Chapter 1: Cellular and Cognitive Neuroscience

Chapter Introduction

The Turtle has walked with you a long way.

In K-12 you met your brain — 86 billion neurons, three pounds of soft tissue, the most complex object in the known universe. At Associates you went into neuroscience proper — neurons and glia, the action potential at ion-channel depth, the major neurotransmitter systems, the four lobes and the limbic anatomy, long-term potentiation and the BDNF cascade, the HPA axis and allostatic load, Posner's attention networks and the reward circuitry, the integrator move that names every other Coach's content as acting on the brain through identifiable neural mechanisms. At the end of Associates you could read a primary neuroscience paper and recognize what you were looking at.

This chapter is the second step of the upper-division spiral.

At the Bachelor's level, Coach Brain goes deeper at three axes simultaneously. The cellular axis goes molecular: where Associates said neurons fire action potentials, Bachelor's writes the Hodgkin-Huxley equations as the mathematical model they actually are, traces saltatory conduction as physics rather than metaphor, and walks synaptic transmission at the molecular detail of the SNARE complex and vesicle fusion. The cognitive axis goes circuit-level: where Associates introduced working memory, Bachelor's enters the dorsolateral prefrontal persistent activity literature (Goldman-Rakic, D'Esposito), the fronto-striatal loops in detail, and the unification of Schultz's dopamine prediction error with Sutton and Barto's temporal-difference learning that made modern computational neuroscience possible. The methodological axis goes core: where Associates referenced fMRI in passing, Bachelor's enters the BOLD signal at signal-detection depth, the multiple-comparisons problem and the dead-salmon paper, the optogenetic and chemogenetic revolution that transformed circuit neuroscience in the 2000s and 2010s, and the reproducibility crisis honestly addressed.

The voice is the same Turtle. Patient. Methodical. Slow and deep. Expects you to keep up. What changes is the methodological consciousness. Upper-division work means you can no longer take a neuroscience finding at face value. You read it as a finding from a specific study using specific methods, in a specific organism or population, and you ask what the design could and could not have shown. This is what distinguishes Bachelor's neuroscience from Associates: the methodological discipline that lets you evaluate claims rather than receive them.

A word about what this chapter is not, before you begin. This chapter is not a diagnostic manual. Depression, anxiety, ADHD, PTSD, OCD, schizophrenia-spectrum conditions, and substance use disorders are real, well-researched, and present in these pages — descriptively, as topics that the field studies at neuroscience depth. They are not framed as conditions for you to diagnose in yourself or in others. The neuroscience belongs here. The diagnostic question belongs in a clinician's office.

A word about being a pre-health or neuroscience-major student, before you begin. Some of you are heading to medical school, doctoral neuroscience research, clinical psychology, neuropsychology, physician assistant programs, or directly into healthcare. The clinical neuropsychiatry in this chapter is calibrated for that pathway: research-grade pathophysiology of mood disorders, anxiety, addiction, and the major psychiatric conditions, framed throughout as recognition and clinical evaluation rather than diagnostic prescription. Patients receive diagnoses, prescriptions, and treatment plans from licensed clinicians. The chapter builds the mechanistic understanding that informs clinical conversation, not confidence to substitute for it.

A word about mental health, before you begin. The late teens and early twenties are a real-incidence peak for the first onset of several mental health conditions. Many people experience their first depressive episode, first major anxiety, first encounter with addictive patterns, or first manic or psychotic episode during the college years. The research-grade depth in this chapter — neurochemistry of depression, dopamine and reward, the HPA-stress-mood axis — is content that may surface things you are working through. If anything you read here lands close to your own experience and you are working through it alone when you do not need to be, the verified crisis resources at the end of this chapter are real. Your college counseling center is real. Your primary care provider is real. The Turtle is patient with you.

This chapter has five lessons.

Lesson 1 is Cellular and Molecular Neuroscience — the Hodgkin-Huxley mathematical model of the action potential, saltatory conduction in myelinated axons as a physics rather than a metaphor, synaptic transmission at the molecular detail of presynaptic calcium entry, SNARE-complex-mediated vesicle fusion, and postsynaptic receptor pharmacology (NMDA, AMPA, kainate, GABA-A, GABA-B, the metabotropic glutamate receptors and their G-protein cascades), and the neuromodulation systems at receptor-and-circuit depth.

Lesson 2 is Neuroplasticity and Memory at Molecular Resolution — the LTP/LTD molecular cascade (CaMKII autophosphorylation, AMPAR trafficking and phosphorylation states, CREB phosphorylation, PKMζ as a candidate maintenance molecule), the Kandel cellular molecular cascade of long-term memory, BDNF/TrkB signaling, the adult human hippocampal neurogenesis controversy at the methodological level (Sorrells 2018 versus Boldrini 2018 — what each study measured, why their conclusions differ, what would resolve it), and spatial memory at circuit depth from O'Keefe's place cells through the Mosers' grid cells (Nobel work).

Lesson 3 is Stress Neurobiology and Pathophysiology — the HPA axis at glucocorticoid- and mineralocorticoid-receptor depth, the GR/MR ratio in different brain regions, McEwen's allostatic load with primary-literature backing, chronic stress effects on hippocampal dendritic morphology and prefrontal function (Sapolsky primate work, Lupien lifespan research), the neuroscience of depression at research-grade depth (the monoamine hypothesis past and present, the inflammatory hypothesis, the glutamate/ketamine paradigm shift, the HPA dysregulation contribution), and the first Bachelor's-tier lateral reference: HPA-metabolic intersection with Coach Food Bachelor's Lesson 2 on energy regulation.

Lesson 4 is Executive Function, Decision Neuroscience, and Reward — dorsolateral prefrontal working memory at neuroimaging-and-electrophysiology depth (Goldman-Rakic, D'Esposito), the fronto-striatal loops in detail, the unification of Schultz's dopamine prediction error with Sutton and Barto's temporal-difference learning, the prefrontal-limbic balance in self-control and emotion regulation, and addiction neurobiology at Berridge wanting/liking and Nestler ΔFosB depth.

Lesson 5 is Research Methods in Neuroscience — fMRI BOLD signal at the signal-detection level (Ogawa 1990 as the foundational anchor), the inferential chain from BOLD to cognition, the spatial-and-temporal-resolution tradeoffs, the dead-salmon and multiple-comparisons problem, single-unit and patch-clamp electrophysiology, EEG/MEG, optogenetics (Boyden/Deisseroth 2005 forward) and chemogenetics (Roth/DREADDs) as the methodological revolution, the reproducibility crisis in neuroscience and psychology honestly addressed, and the application of the five-point evaluation framework to neuroscience claims specifically.

The Turtle is in no hurry. Begin.

---

Lesson 1: Cellular and Molecular Neuroscience

Learning Objectives

By the end of this lesson, you will be able to:

• Describe the Hodgkin-Huxley model as a mathematical model of voltage-gated ion-channel kinetics and identify how the model produces the action potential as an emergent property
• Explain the physics of saltatory conduction in myelinated axons, including why the node-of-Ranvier architecture increases conduction velocity by approximately an order of magnitude
• Walk synaptic transmission from presynaptic calcium influx through SNARE-complex-mediated vesicle fusion to postsynaptic receptor activation
• Distinguish the principal ionotropic glutamate receptors (NMDA, AMPA, kainate), the GABA-A and GABA-B receptors, and the metabotropic glutamate receptors at the level of subunit composition, ionic selectivity, and signaling cascade
• Identify the principal neuromodulation systems (dopaminergic, noradrenergic, serotonergic, cholinergic, histaminergic) and the receptor classes through which each operates

Key Terms

The Hodgkin-Huxley Model as Mathematical Object

At Associates depth, you read that the action potential is generated by voltage-gated sodium and potassium channels. At Bachelor's depth, you meet the mathematical model that made the cellular description possible.

In 1952, Alan Hodgkin and Andrew Huxley published a series of papers in the Journal of Physiology describing the squid giant axon and producing the first quantitative model of the action potential [1]. The squid axon — large enough to accept the glass microelectrodes of the era — let Hodgkin and Huxley apply voltage clamp (holding membrane voltage at chosen values while measuring transmembrane currents) and dissect the ionic currents into separable components. They identified an inward sodium current (responsible for depolarization) and an outward potassium current (responsible for repolarization), characterized the voltage- and time-dependence of each, and wrote a system of four coupled differential equations that, when integrated, reproduced the action potential as an emergent property of the underlying ionic conductances [2].

The equations describe:

• Membrane potential V_m as a function of total ionic current and membrane capacitance.
• Three gating variables: m (sodium channel activation), h (sodium channel inactivation), and n (potassium channel activation), each obeying a first-order rate equation with voltage-dependent rate constants α(V) and β(V).
• Total ionic current as the sum of Na⁺, K⁺, and leak currents, each gated according to its activation and inactivation variables.

The model predicts not only the action-potential waveform but also threshold behavior, the refractory period, the conduction velocity along an axon, and the response to subthreshold current injection. It is one of the most successful quantitative models in biology. Hodgkin and Huxley received the Nobel Prize in 1963.

Three notes that matter at upper-division depth:

First, the model was constructed before the molecular identity of voltage-gated channels was known. Hodgkin and Huxley inferred the existence of voltage-gated sodium and potassium channels — and the gating particles whose stochastic motion produces the macroscopic currents — from electrophysiological observation. The molecular reality (the alpha subunit of Na_v 1.x channels, the four-domain transmembrane architecture, the S4 voltage sensor, the ball-and-chain inactivation mechanism) was confirmed decades later. This is a paradigm example of how careful biophysical modeling can predict molecular biology.

Second, the model is deterministic at the macroscopic level. Individual voltage-gated channels gate stochastically (Markov process); the macroscopic conductance Hodgkin-Huxley described is the ensemble average of many stochastic channels. At small membrane patches with few channels, single-channel stochasticity produces threshold variability that the deterministic model does not capture. This is one of the reasons modern computational neuroscience extends Hodgkin-Huxley with stochastic models in some contexts.

Third, modern neurons express dozens of channel types beyond the squid-axon sodium and potassium. The Hodgkin-Huxley framework generalizes — additional channels are added as additional conductances in the membrane equation, each with its own voltage- and time-dependence — but the resulting models can become high-dimensional. Detailed compartmental models of cortical pyramidal neurons may include 10-15 channel types distributed non-uniformly across dendrites, soma, and axon, with morphology incorporated through cable-theory partial differential equations.

The take-home for upper-division neuroscience is that the action potential is not a metaphor or a cartoon; it is a quantitatively-modeled physical phenomenon, and the model has been one of the foundational successes of biophysics.

Saltatory Conduction: Physics, Not Metaphor

In an unmyelinated axon, action potentials propagate continuously along the membrane: local current from one active region depolarizes adjacent membrane to threshold, that region fires, and the process repeats. Conduction velocity scales approximately with the square root of axon diameter; in a 1 μm unmyelinated axon, velocity is around 1 m/s.

Myelination — the wrapping of axons by oligodendrocyte (CNS) or Schwann cell (PNS) processes — dramatically increases velocity by changing the propagation physics. Myelin acts as an electrical insulator with high resistance and low capacitance. The internodal segment of axon under myelin behaves nearly as a passive cable rather than as active excitable membrane: depolarization propagates electrotonically with minimal current loss across the membrane. At the node of Ranvier — a short bare segment between myelin sheaths — voltage-gated sodium channels are densely clustered, and action potentials are regenerated. The combination produces saltatory conduction: the action potential jumps from node to node, with internodal propagation orders of magnitude faster than membrane-bound regeneration [3].

For a typical myelinated mammalian axon at 5-10 μm diameter, conduction velocity reaches 50-100 m/s — roughly ten times faster than an unmyelinated axon of comparable diameter. The velocity gain comes at a cost: peripheral demyelinating diseases (Guillain-Barré syndrome) and central demyelinating diseases (multiple sclerosis) impair saltatory conduction and produce the functional deficits characteristic of those conditions [4].

The pre-clinical relevance: the architecture of myelinated axons connects directly to the pathophysiology of demyelinating disease, to the developmental window in which myelination occurs (which extends well into the third decade in human prefrontal cortex), and to current neuroscience research on adult oligodendrocyte function and remyelination as targets for MS therapeutics.

