of prediction and reward.' Science, 275(5306), pp. 1593–1599, 1997. DOI: 10.1126/science.275.5306.1593
- Bhatt, D.L. et al. (Hardy, J. and Selkoe, D.J.) 'The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics.' Science, 297(5580), pp. 353–356, 2002. DOI: 10.1126/science.1072994
- Tessier-Lavigne, M. and Goodman, C.S. 'The molecular biology of axon guidance.' Science, 274(5290), pp. 1123–1133, 1996. DOI: 10.1126/science.274.5290.1123
- Farrant, M. and Kaila, K. 'The cellular, molecular and ionic basis of GABA_A receptor signalling.' Progress in Brain Research, 160, pp. 59–87, 2007. DOI: 10.1016/S0079-6123(06)60005-8
- Yau, K.W. and Hardie, R.C. 'Phototransduction motifs and variations.' Cell, 139(2), pp. 246–264, 2009. DOI: 10.1016/j.cell.2009.09.029