Synaptic Transmission at Molecular Detail

When an action potential reaches the axon terminal, it triggers a sequence of events that culminates in neurotransmitter release. At Bachelor's depth the cascade is:

1. Presynaptic depolarization opens voltage-gated Ca²⁺ channels (principally P/Q-type Ca_v 2.1 and N-type Ca_v 2.2 in central terminals).
2. Calcium influx raises intraterminal [Ca²⁺] from ~100 nM (resting) to ~10-100 μM in microdomains adjacent to channels — a thousand-fold local elevation.
3. Synaptotagmin, the presynaptic calcium sensor, binds Ca²⁺ through its C2A and C2B domains. The conformational change triggers vesicle fusion.
4. The SNARE complex — syntaxin and SNAP-25 on the plasma membrane, synaptobrevin (VAMP) on the vesicle — has already zippered into a four-helix bundle that draws the vesicle into close apposition with the membrane. Synaptotagmin's Ca²⁺-bound state catalyzes the final fusion step.
5. Vesicle fusion opens a fusion pore through which neurotransmitter (glutamate, GABA, neuromodulators, peptides) is released into the synaptic cleft.
6. Post-fusion recovery involves SNARE disassembly (NSF and α-SNAP), vesicle endocytosis (clathrin-mediated or other), neurotransmitter reuptake (vesicular transporters), and vesicle re-acidification for refilling. The entire cycle takes seconds to tens of seconds depending on vesicle pool dynamics.

Thomas Südhof shared the 2013 Nobel Prize in Physiology or Medicine with James Rothman and Randy Schekman for the molecular dissection of vesicle trafficking, including the identification of synaptotagmin's calcium-sensor role and the molecular architecture of SNARE-mediated fusion [5]. The presynaptic machinery is one of the most molecularly-characterized signaling systems in cell biology.

The clinical relevance: many neurological diseases involve presynaptic machinery. Botulinum toxin (Clostridium botulinum) cleaves SNARE proteins, blocking neurotransmitter release at neuromuscular junctions and producing flaccid paralysis. Tetanus toxin cleaves the inhibitory-neuron SNAREs preferentially, producing spastic paralysis. Lambert-Eaton myasthenic syndrome involves antibodies against presynaptic Ca²⁺ channels. The molecular detail Südhof's laboratory mapped is not abstract: it is the substrate of clinical neuropathology.

Postsynaptic Receptors: Glutamate, GABA, and the Cascade

Neurotransmitter binds postsynaptic receptors, which fall into two principal classes: ionotropic (ligand-gated ion channels) and metabotropic (G-protein-coupled receptors). The receptor classes for the two major fast neurotransmitters:

Glutamate ionotropic receptors include three families [6]:

• AMPA receptors — heterotetrameric (GluA1-4) channels permeable to Na⁺ and K⁺ (some subunit compositions also permeable to Ca²⁺). Fast kinetics (millisecond opening); the principal mediator of fast excitatory transmission. AMPAR phosphorylation, trafficking, and subunit composition are central to LTP/LTD (Lesson 2).
• NMDA receptors — heterotetrameric (typically GluN1 + GluN2A or GluN2B) channels permeable to Na⁺, K⁺, and notably Ca²⁺. Gating requires both ligand (glutamate) and a voltage-dependent relief of Mg²⁺ block — making them coincidence detectors for pre- and post-synaptic activity. Slow kinetics (tens to hundreds of milliseconds). The principal molecular substrate of associative learning rules.
• Kainate receptors — heterotetrameric (GluK1-5) channels with kinetics and trafficking distinct from AMPA; play roles in modulation of release and in certain forms of plasticity.

Glutamate metabotropic receptors (mGluR) — eight members in three groups — couple to Gq (Group I: mGluR1, mGluR5 → PLC → IP₃/DAG/Ca²⁺) or Gi/o (Groups II and III → inhibition of adenylate cyclase, modulation of synaptic transmission).

GABA receptors are the principal inhibitory machinery [7]:

• GABA-A receptors — pentameric ligand-gated chloride channels. Subunit composition (α1-6, β1-3, γ1-3, δ, ε, π, θ, ρ1-3) determines pharmacology and localization. Benzodiazepines bind at the α/γ subunit interface and potentiate GABA action. Barbiturates, ethanol, certain anesthetics, and neurosteroids also modulate GABA-A receptors at distinct sites. The clinical pharmacology of anxiolysis, sedation, and anesthesia is largely GABA-A pharmacology.
• GABA-B receptors — heterodimeric G-protein-coupled receptors signaling through Gi/o; mediate slow inhibitory currents through K⁺ channel activation and presynaptic Ca²⁺ channel inhibition. Baclofen is a GABA-B agonist used clinically for spasticity.

The fast excitatory-inhibitory balance in cortical circuits — typically 80% glutamatergic pyramidal cells and 20% GABAergic interneurons distributed across morphological and molecular subtypes (parvalbumin, somatostatin, VIP, and others) — is one of the architectural features that produces cortical computation. Disruption of E/I balance is implicated in epilepsy, autism spectrum, and schizophrenia at varying levels of evidence.

Neuromodulation Systems

Beyond fast excitatory and inhibitory transmission, the brain operates a set of neuromodulation systems — small numbers of neurons in specific brainstem and basal-forebrain nuclei that project diffusely throughout the brain, releasing monoamines, acetylcholine, and peptides that modify the function of fast transmission. The principal systems at receptor depth:

Dopaminergic system — Ventral tegmental area (VTA, mesolimbic and mesocortical projections) and substantia nigra pars compacta (SNc, nigrostriatal projection). Dopamine signals through five GPCR families: D1-like (D1, D5; Gs → adenylate cyclase → cAMP) and D2-like (D2, D3, D4; Gi/o → inhibition of adenylate cyclase). Phasic dopamine signals reward prediction error (Schultz, Lesson 4); tonic dopamine modulates motivation, motor function, and working memory. Pharmacology: antipsychotic drugs are predominantly D2 antagonists (with newer atypicals adding 5-HT2A antagonism); levodopa replacement in Parkinson's; stimulants (amphetamine, methylphenidate) act through DAT and VMAT2 [8].

Noradrenergic system — Locus coeruleus (LC) in the pons; projects diffusely. Norepinephrine signals through α1, α2, β1, β2, β3 GPCRs. Phasic LC firing accompanies salient and attention-demanding events; tonic LC firing modulates arousal. The Aston-Jones / Cohen adaptive-gain theory frames LC-norepinephrine as governing the exploit/explore balance. Pharmacology: norepinephrine reuptake inhibitors (NRIs) in depression and ADHD; α2 agonists (clonidine, guanfacine) in attention disorders and hypertension [9].

Serotonergic system — Dorsal and median raphe nuclei in the brainstem. Serotonin (5-HT) signals through a remarkably large family of receptors: 5-HT1A-F, 5-HT2A-C, 5-HT3 (the only ionotropic), 5-HT4, 5-HT5A-B, 5-HT6, 5-HT7. Most are GPCRs with distinct downstream cascades. The serotonin system has been implicated in mood, anxiety, aggression, sleep, and many other processes; the complexity of receptor pharmacology is part of why "the serotonin system" is rarely a single coherent target [10].

Cholinergic system — Basal forebrain (nucleus basalis of Meynert, medial septum, diagonal band) provides cortical and hippocampal cholinergic projection; brainstem nuclei (pedunculopontine, laterodorsal tegmental) provide thalamic and brainstem cholinergic input. Acetylcholine signals through nicotinic receptors (ionotropic) and muscarinic receptors (M1-5, GPCRs). The cholinergic basal forebrain is degenerated in Alzheimer's disease; cholinesterase inhibitors (donepezil, rivastigmine, galantamine) provide modest symptomatic benefit [11].

Histaminergic system — Tuberomammillary nucleus in the hypothalamus; projects diffusely. Histamine signals through H1-H4 GPCRs. The system regulates wakefulness; H1 antagonists are sedating (the older antihistamines). Modafinil and pitolisant act through histaminergic pathways in narcolepsy treatment.

The neuromodulation systems collectively span the state dimension of brain function — arousal, attention, motivation, mood, sleep-wake organization. They operate at slower timescales than fast transmission (hundreds of milliseconds to seconds for receptor activation; minutes to hours for downstream gene expression effects) and shape the responsiveness of fast circuits to incoming information. The Bachelor's-level point is that the brain's computational function emerges from the interaction of fast point-to-point transmission with slow diffuse modulation, not from either alone.

Lesson Check

1. Walk the Hodgkin-Huxley framework. What did Hodgkin and Huxley experimentally observe, and what did they infer? What did the model predict that the molecular biology of the time could not directly confirm?
2. Explain saltatory conduction as physics: what does myelin do to membrane electrical properties, and why does the result speed conduction?
3. Trace synaptic transmission from presynaptic depolarization to postsynaptic receptor activation, naming the role of SNARE complex and synaptotagmin specifically.
4. Compare AMPA, NMDA, and metabotropic glutamate receptors at the level of mechanism. Why is NMDA called a coincidence detector?
5. Identify two pharmacological agents that act on the GABA-A receptor, and explain how their binding modulates receptor function relative to GABA alone.
6. Distinguish phasic from tonic activity in midbrain dopamine neurons and identify what each is thought to signal.

---

Lesson 2: Neuroplasticity and Memory at Molecular Resolution

Learning Objectives

By the end of this lesson, you will be able to:

• Walk the LTP molecular cascade from NMDA-receptor-mediated Ca²⁺ influx through CaMKII autophosphorylation, AMPAR trafficking, and CREB-driven gene expression
• Distinguish early-LTP from late-LTP and identify the protein-synthesis-dependence boundary
• Engage with the adult human hippocampal neurogenesis controversy at the methodological level (Sorrells 2018 versus Boldrini 2018) and articulate what would resolve it
• Describe place cells (O'Keefe) and grid cells (Moser, Moser) and their roles in spatial navigation and memory
• Identify the principal techniques (engram tagging, optogenetics) by which Tonegawa and colleagues have demonstrated necessity-and-sufficiency of specific neural populations for specific memories

Key Terms

LTP at Molecular Depth

At Associates depth, you read that LTP is the cellular substrate of learning, mediated principally through NMDA receptors at hippocampal CA1 synapses. At Bachelor's depth, the molecular cascade is:

1. Glutamate release from CA3 presynaptic terminals binds postsynaptic AMPA receptors, depolarizing the CA1 dendritic spine.
2. Sufficient depolarization relieves the voltage-dependent Mg²⁺ block of NMDA receptors. NMDARs become permeable to Na⁺, K⁺, and notably Ca²⁺.
3. Ca²⁺ influx through NMDA receptors, supplemented by Ca²⁺ release from intracellular stores, raises spine [Ca²⁺] into the micromolar range.
4. Ca²⁺/calmodulin activates CaMKII. CaMKII is a dodecameric holoenzyme; activated subunits autophosphorylate at Thr286, producing a persistent active state that outlasts the original Ca²⁺ signal. This autophosphorylation is one of the principal candidate mechanisms for the persistence of LTP that distinguishes it from transient synaptic facilitation [12][13].
5. Activated CaMKII has many substrates. The best-characterized include:
   • AMPA receptors — CaMKII phosphorylates GluA1 at Ser831, increasing single-channel conductance.
   • Stargazin / TARP — auxiliary AMPAR subunits whose phosphorylation regulates receptor trafficking into the postsynaptic density.
   • AMPAR insertion — additional AMPARs are trafficked to the synaptic membrane from recycling endosomes, increasing the postsynaptic response to a given quantum of glutamate.
6. In parallel, other kinases (PKA, PKC, MAPK) are activated. PKA-driven cAMP signaling and MAPK signaling phosphorylate CREB at Ser133. Phospho-CREB drives transcription of plasticity-related genes including BDNF, Arc, Zif268, and others [14][15].
7. Late-phase LTP requires this gene-expression-dependent step. Protein-synthesis inhibitors administered within the first 1-2 hours after induction prevent late-LTP without blocking early-LTP. The protein-synthesis-dependence boundary is one of the operational definitions distinguishing the two phases.
8. The newly-synthesized proteins return to the activated synapse — synaptic tagging and capture (Frey and Morris 1997) — to maintain the structural and functional changes that constitute late-LTP. Spine enlargement, new spine formation, and changes in postsynaptic-density composition occur over hours to days following induction.

The cellular cascade has been extended in important ways. PKMζ — a constitutively active atypical PKC isoform — was proposed by Sacktor and colleagues as a maintenance molecule for LTP and long-term memory, with experiments showing that ZIP (a peptide inhibitor) erased established memories in rodent models [16]. The PKMζ hypothesis attracted substantial attention. Subsequent conditional knockout studies (Volk, Bauer, Lee, Huganir 2013) showed that PKMζ-knockout mice retain normal LTP and memory, complicating the strict PKMζ-required model [17]. The current research-grade picture is that PKMζ is sufficient (when present) but not strictly necessary (other mechanisms can compensate when it is absent throughout development) — a more nuanced account than the original proposal. This is one of the recurring patterns at upper-division depth: a striking proposal, an apparent confirmation, and a more careful subsequent picture.

LTD — the complementary process — operates through different intracellular machinery. Low-frequency stimulation produces moderate (rather than large) postsynaptic Ca²⁺ elevations that preferentially activate phosphatases (calcineurin, PP1) rather than kinases. Dephosphorylation of GluA1, AMPAR endocytosis, and reduction in postsynaptic conductance produce the persistent weakening that constitutes LTD [18]. The bidirectional Ca²⁺-amplitude-dependent regulation — high Ca²⁺ activates kinases and produces LTP; moderate Ca²⁺ activates phosphatases and produces LTD — provides one mechanism by which the same synapse can be strengthened or weakened depending on input pattern.

The clinical implications are extensive but largely indirect. NMDA-receptor function is implicated in schizophrenia (the NMDA-hypofunction hypothesis, supported by the psychotomimetic effects of NMDA antagonists like ketamine and PCP at high doses), in stroke (excitotoxicity), in chronic pain (central sensitization), and in autism spectrum (with some monogenic forms involving NMDA-pathway genes). Drugs acting on NMDA function — memantine in Alzheimer's, ketamine in treatment-resistant depression, esketamine intranasal — represent some of the molecular handles by which LTP-relevant biology meets clinical neuroscience [19].

Kandel's Cellular Cascade of Long-Term Memory

Eric Kandel's laboratory at Columbia worked out the cellular and molecular basis of long-term memory using Aplysia californica, a marine mollusk with a numerically simple nervous system (~20,000 neurons) and large identifiable cells suitable for intracellular recording. The Aplysia work, conducted from the late 1960s through the 1990s, established several principles that generalize to vertebrate memory [20]:

• Short-term sensitization of the gill-withdrawal reflex involves serotonergic facilitation of sensory-motor neuron synapses. The cellular mechanism: serotonin activates Gs → adenylate cyclase → cAMP → PKA → phosphorylation of presynaptic K⁺ channels (broadening the action potential and prolonging Ca²⁺ influx) and direct phosphorylation of the release machinery. The behavioral effect lasts minutes to hours.
• Long-term sensitization (lasting days to weeks) requires the same initial signaling cascade plus PKA translocation to the nucleus, CREB phosphorylation, and gene transcription. CREB drives expression of CCAAT/enhancer-binding protein (C/EBP), of new synaptic-machinery proteins, and of structural changes including the growth of new synapses.
• The molecular switch between short- and long-term memory is the gene-expression step. Without it, sensitization is transient; with it, sensitization is persistent and accompanied by structural change.
• CREB-1 (a transcriptional activator) and CREB-2 (a repressor) form a balance; tonic CREB-2 repression must be relieved to permit the long-term memory cascade. The result is that long-term memory formation requires removing inhibition as well as adding activation — explaining in part why long-term memory is a regulated rather than automatic process [21].

The Aplysia work and parallel work in Drosophila (Tully, Greenspan) and mammals (Bourtchouladze, Silva, others) converged on the CREB-dependent gene-expression-dependent picture of long-term memory. Kandel received the 2000 Nobel Prize in Physiology or Medicine for this body of work. The cellular cascade he mapped operates with surprising conservation from invertebrate sensitization to vertebrate hippocampal long-term memory.

BDNF/TrkB Signaling and Adult Plasticity

BDNF (brain-derived neurotrophic factor) signals through the TrkB (tropomyosin receptor kinase B) receptor tyrosine kinase. The cascade:

1. BDNF binds TrkB → dimerization and autophosphorylation of intracellular tyrosines.
2. Phosphorylated TrkB recruits adapter proteins (Shc, FRS2) and propagates signal through three principal cascades: PI3K/Akt (cell survival, growth), Ras/Raf/MEK/ERK (transcriptional regulation, synaptic plasticity), and PLCγ/IP₃/DAG/Ca²⁺ (synaptic plasticity, gene expression).
3. Downstream targets include CREB (transcriptional activation), AMPAR trafficking machinery (potentiating synaptic transmission), and the autophagy/UPR machinery (cell-state regulation).

BDNF is one of the most-studied molecules in neuroplasticity research. Activity-dependent BDNF release supports LTP maintenance, dendritic growth, spine formation, and adult-onset plasticity. BDNF expression is upregulated by exercise (the Erickson 2011 randomized trial in older adults found exercise-induced hippocampal volume change correlated with serum BDNF elevation, with the framing as supportive evidence for the BDNF mechanism [22]). BDNF is reduced in depression and in chronic stress; antidepressant treatment (including SSRIs and ketamine) restores BDNF expression in animal models, contributing to the neurotrophic hypothesis of depression that Lesson 3 returns to [23].

A polymorphism in BDNF — Val66Met — affects activity-dependent BDNF release; carriers of the Met allele show subtle differences in hippocampal volume, episodic memory performance, and treatment response in some psychiatric conditions. The polymorphism is one of the most-studied behavioral-genetic variants in neuroscience; effect sizes are modest, and replication has been variable, consistent with the broader pattern of behavioral-genetic associations [24].

The Adult Human Hippocampal Neurogenesis Controversy

Adult neurogenesis — the generation of new neurons in adult brain — was established in rodents in the 1990s through 2000s. Newborn neurons in the dentate gyrus subgranular zone integrate into hippocampal circuits, contribute to certain forms of pattern separation, and respond to environmental enrichment, exercise, and antidepressant treatment. The rodent picture is well-established [25].

Whether adult human hippocampal neurogenesis occurs at functionally meaningful rates is contested. Two 2018 papers reached opposite conclusions:

• Sorrells et al. 2018 in Nature: Studying surgical and postmortem human hippocampal samples across the lifespan with antibody staining for neurogenesis markers (DCX, PSA-NCAM), the authors found abundant neurogenic markers in childhood, declining sharply through adolescence, and reaching essentially undetectable levels in adults. Their conclusion: adult human hippocampal neurogenesis is rare or absent [26].
• Boldrini et al. 2018 in Cell Stem Cell: Studying postmortem human hippocampi with stereological counting and a different set of markers (including Ki67, DCX, PCNA), the authors found preserved neurogenic activity across the adult lifespan — including in older adults. Their conclusion: adult human hippocampal neurogenesis persists throughout aging [27].

The methodological controversy at upper-division depth:

The two studies differed in tissue source (Sorrells: more surgical samples; Boldrini: more postmortem), tissue processing (postmortem interval and fixation conditions affect immunoreactivity of neurogenesis markers), the specific markers used, the quantification approach, and the inclusion criteria. The DCX (doublecortin) immunoreactivity that anchors much of the field has been argued by Sorrells and others to potentially reflect persistent immature-but-not-newly-generated neurons, complicating its use as a strict neurogenesis marker.

Subsequent work has continued to produce mixed results. A 2019 Nature Medicine paper from the Llorens-Martín group reported preserved adult human neurogenesis using improved fixation protocols (Moreno-Jiménez et al.) [28]. A 2019 Cell Stem Cell paper using single-nucleus RNA sequencing from human hippocampus (Habib et al.) did not detect transcriptomic signatures of newborn neurons in adult samples. The methodological question — whether the right markers are being measured under the right conditions — remains unresolved at the level of consensus.

For the Bachelor's audience, the controversy is pedagogically useful precisely because it is unresolved. The standard inferential question — given conflicting findings, what would resolve the disagreement? — has identifiable answers:

• Direct lineage tracing in living human brain (technically infeasible)
• ¹⁴C bomb-pulse dating of human hippocampal DNA (Spalding 2013 reported evidence for ongoing turnover [29]; methodologically distinct from histological approaches)
• Consensus methodology for tissue handling and marker validation
• Larger samples and pre-registered analytical pipelines to reduce researcher-degrees-of-freedom contributions

The upper-division take-away is not which side is right but how to read a methodologically contested research area. The neuroscience of adult human neurogenesis remains an active area of work; clinical translation (the idea that interventions might support neurogenesis to improve mood, memory, or aging outcomes) is more speculative than the popular press has often implied.

Place Cells, Grid Cells, and Spatial Memory

In 1971, John O'Keefe and Jonathan Dostrovsky reported in Brain Research that single hippocampal pyramidal neurons in freely-moving rats fired selectively when the animal occupied specific locations in an environment — place cells [30]. The discovery established that the hippocampus encodes spatial location; an animal's position can in principle be reconstructed from a population of place cell activities. O'Keefe and Lynn Nadel's 1978 monograph The Hippocampus as a Cognitive Map developed the theoretical framework. O'Keefe received a share of the 2014 Nobel Prize in Physiology or Medicine for this work [31].

In 2005, Edvard Moser and May-Britt Moser's laboratory in Trondheim reported in Nature that entorhinal cortex neurons fired in a regular hexagonal grid pattern across an environment — grid cells [32]. Grid cells provide a metric on space; their hexagonal firing fields are consistent across environments (the grid spacing and orientation are relatively stable, but the grid translates with respect to environmental landmarks). The entorhinal-hippocampal circuit thus provides both metric (grid cells in entorhinal cortex) and place-specific (place cells in hippocampus) representations of space, with the two interacting through the perforant path. The Mosers shared the 2014 Nobel with O'Keefe.

Subsequent work has identified additional cell types in the spatial-memory circuit: head-direction cells (orientation), border cells (proximity to environmental boundaries), object-vector cells (relative position to specific objects), and time cells (temporal position within a sequence of events). The collective architecture of entorhinal-hippocampal spatial-temporal representation is one of the best-mapped cognitive circuits in mammalian neuroscience.

The clinical relevance: the entorhinal cortex is one of the first regions affected in Alzheimer's disease, with neurofibrillary tangles appearing in the entorhinal cortex (Braak stage I-II) well before the cortex more broadly. The early disruption of spatial-temporal encoding helps explain why disorientation and topographic memory loss are among the earliest symptoms of the disease. Grid-cell-like signals can be measured in human entorhinal cortex with high-resolution fMRI; reduced grid-cell-pattern signal has been observed in young APOE4 carriers (a major AD risk allele) decades before cognitive symptoms typically appear [33].

Engram Tagging and the Causal Demonstration of Memory

Through the 2000s and 2010s, Susumu Tonegawa's laboratory at MIT and others used the convergence of optogenetics (Lesson 5) and activity-dependent gene expression to identify engrams — the specific neural populations whose activity supports a specific memory — and to demonstrate causal sufficiency for memory recall.

The general methodology:

1. Tag — A mouse expresses a tetracycline-controlled construct in which neurons active during a specific experience (typically conditioning) drive expression of an effector (channelrhodopsin for optogenetics, hM3Dq for chemogenetics) in those active neurons. Activity-dependent immediate-early-gene promoters (c-Fos, Arc) drive the tagging.
2. Retrieve — In a different context (where the original cue is absent), the tagged neurons are artificially reactivated by light (optogenetics) or by ligand (chemogenetics for hM3Dq systems).
3. Observe — The animal expresses the memory-associated behavior in the new context, demonstrating that artificial reactivation of the tagged ensemble is sufficient to reinstate the memory state [34].

Subsequent work has demonstrated engram-suppression-mediated forgetting, false-memory generation through reactivation of one engram during a different experience, and engram-based rescue of memory deficits in models of Alzheimer's. The engram-tagging methodology has transformed memory research from a primarily correlational science to one capable of causal manipulation of specific memory representations.

The Bachelor's-level take-away: the cellular and molecular biology of memory has progressed from "memory involves synaptic change" (a verbal description) to "specific synaptic changes in specific neural populations are necessary and sufficient for specific memories" (a causal demonstration). The methods that made this possible are part of Lesson 5; the conceptual achievement is part of how upper-division neuroscience has shifted in the last two decades.

Lesson Check

1. Walk the LTP molecular cascade from NMDAR Ca²⁺ influx through CaMKII autophosphorylation, AMPAR trafficking, and CREB-driven gene expression. Identify the protein-synthesis-dependence boundary between early- and late-LTP.
2. Distinguish LTP and LTD at the level of postsynaptic Ca²⁺ amplitude and downstream kinase/phosphatase balance.
3. Describe Kandel's cellular cascade of long-term memory in Aplysia, including the molecular role of CREB and the CREB-1/CREB-2 balance.
4. Compare Sorrells 2018 and Boldrini 2018 on adult human hippocampal neurogenesis. What did each study measure, why might their conclusions differ, and what kind of evidence would resolve the controversy?
5. Describe place cells and grid cells and identify what each cell type contributes to the spatial-memory system.
6. Explain the engram-tagging methodology and what it demonstrates causally about memory storage.

---

Lesson 3: Stress Neurobiology and Pathophysiology

Learning Objectives

By the end of this lesson, you will be able to:

• Trace the HPA axis from PVN through anterior pituitary to adrenal cortex at receptor and gene-expression depth
• Distinguish glucocorticoid receptor (GR) and mineralocorticoid receptor (MR) function and identify the regional MR/GR ratio in hippocampus, prefrontal cortex, and amygdala
• Describe allostatic load (McEwen) with the supporting primary literature, including the dendritic-morphology research in hippocampus and PFC (Sapolsky, McEwen) and the lifespan stress effects (Lupien)
• Engage with the principal hypotheses of depression neurobiology: monoamine hypothesis (past and present), inflammatory hypothesis, glutamate/ketamine paradigm, HPA dysregulation
• Identify the principal anxiety neurocircuitry — amygdala, hippocampus, BNST, prefrontal regulation — and the limits of single-circuit models
• Articulate the HPA-metabolic intersection in cross-reference with Coach Food Bachelor's Lesson 2 on energy regulation

Key Terms

The HPA Axis at Receptor and Gene-Expression Depth

At Associates depth, you traced the HPA axis from the paraventricular nucleus (PVN) to cortisol release. At Bachelor's depth, the cascade has more detail and the receptor biology becomes essential.

1. Stressor input — Sensory, interoceptive, or cognitive inputs converge on the PVN parvocellular neurons. The amygdala can drive PVN activation directly and indirectly through brainstem relays; the prefrontal cortex provides inhibitory regulation. The integration is complex and context-dependent.
2. PVN CRH release — Corticotropin-releasing hormone (CRH, also called CRF) is released into the hypophyseal portal circulation from PVN terminals in the median eminence. Co-released vasopressin (AVP) potentiates the CRH signal at the pituitary.
3. Anterior pituitary ACTH release — CRH binds CRH-R1 receptors on anterior pituitary corticotropes, driving release of adrenocorticotropic hormone (ACTH) into systemic circulation.
4. Adrenal cortex cortisol synthesis — ACTH binds MC2R receptors on adrenal zona fasciculata cells, activating the steroidogenic cascade (cholesterol → pregnenolone → progesterone → 11-deoxycortisol → cortisol). Synthesis takes several minutes; the cascade is not pre-formed and stored.
5. Cortisol distribution and receptor binding — Cortisol enters circulation, mostly bound to corticosteroid-binding globulin (CBG) with a free fraction (~10%) biologically active. Free cortisol crosses cell membranes and binds intracellular receptors.

The receptor biology is where Bachelor's depth becomes essential. Cortisol acts on two nuclear hormone receptors with markedly different properties [35]:

• Mineralocorticoid Receptor (MR) — Higher affinity for cortisol (Kd ~0.5 nM). Nearly saturated at basal cortisol levels. Distributed prominently in hippocampus, especially CA1 and CA2, with lower expression elsewhere in brain.
• Glucocorticoid Receptor (GR) — Lower affinity for cortisol (Kd ~5 nM). Substantially occupied only at stress-elevated cortisol levels. Broadly distributed throughout brain, including hippocampus, prefrontal cortex, amygdala, and peripheral tissues.

The MR/GR ratio in different brain regions has functional consequences. In hippocampus, the high MR expression means that basal cortisol tonically supports MR-driven functions (some forms of plasticity, glucose handling, the diurnal cortisol rhythm's modulation of memory). At stress-elevated cortisol, GR occupation rises and produces the well-characterized stress-response effects. The MR/GR balance idea (de Kloet and colleagues) holds that the relative activation of the two receptor systems shapes the brain's stress response — neither MR alone nor GR alone, but the balance between them, mapping onto distinct behavioral and physiological outputs [36].

The downstream effects of GR activation are extensive. GR is a nuclear hormone receptor that binds glucocorticoid response elements (GREs) in target gene promoters, regulating transcription of hundreds of genes. The transcriptional program is tissue-specific (different genes are accessible in different cell types) and time-dependent (acute versus chronic activation produces different transcriptomes). Some immediate effects of GR activation in brain include energy mobilization, suppression of immune response, enhancement of memory consolidation for emotionally relevant material, and feedback regulation of CRH and ACTH release (the negative feedback that limits HPA-axis duration).

The HPA axis is also subject to negative feedback: cortisol acts on GR in the hippocampus, PVN, and anterior pituitary to reduce CRH and ACTH release, terminating the response. The feedback's failure — HPA dysregulation — is one of the most consistently observed neuroendocrine findings in major depressive disorder, with elevated cortisol, blunted dexamethasone suppression test responses, and altered diurnal cortisol patterns observed in many but not all depressed patients [37].

Chronic Stress, Allostatic Load, and Brain Morphology

Bruce McEwen's allostatic load framework, introduced with Eliot Stellar in 1993, conceptualizes the cumulative cost of repeated stress-response activation as the price the body pays for repeated adaptation [38]. The four allostatic states McEwen named:

1. Allostasis — appropriate response to a current stressor.
2. Allostatic load — accumulation of stress-response cost over time.
3. Allostatic overload Type 1 — energy demands exceed availability (a survival situation).
4. Allostatic overload Type 2 — high allostatic load without survival pressure (the chronic-stress situation typical of modern environments).

The brain-morphology consequences of chronic allostatic load have been mapped at substantial detail [39]:

• Hippocampal dendritic atrophy — In CA3 pyramidal neurons of chronically-stressed rodents, apical dendrites retract and lose branches. Sapolsky's primate work documented hippocampal atrophy in chronically-stressed olive baboons and in primates with extended cortisol elevation. Human imaging studies show reduced hippocampal volume in patients with chronic PTSD, major depression, Cushing's syndrome (a clinical hypercortisolism), and chronic high-stress occupational contexts.
• Prefrontal cortex effects — Medial PFC dendritic remodeling and reduced spine density follow chronic stress in rodents. Functional consequences include impaired working memory and executive function. Human imaging studies show PFC volume reductions in chronic-stress and depression contexts.
• Amygdala effects — In contrast to hippocampus and PFC, the basolateral amygdala typically shows hypertrophy (dendritic branching increases, spine density rises) following chronic stress. The functional consequence is elevated amygdala reactivity to threat — biased threat-detection and fear-learning that persists after the original stressor.

The collective effect — reduced hippocampal volume and function, impaired PFC regulation, hypertrophic amygdala threat-reactivity — produces a brain biased toward threat-detection and away from contextual flexibility. This morphological pattern overlaps substantially with the imaging findings in PTSD, chronic anxiety, and major depression, supporting the framework that these conditions involve persistent neuroplastic changes from chronic stress rather than (or in addition to) acute neurochemical disturbance.

Sonia Lupien and colleagues extended the framework across the lifespan, documenting that stress effects on brain development are most consequential during specific developmental windows: prenatal (placental cortisol exposure), early childhood (HPA-axis programming), adolescence (PFC maturation continues into the third decade), and aging (cumulative load over time) [40][41]. The lifespan framework has shaped contemporary understanding of how childhood adversity propagates into adult mental health.

Depression Neuroscience: Four Frameworks

The neuroscience of depression has gone through several frameworks. The Bachelor's-level approach is to hold all of them as partially-correct and complementary rather than as competing alternatives.

The monoamine hypothesis — emerged from the 1950s-1960s observations that drugs depleting monoamines (reserpine) sometimes produced depression and drugs elevating monoamine signaling (MAOIs, tricyclics, later SSRIs) sometimes treated it. The hypothesis frames depression as a serotonergic and/or noradrenergic deficiency. The framework has limits that have become clearer over time [42]:

• SSRIs elevate synaptic serotonin within hours but produce clinical effects only over weeks, indicating the immediate biochemistry is not the proximal mechanism.
• Many patients fail to respond to SSRI treatment (treatment-resistant depression is a substantial clinical category).
• Genetic and imaging studies have not consistently identified primary serotonergic abnormalities in depressed patients at the level the strong monoamine hypothesis would predict.

The current research-grade view: SSRIs and related drugs work for some patients through mechanisms that include monoaminergic effects but extend through downstream neuroplastic adaptations (BDNF upregulation, hippocampal neurogenesis-related changes in some models). The monoamine hypothesis was partially right; the strict form is not adequate.

The inflammatory hypothesis — more recent — frames depression as involving chronic low-grade inflammation, with pro-inflammatory cytokines (IL-6, TNF-α, CRP) elevated in some depressed patients, particularly those with treatment resistance, somatic symptom predominance, and certain medical comorbidities. Cytokine challenge (interferon-α for hepatitis treatment) produces depression-like syndromes in some patients. Inflammation-targeting interventions (anti-inflammatory adjuncts in clinical trials) have shown modest effects in inflammation-elevated depression subgroups [43]. The hypothesis is well-supported for some depressed patients and not for others — consistent with the broader recognition that depression is etiologically heterogeneous.

The glutamate / ketamine paradigm — emerged from the early-2000s observation that subanesthetic ketamine (an NMDA receptor antagonist) produces rapid (hours) antidepressant effects, including in treatment-resistant depression. The finding broke the monoamine framework (the antidepressant effect is too fast to involve weeks of SSRI-style adaptation) and shifted attention to glutamatergic mechanisms [44][45]. Subsequent work has identified ketamine's metabolite (R,S-hydroxynorketamine) as contributing to the antidepressant effect, the downstream AMPAR-trafficking and BDNF/TrkB cascade as part of the mechanism, and the rapid synaptic restoration (rather than monoaminergic adjustment) as a candidate substrate. The clinical translation includes intravenous racemic ketamine and the FDA-approved intranasal esketamine for treatment-resistant depression [46].

HPA dysregulation — A subset of depressed patients show elevated cortisol, blunted dexamethasone suppression, flattened diurnal rhythm, and reduced GR sensitivity. The HPA findings overlap with those in chronic stress and in PTSD. Mechanisms by which HPA dysregulation contributes to depression include hippocampal damage from chronic glucocorticoid exposure, GR-mediated transcriptional changes in mood-relevant circuits, and the metabolic effects of chronic cortisol (the intersection Lesson 3 returns to below) [47].

The contemporary research view holds all four frameworks: depression is a heterogeneous syndrome with multiple contributing mechanisms, and different patients likely have different combinations of monoaminergic, inflammatory, glutamatergic, and HPA contributions. Treatment heterogeneity (some patients respond to SSRIs, some to ketamine, some to behavioral interventions, some to combination approaches) is consistent with mechanistic heterogeneity. The Bachelor's-level reading discipline is to recognize that "the cause of depression" framings, in any direction, oversimplify what is genuinely a multi-mechanism research area.

Anxiety Neurocircuitry

Anxiety disorders — generalized anxiety, panic, social anxiety, specific phobia, separation anxiety, and the formally-related conditions like OCD and PTSD — engage overlapping but distinguishable neural circuits. The Bachelor's-level survey:

• Amygdala — Central to fear conditioning, threat detection, and rapid behavioral response. The basolateral amygdala (BLA) integrates sensory and contextual information; central amygdala (CeA) outputs to autonomic, neuroendocrine, and behavioral systems. Imaging in anxiety disorders consistently shows elevated amygdala reactivity to threat-relevant stimuli [48].
• Bed Nucleus of the Stria Terminalis (BNST) — An extended-amygdala structure increasingly implicated in sustained threat anticipation and contextual fear, distinct from the amygdala's rapid phasic responses. The amygdala/BNST distinction (Walker, Davis, and colleagues) frames the amygdala as handling rapid imminent threat and BNST as handling sustained or ambiguous threat — relevant to generalized anxiety and panic vulnerability.
• Hippocampus — Contextual learning and the modulation of fear responses by context. Hippocampal dysfunction contributes to over-generalization of fear from a specific context to many contexts (a feature of PTSD and generalized anxiety).
• Insula — Interoceptive awareness and visceral salience. Anterior insula activity scales with subjective anxiety in many imaging studies.
• Prefrontal regulation — Ventromedial and dorsolateral PFC provide top-down regulation of amygdala reactivity. Reduced PFC regulation is observed in many anxiety conditions and is a candidate target of exposure-based and cognitive therapies, which appear to operate in part through strengthening PFC-amygdala inhibitory connectivity.

The pharmacology of anxiety treatment overlaps substantially with depression treatment (SSRIs are first-line for most anxiety disorders), reflecting the substantial co-occurrence and shared mechanisms. Benzodiazepines, acting through GABA-A potentiation, provide rapid symptomatic relief but carry dependence risk with chronic use. Newer agents (SNRIs, pregabalin for some conditions) and behavioral therapies (CBT, exposure-based therapies) provide additional options. The pre-clinical relevance for this chapter is that anxiety is a tractable research area in mechanistic terms — the circuits are accessible, the behavioral readouts are replicable, and the gap to clinical translation is closer than in many psychiatric conditions.

The HPA-Metabolic Intersection: Cross-Reference to Coach Food Bachelor's

The first Bachelor's-tier intra-tier lateral reference: chronic HPA dysregulation has direct metabolic consequences that connect Coach Brain Bachelor's Lesson 3 with Coach Food Bachelor's Lesson 2 (energy regulation and homeostasis).

Cortisol's peripheral metabolic effects are well-characterized. Acute cortisol elevation promotes:

• Hepatic gluconeogenesis (increased glucose-6-phosphatase, PEP carboxykinase expression)
• Peripheral lipolysis (free fatty acid mobilization)
• Reduced glucose uptake in muscle and adipose
• Modulation of insulin secretion

These are adaptive responses to acute stress, providing energy substrate for fight-or-flight demand. Sustained or chronic cortisol elevation, however, produces persistent metabolic effects that contribute to:

• Insulin resistance — Chronic glucocorticoid exposure impairs insulin signaling at multiple steps. In Cushing's syndrome (pathological hypercortisolism), insulin resistance is among the most consistent metabolic features.
• Visceral adipose deposition — Glucocorticoids preferentially promote visceral over subcutaneous fat accumulation. Cortisol-elevated states (Cushing's, chronic stress, exogenous glucocorticoid therapy) consistently show increased visceral adipose tissue and the cardiometabolic risk that accompanies it.
• Ectopic lipid — Coach Food Bachelor's Lesson 4 walked through ectopic lipid as the molecular spine of metabolic syndrome (DAG accumulation → PKCε/θ activation → IRS serine phosphorylation → insulin resistance). Chronic cortisol elevation contributes to this pathway through visceral lipid deposition and altered substrate flux.
• HPA dysregulation in metabolic syndrome — Reciprocally, metabolic syndrome and visceral obesity are associated with altered HPA function (flattened diurnal cortisol, blunted post-stress recovery), suggesting the relationship runs in both directions.

The integration: chronic stress and chronic metabolic dysfunction are not separate problems. The HPA axis at the depth Coach Brain teaches and the energy regulation circuit at the depth Coach Food teaches are the same system, observed from different angles. Stress-induced metabolic effects and metabolism-induced stress-axis effects coexist and reinforce each other in many real clinical contexts (chronic-stress depression with weight gain; metabolic syndrome with sleep disruption and mood symptoms; the bidirectional relationships in burnout populations).

For pre-health students, this intersection is the kind of cross-system integration that distinguishes upper-division neuroscience and metabolic medicine from lower-division survey treatments. The clinical relevance: depression and metabolic syndrome co-occur at rates substantially above chance; treatment of one without addressing the other often produces incomplete results; integrated care models that handle both are increasingly recognized as the appropriate clinical response. The chapter does not prescribe such care; it identifies the biology that informs it.

Lesson Check

1. Trace the HPA axis from PVN through anterior pituitary to adrenal cortex, identifying the hormones and receptors at each step.
2. Distinguish glucocorticoid receptor (GR) and mineralocorticoid receptor (MR) function, and identify the regional differences in their expression that produce distinct functional consequences.
3. Describe the principal brain-morphology consequences of chronic allostatic load — hippocampus, PFC, amygdala — and identify the functional implications.
4. Walk the four principal frameworks of depression neuroscience (monoamine, inflammatory, glutamate/ketamine, HPA dysregulation). What does each contribute, and where does each have limits?
5. Identify three principal nodes of the anxiety neurocircuitry and the role of each.
6. Articulate the HPA-metabolic intersection. How does chronic cortisol elevation contribute to insulin resistance, and how does metabolic syndrome alter HPA function?

---

Lesson 4: Executive Function, Decision Neuroscience, and Reward

Learning Objectives

By the end of this lesson, you will be able to:

• Describe persistent activity in dorsolateral prefrontal cortex during working memory and identify the principal neuroimaging and electrophysiological evidence for it
• Walk the fronto-striatal loops (Alexander, DeLong, Strick 1986) and their functional specializations: motor, oculomotor, dorsolateral prefrontal, lateral orbitofrontal, anterior cingulate
• Articulate the unification of Schultz's dopamine prediction error with Sutton and Barto's temporal-difference learning that grounded modern computational neuroscience
• Distinguish Berridge's wanting (incentive salience) from liking (hedonic pleasure) and discuss the implications for addiction
• Identify Nestler's ΔFosB as a candidate molecular substrate for the persistence of addiction
• Describe the prefrontal-limbic balance in self-control and emotion regulation at the level of imaging and neuropharmacology

Key Terms

Persistent Activity and Working Memory in Prefrontal Cortex

In 1971, Joaquin Fuster and colleagues recorded from monkey prefrontal cortex during delayed-response tasks and observed neurons whose firing was elevated during the delay period — the interval between cue presentation and required response — and selective for the cued stimulus or location. This persistent activity became the foundation of the prefrontal-working-memory framework that Patricia Goldman-Rakic and colleagues developed through the 1980s and 1990s [49].

Goldman-Rakic's research mapped the dorsolateral prefrontal cortex (DLPFC) as a working-memory substrate with several specific properties [50]:

• Domain-specific organization — Different DLPFC subregions appear to handle different content (spatial vs. object vs. verbal working memory), though the topography is less strict than the original maps suggested.
• Recurrent excitation — Local recurrent circuitry (pyramidal cell collaterals among layer 3 neurons) is thought to support sustained activity through positive feedback, with inhibitory interneurons (parvalbumin-expressing fast-spiking cells) shaping the dynamics.
• Dopaminergic modulation — DLPFC pyramidal cells express D1 dopamine receptors prominently; phasic D1 activation gates entry of information into working-memory representations; tonic D1 activation supports maintenance. The Cools-Robbins inverted-U relationship — optimal performance at intermediate D1 stimulation, impairment at either too-low or too-high — has been a robust finding.
• Lateralization and integration — Working memory engages bilateral DLPFC with content-specific lateralization, posterior parietal cortex (sensory representation), and basal ganglia (attentional gating).

Mark D'Esposito and colleagues extended the framework to human cognitive neuroscience using fMRI, demonstrating sustained DLPFC and parietal activity during the delay period of working-memory tasks in humans and characterizing the relationships between activity, performance, and capacity limits [51]. The combined animal-electrophysiology and human-neuroimaging literatures converged on a reasonably consistent picture: DLPFC is a substantial substrate of working memory through persistent activity in recurrent local circuits, modulated by dopamine, with inputs from posterior cortex providing content.

Recent work has complicated the pure persistent-activity model. Some studies have observed activity-silent working memory — periods in which delay-period firing is low but the memory representation is preserved in synaptic short-term plasticity that can be re-read by subsequent input. The current research-grade view holds both: persistent activity is one mechanism; activity-silent dynamics are another; the brain likely uses different mechanisms in different conditions and for different content [52].

Fronto-Striatal Loops: The Alexander Framework

In 1986, Garrett Alexander, Mahlon DeLong, and Peter Strick published a foundational Annual Review of Neuroscience paper proposing that the basal ganglia and frontal cortex are organized as a set of parallel circuits — fronto-striatal loops — each linking a specific frontal cortical region to a specific striatal target and returning through pallidum and thalamus to the originating cortex [53]. The five loops identified:

1. Motor loop — Primary motor cortex / supplementary motor area → putamen → globus pallidus → ventrolateral thalamus → motor cortex. Motor sequencing and execution.
2. Oculomotor loop — Frontal eye fields → caudate nucleus → globus pallidus / substantia nigra reticulata → ventral anterior thalamus → frontal eye fields. Saccadic eye movements.
3. Dorsolateral prefrontal loop — DLPFC → dorsolateral head of caudate → globus pallidus → ventral anterior thalamus → DLPFC. Working memory and executive function.
4. Lateral orbitofrontal loop — Lateral orbitofrontal cortex → ventromedial caudate → globus pallidus → thalamus → lateral OFC. Behavioral inhibition and response selection.
5. Anterior cingulate loop — Anterior cingulate cortex → ventral striatum (nucleus accumbens) → ventral pallidum → thalamus → ACC. Motivation and limbic processing.

Each loop carries a direct pathway (cortex → striatum → globus pallidus internal / SNr → thalamus → cortex; net excitatory effect on cortex) and an indirect pathway (with an additional inhibitory step through globus pallidus external; net inhibitory effect on cortex). The balance between direct and indirect pathways within each loop shapes the loop's net output. Dopaminergic input from substantia nigra (motor and oculomotor loops) and VTA (prefrontal and limbic loops) modulates the D1-direct and D2-indirect pathways differently, biasing loop output.

The clinical relevance is extensive. Parkinson's disease — degeneration of nigrostriatal dopamine neurons — produces motor symptoms primarily through disruption of the motor loop. Huntington's disease — degeneration of striatal medium spiny neurons — produces motor and cognitive symptoms through disruption across multiple loops. OCD has been mapped substantially onto orbitofrontal-striatal-thalamic circuit dysfunction, with deep brain stimulation of internal capsule / nucleus accumbens regions effective in treatment-resistant cases. ADHD involves frontostriatal-circuit alterations that have been mapped in neuroimaging studies, with stimulant medications acting through dopaminergic enhancement at these circuit levels.

The Alexander framework is one of the few neuroscience papers from the 1980s that has held up across three decades of subsequent research. It is required reading for any pre-clinical neurology, neuropsychiatry, or movement-disorders pathway.

Schultz and the Reinforcement Learning Unification

Wolfram Schultz's recordings from midbrain dopamine neurons in awake-behaving monkeys, beginning in the 1980s and continuing through the present, established that phasic dopamine signals reward prediction error (RPE): the difference between actual and expected reward, with positive RPE encoded by phasic increases and negative RPE encoded by phasic decreases [54]. The findings (summarized in Schultz 1998 Journal of Neurophysiology and subsequent reviews):

• Unexpected reward — Sharp phasic increase at reward delivery.
• Cue predicting reward — Across learning, the phasic response shifts to the predictive cue; the reward itself, now predicted, produces no phasic response.
• Predicted reward omitted — A phasic decrease (dip below baseline) at the time the reward should have arrived. This is the negative-prediction-error signal.
• Higher-than-expected reward — Phasic increase scaled to the magnitude of the unexpected component.

The signal is not "reward" — it is prediction error, the discrepancy between expected and actual outcome. Schultz's empirical characterization aligned remarkably with a contemporaneous theoretical development in reinforcement learning [55].

Andrew Barto, Richard Sutton, and colleagues had developed temporal-difference (TD) learning — a computational algorithm in which an agent learns the value of states by updating estimates based on prediction error: V(s_t) ← V(s_t) + α[r_t + γV(s_{t+1}) − V(s_t)]. The bracketed term is the TD error, and its algorithmic role is to drive learning of state values.

The unification: the TD error in Sutton and Barto's algorithm has the exact mathematical form of Schultz's empirical dopamine signal. The match was striking — a computational model developed for engineering problems matched the firing of a specific population of neurons in a specific brain region with high precision. The 1996 Schultz, Dayan, Montague Science paper formalized the correspondence and grounded modern computational neuroscience: a specific neural population implements a specific algorithm with a specific computational function in reinforcement learning [56].

Subsequent work has elaborated the picture. Dopamine signals encode RPE for many forms of reward (food, fluid, sex, music in some studies, social reward in others), with some controversy about whether all dopamine neurons signal the same value-based RPE or whether subpopulations signal somewhat different quantities (salience, novelty, aversive prediction error). Wolfram Schultz received a share of the 2017 Brain Prize for this body of work.

The Bachelor's-level take-away is that the brain's reward learning is computationally well-characterized: specific neurons compute specific quantities that drive learning of specific value functions through specific synaptic plasticity. This is one of the more complete cellular-to-computational mappings in neuroscience and a model for the kind of integration that upper-division work points toward.

Berridge: Wanting versus Liking

Kent Berridge and colleagues at Michigan developed a complementary framework that complicates the simple "dopamine = reward" picture in important ways. The wanting/liking distinction [57][58]:

• Wanting (incentive salience) — The motivational pull of a cue that predicts reward; the urge to obtain the cued outcome. Mediated by mesolimbic dopamine (VTA → nucleus accumbens, especially the medial shell).
• Liking (hedonic pleasure) — The actual pleasure produced by consummatory experience. Mediated by hedonic hotspots — small subregions of nucleus accumbens shell, ventral pallidum, and brainstem parabrachial nucleus — that signal through opioid and endocannabinoid receptors.

The framework rests on dissociations:

• Dopamine-depleted rats still show normal "liking" reactions to sucrose (orofacial hedonic reactions characteristic of pleasurable taste) but fail to seek out sucrose voluntarily — they want it less but like it the same.
• Conversely, electrical or chemical stimulation of nucleus accumbens shell elevates "wanting" (food-seeking effort) without elevating "liking" (orofacial responses to taste).
• Pharmacological dissociations consistently separate the two functions onto different neurochemical substrates.

The implications for addiction are substantial. The incentive sensitization theory (Robinson and Berridge 1993, updated 2008) proposes that repeated drug use progressively sensitizes the wanting system without proportionally sensitizing the liking system. The result: addicted individuals experience strong urges to use drugs even when the actual pleasure from use has diminished or disappeared. This pattern matches the clinical experience of late-stage addiction more accurately than a simple hedonic-treadmill model.

The framework also speaks to non-drug compulsions. Variable-reinforcement-schedule technologies (social media notifications, slot-machine-style apps, certain games) engage dopaminergic prediction-error systems through unpredictable rewards, generating strong wanting without proportional liking. The biology that evolved to drive food-seeking and pair-bonding can be engaged by engineered environmental contingencies that the brain did not evolve to handle. The Bachelor's-level reading: the neuroscience of motivation explains why some modern environments are uncomfortable to live in, not as a moral commentary but as a mechanistic observation.

Nestler and ΔFosB: A Candidate Persistence Molecule

Eric Nestler's laboratory at Mount Sinai has worked for two decades on the molecular substrates of addiction's persistence — the property that distinguishes addiction from ordinary preferences and makes recovery so difficult. The candidate substrate Nestler proposed is ΔFosB, a truncated FosB transcription-factor variant with several unusual properties [59]:

• Unusual stability — Most immediate-early-gene products (c-Fos, FosB, JunB) are degraded within hours. ΔFosB has a half-life of weeks, allowing it to accumulate in nucleus accumbens neurons with repeated exposure to drugs of abuse or to natural rewards encountered at high frequency.
• Transcriptional cascade — Accumulated ΔFosB drives transcription of genes including AMPAR subunits (GluA2), CDK5, and others. The transcriptional program shifts the molecular state of NAc neurons in directions that elevate sensitivity to reward cues and weaken cognitive control.
• Necessity demonstrated — Inducible viral overexpression of ΔFosB in nucleus accumbens enhances drug-induced behavioral sensitization and reward-related behaviors. Reduction of ΔFosB (through transgenic models or pharmacological approaches) reduces these behaviors.

ΔFosB is not the sole molecular substrate of addiction. The chronic neuroplastic changes in addiction span structural changes in dendritic spines, alterations in AMPAR composition (calcium-permeable GluA2-lacking AMPARs in some addiction stages), modifications of glutamatergic transmission, alterations in glial function, and others. The molecular landscape is multilayered. ΔFosB is one well-characterized component that helps explain persistence; it is not a comprehensive theory of addiction.

The clinical relevance of the molecular work: it grounds the recognition that addiction is a brain disease in the sense that the brain undergoes durable molecular and structural changes during the disease process — without erasing the reality that addiction is also a behavioral and social condition shaped by environment, choice, and life context. The molecular work informs clinical conversation; it does not replace it.

Prefrontal-Limbic Balance and Self-Control

The neuroscience of self-control and emotion regulation maps substantially onto the prefrontal-limbic balance: prefrontal regions (particularly ventromedial and dorsolateral PFC) regulate limbic regions (amygdala, ventral striatum) through largely inhibitory projections. The balance produces:

• Cognitive emotion regulation — Reappraisal, suppression, and other regulatory strategies engage DLPFC and ventrolateral PFC, modulating amygdala reactivity. The Ochsner-Gross framework maps the cognitive-control literature onto this architecture.
• Delay discounting — The willingness to wait for a larger later reward over an immediate smaller one engages medial PFC and is impaired by PFC dysfunction (lesions, age-related decline, addiction).
• Effortful response inhibition — The Go/No-Go and stop-signal tasks engage right inferior frontal gyrus and pre-SMA, with the cognitive control engaging these regions over basal ganglia indirect-pathway connections.

The development of self-control across adolescence and young adulthood corresponds to the protracted maturation of prefrontal cortex, which continues through the mid-twenties. The mismatch between earlier-maturing limbic systems and later-maturing prefrontal regulation is the standard neurodevelopmental account of adolescent risk-taking propensity. The clinical relevance: many of the conditions associated with impaired self-control — ADHD, addiction, certain personality traits — engage variants of the prefrontal-limbic balance, and treatment approaches (medication, behavioral therapy, mindfulness-based interventions) can be understood in part as interventions on this balance.

Lesson Check

1. Describe persistent activity in DLPFC during working memory and identify the principal neuroimaging and electrophysiological evidence for the framework. What complicates the simple persistent-activity model?
2. Walk the Alexander fronto-striatal loop framework. Identify the five loops and their functional specializations.
3. Explain Schultz's reward prediction error finding and articulate how it unified with Sutton and Barto's temporal-difference learning algorithm. What did the unification establish about brain function?
4. Distinguish Berridge's wanting (incentive salience) from liking (hedonic pleasure). What dissociations support the distinction? What does the framework explain about late-stage addiction?
5. Describe Nestler's ΔFosB proposal as a candidate molecular substrate for addiction persistence. What gives ΔFosB its unusual properties?
6. Identify the principal architecture of prefrontal-limbic balance in self-control and emotion regulation. Why does the protracted PFC maturation through the mid-twenties shape adolescent risk-taking?

---

Lesson 5: Research Methods in Neuroscience

Learning Objectives

By the end of this lesson, you will be able to:

• Describe the BOLD signal at signal-detection depth, including its hemodynamic basis, spatial and temporal resolution, and the inferential chain from BOLD to cognition
• Identify the principal limits of fMRI: the multiple-comparisons problem, the dead-salmon paper (Bennett 2009), false-positive rates in cluster-based inference (Eklund 2016), and small-sample power issues
• Distinguish single-unit recording, multi-unit recording, patch clamp, calcium imaging, and voltage imaging at the level of what each measures and the resolution-throughput tradeoffs
• Explain optogenetics (Deisseroth and colleagues, 2005 forward) and chemogenetics (Roth and colleagues, DREADDs) and identify what these tools enabled methodologically
• Articulate the reproducibility crisis in neuroscience and psychology and the principal reforms that have followed (pre-registration, larger samples, transparent reporting)
• Apply the five-point evaluation framework to neuroscience claims

Key Terms

The BOLD Signal: Ogawa 1990 as Foundational Anchor

The foundational anchor for this chapter is Seiji Ogawa, Tso-Ming Lee, Alan Kay, and David Tank's 1990 Proceedings of the National Academy of Sciences paper Brain magnetic resonance imaging with contrast dependent on blood oxygenation, which described what became the BOLD signal — the basis of modern functional magnetic resonance imaging [60].

Ogawa and colleagues at AT&T Bell Laboratories made the foundational observation by working with deoxyhemoglobin's magnetic properties. Oxyhemoglobin is diamagnetic; deoxyhemoglobin is paramagnetic. The presence of paramagnetic deoxyhemoglobin in cerebral capillaries and venules produces local magnetic-field inhomogeneities that shorten the apparent transverse relaxation time (T2*) of nearby protons. Brain regions with elevated neural activity show a complex hemodynamic response: local oxygen consumption rises, but local blood flow rises more (the "uncoupling" Roy and Sherrington observed in 1890), producing a net decrease in local deoxyhemoglobin concentration. The decreased paramagnetic burden lengthens T2* and produces a signal increase in T2*-weighted images. This is the BOLD signal.

Ogawa's 1990 paper demonstrated the contrast in animals; subsequent human applications (Kwong, Belliveau, and others 1992) extended it to functional human imaging. By the late 1990s, fMRI had displaced the older positron emission tomography (PET) methods for most cognitive-neuroscience research, principally because fMRI uses no radioactive tracer (allowing repeated scanning) and offers better spatial resolution.

What BOLD measures and what it does not:

BOLD measures the local hemodynamic response. It is not a direct measure of neural activity. The hemodynamic response is delayed (peaking 5-6 seconds after the neural event in healthy young adults), spatially blurred (the vascular bed serves regions larger than the underlying neural activation), and modulated by factors beyond neural activity (blood pressure, CO₂, neurovascular coupling integrity, age, medication, caffeine intake). Inferring "the brain region was active" from "the BOLD signal increased" requires assuming that hemodynamic coupling is intact and proportional to neural activity — an assumption usually reasonable in healthy adults but with documented exceptions.

Logothetis and colleagues in the 2000s combined fMRI with simultaneous electrophysiology to map the relationship between BOLD and underlying neural signals [61]. Their findings: BOLD signal correlates most strongly with local field potential (LFP) in the gamma range, which reflects local synaptic activity (inputs and dendritic processing) more than with action-potential output of the recorded region. The implication: fMRI signal can reflect synaptic activity in a region even when output firing is unchanged, and conversely can fail to detect activity that occurs without commensurate gamma-band synaptic activity.

Spatial resolution — Typical fMRI voxel sizes are 2-3 mm cubed, containing approximately 100,000 to 1,000,000 neurons. The cognitive-level inferences fMRI supports are inherently averaged across these neural populations; single-neuron-resolution conclusions are not accessible.

Temporal resolution — The hemodynamic response is inherently slow; typical fMRI sampling at 1-2 second TR is constrained more by the slow hemodynamic blurring than by acquisition speed. Sub-second cognitive events that are clearly resolvable in electrophysiology are blurred together in BOLD.

For Bachelor's-level neuroscience, the methodological literacy includes understanding what BOLD does and does not measure — and reading published fMRI papers accordingly.

The Multiple-Comparisons Problem and the Dead Salmon

A standard fMRI analysis involves performing a statistical test at each of tens of thousands of voxels. Without correction, the false-positive rate at the standard p < 0.05 threshold compounds across voxels: 5% of voxels in a brain volume with no real activation will show "significant" effects by chance. The need for correction is widely recognized, but the specific method and its rigor have been a methodological concern.

Bennett and colleagues' 2009 Journal of Serendipitous and Unexpected Results paper Neural correlates of interspecies perspective taking in the post-mortem Atlantic Salmon (the dead salmon paper) provided a memorable demonstration. The authors placed a dead Atlantic salmon in an MRI scanner, ran a standard cognitive-neuroscience paradigm (the salmon was shown photographs of humans in social situations and asked to consider the human's emotional state), and performed standard fMRI analysis with uncorrected p < 0.001. The result: voxels in the dead salmon's brain showed "significant" task-related activity [62]. The paper, originally a poster presentation at the Human Brain Mapping conference, made a serious methodological point with humor: without appropriate multiple-comparisons correction, fMRI analyses produce "results" even from a dead fish.

Eklund and colleagues' 2016 PNAS paper Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates extended the methodological critique. By running thousands of resting-state datasets (subjects at rest, not performing tasks) through standard cluster-based fMRI inference algorithms, they showed that several widely-used analysis pipelines produced false-positive rates substantially higher than the nominal 5% — in some configurations approaching 70% [63]. The paper prompted substantial revision of fMRI analytical practice, including more conservative cluster thresholds, alternative inference methods (TFCE, FWE correction at the voxel level), and increased use of permutation-based methods.

The combined methodological concerns mean that fMRI claims should be read with awareness of:

• The correction method applied
• The sample size (small-N fMRI studies have substantial concerns about power and false-positive inflation)
• The pre-registration status (post-hoc analytical flexibility — the "garden of forking paths" — is a known source of inflated findings)
• The replication status (single-study findings deserve more skepticism than findings replicated across multiple groups and pipelines)

The discipline of reading fMRI papers is one of the core upper-division literacies.

Electrophysiology and Cellular-Resolution Methods

Beyond fMRI, the neuroscience methods toolkit includes higher-resolution approaches that operate at cellular and subcellular levels [64]:

Single-unit recording — A glass or metal microelectrode positioned near a single neuron records action potentials of that neuron. Yields exquisite resolution (single-cell, sub-millisecond) but limited throughput (one or a few cells per electrode penetration). The foundational methodology for Schultz's dopamine prediction error work, O'Keefe's place cell discovery, and Goldman-Rakic's persistent-activity studies.

Multi-unit / multi-electrode recording — Arrays of electrodes (Utah arrays, polytrodes, Neuropixels probes) record from tens to thousands of neurons simultaneously. The Neuropixels probe, introduced in 2017, can record from ~300 neurons across a 4-mm shank — a substantial increase in throughput that has enabled population-coding analyses at scales previously infeasible.

Patch clamp — A glass micropipette forms a high-resistance seal with the cell membrane and provides direct electrical access to the cell interior. Yields the highest signal-to-noise electrophysiology and supports both passive (voltage) and active (current injection) protocols. Used extensively for synaptic-physiology studies, ion-channel characterization, and detailed analysis of single-neuron computation. Sakmann and Neher received the 1991 Nobel Prize for development of the patch clamp.

Calcium imaging — Fluorescent calcium indicators (synthetic dyes like Oregon Green BAPTA-1; genetically-encoded indicators like GCaMP) visualize calcium transients in many neurons simultaneously. Calcium reflects neural activity indirectly (action potentials produce calcium transients), with temporal resolution limited by the indicator's kinetics and the calcium-buffering machinery. Two-photon microscopy enables sub-cellular resolution in vivo. Calcium imaging has been transformative for circuit-level neuroscience over the last 15 years.

Voltage imaging — Newer fluorescent voltage indicators measure membrane voltage more directly than calcium imaging. Temporal resolution can approach action-potential timescales with the best modern indicators. The methodology is still maturing relative to calcium imaging.

EEG and MEG — Electroencephalography measures voltage fluctuations at the scalp arising from synchronous postsynaptic potentials in superficial cortical neurons. Magnetoencephalography measures the associated magnetic fields. Both offer millisecond-scale temporal resolution but limited spatial resolution (centimeter scale in EEG, somewhat better in MEG). Source-localization analyses attempt to infer underlying cortical generators. EEG is widely used in clinical neurology (epilepsy, sleep studies) and in cognitive neuroscience (event-related potentials, oscillation analyses).

The Bachelor's-level point is that different methods answer different questions. Single-unit recording is appropriate for studying how individual neurons compute; fMRI is appropriate for whole-brain mapping of cognitive operations in humans; calcium imaging is appropriate for circuit-level dynamics in genetically-defined populations. No single method is the right method for all questions, and integrated multi-method studies are increasingly the norm in serious neuroscience research.

Optogenetics and Chemogenetics: The Methodological Revolution

In 2005, Edward Boyden, Karl Deisseroth, and colleagues at Stanford reported that channelrhodopsin-2 — a light-gated cation channel from the alga Chlamydomonas reinhardtii — could be expressed in mammalian neurons and used to evoke action potentials with millisecond-precision light pulses [65]. The methodology, named optogenetics, transformed neuroscience by providing a tool for causal manipulation of genetically-defined neural populations at temporal precision matching neural activity.

The optogenetic toolkit expanded rapidly. Halorhodopsin and archaerhodopsin provide inhibitory functions (light-driven hyperpolarization). Step-function opsins (ChR2 variants with prolonged open states) extend activation timescales. Red-shifted opsins (ChrimsonR) enable deeper-tissue penetration. Variant kinetics support different experimental protocols. The methodology is now standard in circuit-neuroscience laboratories worldwide.

In parallel, Bryan Roth and colleagues developed Designer Receptors Exclusively Activated by Designer Drugs (DREADDs) — engineered GPCRs that respond to an otherwise inert ligand (clozapine-N-oxide, CNO) [66]. The DREADD toolkit includes activating receptors (hM3Dq, signaling through Gq → PLC → cellular depolarization) and inhibitory receptors (hM4Di, signaling through Gi/o → hyperpolarization). DREADDs operate on slower timescales than optogenetics (minutes to hours) but offer the advantage of systemic ligand administration without requiring implanted optical fibers.

What these tools enabled methodologically:

• Causal demonstration of necessity and sufficiency — Activating or silencing a specific population of neurons and observing the behavioral consequence allows direct causal inference, beyond the correlational evidence available from observation alone. This is the methodology that enabled engram-tagging demonstrations (Lesson 2).
• Circuit dissection — Projection-specific manipulations (activating only the neurons in region A that project to region B) allow tests of specific anatomical pathways' contributions to behavior.
• Temporal precision — Optogenetic millisecond resolution enables tests of when in a behavioral sequence a circuit contributes, not merely whether.
• Population specificity — Cre-dependent expression strategies enable manipulation of cell-type-specific populations (e.g., parvalbumin interneurons specifically; D1- versus D2-expressing medium spiny neurons specifically).

The clinical translation is partial. Optogenetic vision restoration in retinal degeneration is in clinical trials (Sahel and colleagues reported preliminary results in 2021). Deep brain stimulation, while not optogenetic, has been informed by circuit-level neuroscience that optogenetic studies illuminated. Direct optogenetic clinical applications beyond vision remain at early stages; the methodology is, however, foundational for the basic-science work that informs future clinical interventions.

The Reproducibility Crisis in Neuroscience and Psychology

In 2015, the Open Science Collaboration published an attempt to replicate 100 psychology studies, finding that only ~36% produced statistically significant results in the same direction as the original, with effect sizes roughly half the original on average [67]. The paper crystallized concerns that had been accumulating across multiple fields: a substantial fraction of published findings appeared not to replicate when tested directly.

The reproducibility-crisis literature has documented several contributing factors [68][69]:

• Underpowered studies — Small samples relative to true effect sizes produce inflated effect-size estimates conditional on detection (the winner's curse). Published estimates from small studies overstate the true effect.
• Publication bias — Negative findings are systematically less likely to be published. Meta-analyses that synthesize only published findings overstate true effects.
• Garden of forking paths — Post-hoc analytic flexibility (choosing among many possible exposure definitions, outcome definitions, subgroup analyses, statistical models, and reporting rules) generates many opportunities for chance findings to be reported as discoveries.
• Multiple comparisons inadequately handled — As fMRI's dead-salmon example illustrates, unaddressed multiple-comparisons issues produce false-positive inflation.
• Direct sample-size requirements — Effect sizes typical of psychology and neuroscience often require samples larger than the discipline has historically used to detect reliably.

The field's response has included:

• Pre-registration — Many journals now accept and even require pre-registered analytic plans, constraining post-hoc flexibility.
• Registered reports — Articles peer-reviewed based on design and analysis plan before data collection, with publication committed regardless of result outcome. The format aligns incentives away from significance-chasing.
• Larger samples — Recognition of historical underpowering has driven calls for larger samples and prospective power calculations.
• Transparent reporting — Standardized checklists (CONSORT for clinical trials, ARRIVE for animal studies, COBIDAS for neuroimaging) encourage complete reporting of methods.
• Open data and code — Increasing availability of data and analysis code permits independent replication and re-analysis.
• Multi-site replication efforts — Coordinated multi-laboratory replications (the Many Labs projects in psychology, ENIGMA in neuroimaging) provide direct evidence on replicability.

The Bachelor's-level point is not that psychology and neuroscience are unreliable as fields; it is that consumers of the literature should hold individual findings more loosely than the publication conventions of the field have historically suggested. Synthesis across studies — particularly across pre-registered studies, registered reports, and multi-site replications — provides more reliable knowledge than reaction to individual reports. The discipline of reading the literature this way is a core upper-division literacy.

The Five-Point Evaluation Framework Applied to Neuroscience

The five-point framework introduced in earlier chapters extends straightforwardly to neuroscience claims:

1. What is the proposed mechanism, and is it biologically plausible? A claimed brain-behavior relationship with no plausible neural mechanism is more likely a false positive than a real signal. A claim grounded in known circuitry is not necessarily right but is more credible.

2. What is the design and methodology that produced the strongest evidence? Correlational fMRI in humans is hypothesis-generating. Single-unit recording in awake-behaving primates is more direct evidence for neural correlates. Causal manipulation through optogenetics or chemogenetics is decisive for necessity and sufficiency claims. Multiple converging methods provide the strongest grounding.

3. What is the effect size and sample size? A small fMRI study with ten participants reporting a brain-behavior correlation has high false-positive risk. A multi-thousand-participant study with appropriate corrections has different epistemic standing. Effect sizes from underpowered studies are systematically inflated.

4. Has the finding replicated, and across what populations, species, and methods? Single-laboratory findings in a single species should be held loosely. Findings replicated across multiple laboratories, multiple species, and multiple methods carry more weight.

5. What does the claim mean for clinical practice or personal application, and is the translation appropriate? A mouse optogenetic finding does not automatically translate to a recommendation for human behavior or clinical practice. The translation from research to practice requires clinical context, patient-specific factors, and judgment that the research alone cannot provide.

Most popular neuroscience claims — "the brain region that does X," "the chemical that makes you Y," "the neuroscience of Z" — fail at point 2 (single-method, often single-study correlational evidence) or point 5 (over-translation from research to practical recommendation). Pre-medical and neuroscience-major students benefit from training that flags these failures by structure.

The Turtle's Integrator Position at Bachelor's: Receiver, Deepened

A closing structural point. At Associates depth, the Turtle's integrator position was named as receiver — the brain receives inputs from every other modality through identifiable neural mechanisms.

At Bachelor's depth, the receiver position deepens at circuit-and-molecular level. The brain is not abstractly "the integrator"; it is a specific computational system that receives substrate (Food: amino acids supporting mTORC1-driven protein synthesis, glucose supporting glycolysis, lipid components of myelin, vitamin D modulating neural function), internal environment (Water: the electrolyte composition that gates neural excitability and the chemical milieu of synaptic function), synchronizer timing (Light: SCN entrainment and the temporal organization of brain states), consolidation cycles (Sleep: the temporal pass that closes the day's plasticity loops through SWS-coupled hippocampal-cortical replay and glymphatic clearance), active-output proprioception (Move: BDNF and exercise-induced plasticity that Lesson 2 referenced), system-probe input (Cold: cold-induced sympathetic and norepinephrine signaling that engages arousal circuits), adaptive-load input (Hot: heat-stress signaling that shapes neural function under thermal challenge), and interface control (Breath: voluntary-autonomic regulation that engages prefrontal-brainstem connectivity).

The receiver position is the molecular and circuit-level ground on which the integration happens. Specific neural substrates handle specific inputs — the gut-vagus pathway for some interoceptive signals, the SCN for light, the brainstem chemoreceptors for breath-related blood-gas information, the proprioceptive thalamocortical pathways for movement, and so on. The brain does not integrate abstractly; it integrates through identifiable cellular machinery.

Whether a Bachelor's-tier specialization produces a distinct integrator position beyond the ten named at Associates is, as Food Bachelor's noted, an open question. The candidate from upper-division depth is something like regulator or closed-loop control system, capturing the homeostatic-control framing that Lesson 3 stress neurobiology and the HPA-metabolic intersection introduced. The Turtle does not commit to a new position here; the ten-position ontology is real and currently complete. Subsequent Bachelor's chapters across the remaining Coaches will inform whether the ontology needs to expand or whether the existing ten suffice when deepened.

Mental-Health Vigilance at Pre-Health Depth

A closing word, before the lesson ends. The Bachelor's-level audience for this chapter includes neuroscience majors, pre-medical, pre-clinical psychology, cognitive science, and other pathways into mental-health-adjacent careers. These programs carry elevated familiarity with the conditions Lesson 3 named (depression, anxiety, ADHD, addiction, the schizophrenia spectrum) — which is appropriate for the pathway but also means the late-teens / early-twenties first-onset window for several of these conditions coincides with the time many students in these programs are reading this material.

The Turtle holds the protective frame from every prior tier:

• The neuroscience in this chapter describes brain biology and pathophysiology at research-grade depth. It is not a diagnostic manual.
• Recognition of patterns in oneself or others is part of life literacy. Diagnosis and treatment are the work of licensed clinicians.
• The late-teens / early-twenties window is real. Many people experience first onset of depression, anxiety, addiction, or other conditions in this window. Asking for help is among the most adult things you can do.

Verified resources (current at this chapter's writing; re-verify before publication):

• 988 Suicide and Crisis Lifeline — call or text 988, 24/7
• Crisis Text Line — text HOME to 741741, 24/7
• National Alliance for Eating Disorders helpline — (866) 662-1235, weekdays 9 a.m.-7 p.m. Eastern, for eating-disorder-adjacent concerns specifically

Important note: The older NEDA helpline (1-800-931-2237) was discontinued in 2023 and is no longer functional. Do not cite it.

College health centers, college counseling centers, primary care providers, and the campus support resources specific to your program are also real. The neuroscience here is the work of the chapter. Asking for help when you need it is the work of life.

Lesson Check

1. Describe the BOLD signal at the level of its hemodynamic and electromagnetic basis. What does fMRI measure, and what does it not measure?
2. Identify the multiple-comparisons problem in fMRI and describe what the dead-salmon paper and the Eklund 2016 analysis added to methodological awareness.
3. Distinguish single-unit recording, calcium imaging, and EEG/MEG at the level of what each measures and the resolution-throughput tradeoffs.
4. Explain optogenetics and chemogenetics. What methodological capability did these tools add to neuroscience that prior methods did not provide?
5. Articulate the reproducibility crisis in neuroscience and psychology. Identify three principal contributors and three principal reforms.
6. Apply the five-point evaluation framework to a recent neuroscience claim of your choosing.

---

End-of-Chapter Activity

Activity: Read a Primary Neuroscience Paper and Evaluate It Against the Methodological Frame

This activity applies the methodological consciousness Lesson 5 named to a concrete research artifact. The goal is to demonstrate upper-division literacy in reading a neuroscience paper.

Step 1 — Select a paper. Pick a primary neuroscience research paper published in the last five years in a major journal (Nature, Science, Neuron, Cell, Nature Neuroscience, Journal of Neuroscience, Cerebral Cortex, NeuroImage, Biological Psychiatry, or similar). Pick one whose topic interests you. Note title, authors, journal, year.

Step 2 — Identify the design and methodology. What did the authors do? Specify the species (rodent, primate, human), the methodology (fMRI, single-unit, calcium imaging, optogenetics, electrophysiology, behavior), the cohort or sample size, and the principal analytic approach. Identify which of Lesson 5's methodological frameworks applies.

Step 3 — Specify the methodological strengths and limits. What does this design allow the authors to claim? What does it not allow? Where are the chronic problems of the method most likely to operate? (For fMRI: multiple comparisons, sample size, BOLD-inferential chain. For animal optogenetics: species generalization to humans. For correlational human studies: causal inference limits.)

Step 4 — Read the effect size in context. What is the magnitude of the reported effect? How does it compare to population variation, measurement noise, and the expected size for studies in this domain? Are confidence intervals or effect-size estimates given, or are reports only at statistical-significance level?

Step 5 — Evaluate the discussion section critically. Does the authors' discussion acknowledge methodological limits appropriately? Are clinical or practical implications stated with appropriate caveats? Does the paper acknowledge alternative explanations and replication context?

Step 6 — Apply the five-point framework. Walk the paper through mechanism plausibility, study design adequacy, effect size in context, replication status, and appropriate translation. Write a one-paragraph summary of what the paper demonstrates and what it does not.

Deliverable. A 1500-2500 word written analysis with citations to the paper and at least three additional context sources. Submit with your name, course identifier, and a one-paragraph reflection on what the exercise taught you about reading neuroscience research.

Optional extension for graduate-school-bound students. Identify a methodologically stronger study addressing the same question, or specify what an ideal study would look like. For pre-medical students: translate the finding into the clinical-conversation language you might use with a patient or family member — descriptive, qualified, with appropriate uncertainty and respect for the clinical conversation the research alone cannot conduct.

---

Vocabulary Review

---

Chapter Quiz

Bachelor's-level quiz. Combination of short-answer mechanistic questions, scenario-based application, and methodological critique. Aim for 3-6 sentences per response; show circuit-and-molecular-level specificity; cite primary literature where appropriate.

1. Describe the Hodgkin-Huxley model as a mathematical object. What did Hodgkin and Huxley experimentally observe in the squid giant axon, and what did the model infer about molecular biology that was confirmed decades later?

2. Walk synaptic transmission from presynaptic depolarization through SNARE-complex-mediated vesicle fusion to postsynaptic receptor activation. Identify the role of synaptotagmin specifically.

3. Compare NMDA, AMPA, and metabotropic glutamate receptors at the level of structure, ionic permeability, and signaling. Why is NMDA called a coincidence detector?

4. Trace the LTP molecular cascade from NMDAR Ca²⁺ influx through CaMKII autophosphorylation, AMPAR trafficking, and CREB phosphorylation. Identify the protein-synthesis-dependence boundary between early- and late-LTP.

5. Describe Eric Kandel's contribution to memory research. Walk the Aplysia long-term sensitization cascade and identify the role of CREB-1 versus CREB-2.

6. Compare Sorrells 2018 and Boldrini 2018 on adult human hippocampal neurogenesis. What did each measure, why might their conclusions differ, and what kind of evidence would resolve the controversy?

7. Describe place cells (O'Keefe) and grid cells (Moser, Moser). What does each cell type contribute to the spatial-memory system, and how does early Alzheimer's pathology relate to this circuit?

8. Trace the HPA axis from PVN through anterior pituitary to adrenal cortex. Distinguish MR and GR function and identify the regional difference in their expression.

9. Describe chronic-stress effects on hippocampal, prefrontal, and amygdala morphology. What are the functional consequences of this pattern?

10. Walk the four principal frameworks of depression neuroscience (monoamine, inflammatory, glutamate/ketamine, HPA dysregulation). What does each contribute, and where does each have limits?

11. Articulate the HPA-metabolic intersection in cross-reference with Coach Food Bachelor's Lesson 2. How does chronic cortisol elevation contribute to ectopic lipid and insulin resistance, and how does metabolic syndrome alter HPA function?

12. Walk the Alexander fronto-striatal loop framework. Identify the five loops, their functional specializations, and one clinical condition associated with disruption of each.

13. Explain Schultz's reward prediction error finding and articulate how it unified with Sutton and Barto's temporal-difference learning. What does the unification establish about brain function and computation?

14. Distinguish Berridge's wanting from liking. What dissociations support the distinction, and what does the framework explain about late-stage addiction and about variable-reinforcement-schedule technologies?

15. Describe the BOLD signal at the level of its hemodynamic basis. What does Logothetis's work indicate about what fMRI signal corresponds to in underlying neural activity?

16. Articulate the multiple-comparisons problem in fMRI. What did the dead-salmon paper and the Eklund 2016 cluster-failure paper add to methodological awareness?

17. Explain optogenetics and chemogenetics. What methodological capability do these tools add that prior methods did not provide? What clinical translation has emerged or is emerging?

18. Articulate the reproducibility crisis in neuroscience and psychology. Identify three contributing factors and three reforms. How should pre-health and neuroscience students hold this knowledge as they read the literature?

19. Apply the five-point evaluation framework to the claim "dopamine is the happiness chemical." Where does the claim succeed, and where does it fail?

20. Articulate the Turtle's integrator position — receiver — at Bachelor's depth. Distinguish it from substrate (Food), internal environment (Water), synchronizer (Light), consolidation (Sleep), active output (Move), interface (Breath), system probe (Cold), and adaptive load (Hot).

---

Instructor's Guide

Pacing Recommendations

This chapter is designed for 18-22 class periods of approximately 50 minutes each — a full-semester upper-division undergraduate course in cognitive neuroscience, neuroscience methods, or systems and behavioral neuroscience. The depth and citation density are calibrated for upper-division coursework; lower-division survey students will struggle without Brain Associates as immediate prerequisite.

Suggested distribution:

• Lesson 1 — Cellular and Molecular Neuroscience: 4-5 class periods. Period 1: Hodgkin-Huxley framework and the action potential as mathematical object. Period 2: saltatory conduction; demyelinating disease. Period 3: synaptic transmission at SNARE and synaptotagmin depth. Period 4: glutamate receptor pharmacology (NMDA, AMPA, kainate, mGluR) and GABA receptors. Period 5: neuromodulation systems survey at receptor depth.

• Lesson 2 — Neuroplasticity and Memory: 4-5 class periods. Period 1: LTP molecular cascade. Period 2: Kandel and Aplysia. Period 3: BDNF/TrkB; adult neurogenesis controversy with Sorrells/Boldrini paired reading. Period 4: place cells, grid cells; entorhinal pathology in early AD. Period 5: engram tagging and the optogenetics-memory intersection.

• Lesson 3 — Stress and Pathophysiology: 4 class periods. Period 1: HPA axis at GR/MR receptor depth. Period 2: allostatic load and brain morphology consequences. Period 3: depression neuroscience with the four frameworks. Period 4: anxiety neurocircuitry and the HPA-metabolic intersection (the cross-reference to Food Bachelor's).

• Lesson 4 — Executive Function and Reward: 3-4 class periods. Period 1: persistent activity, DLPFC, working memory neuroscience. Period 2: Alexander fronto-striatal loops. Period 3: Schultz dopamine RPE and the Sutton-Barto unification. Period 4: Berridge wanting/liking, Nestler ΔFosB, addiction neurobiology.

• Lesson 5 — Research Methods: 3-4 class periods. Period 1: BOLD signal at signal-detection depth. Period 2: multiple comparisons, dead salmon, Eklund 2016. Period 3: electrophysiology, calcium imaging, optogenetics, chemogenetics. Period 4: reproducibility crisis and reform; five-point framework synthesis.

• End-of-chapter activity: Assigned across two weeks as out-of-class work; in-class peer review of methodological analyses can substitute for one class period.

• Quiz / assessment: One to two class periods.

Sample Answers to Selected Quiz Items

Q1 — Hodgkin-Huxley framework. Observation: voltage clamp recordings in squid giant axon identified separable inward (Na⁺) and outward (K⁺) currents with characteristic voltage- and time-dependence. Inference: voltage-gated channels with gating particles whose stochastic motion produces the macroscopic conductance; four coupled differential equations (one for V_m, three for gating variables m, h, n) predict action-potential waveform, threshold, refractory period, conduction velocity. The model was constructed before molecular identity of voltage-gated channels was known; the molecular reality (Na_v 1.x alpha subunit, four-domain transmembrane architecture, S4 voltage sensor, ball-and-chain inactivation) was confirmed decades later. The model is one of the paradigm successes of quantitative biology.

Q6 — Sorrells vs. Boldrini. Sorrells 2018 (Nature): immunostaining for DCX, PSA-NCAM in surgical and postmortem hippocampal samples; abundant neurogenic markers in childhood declining to undetectable in adults; conclusion adult human hippocampal neurogenesis is rare or absent. Boldrini 2018 (Cell Stem Cell): postmortem hippocampi, stereological counting, different markers (Ki67, DCX, PCNA); preserved neurogenic activity across adult lifespan including older adults. Differences may arise from tissue source (surgical vs. postmortem), tissue processing (postmortem interval and fixation affect DCX immunoreactivity), markers used, quantification approach. Resolution paths: improved fixation protocols (Moreno-Jiménez 2019); single-nucleus RNA-seq for transcriptomic neurogenesis signatures (Habib 2019 negative); ¹⁴C bomb-pulse dating (Spalding 2013); pre-registered analytical pipelines for histological neurogenesis quantification. The controversy illustrates the value of methodological literacy in reading conflicting research.

Q11 — HPA-metabolic intersection. Chronic cortisol elevation contributes to insulin resistance and ectopic lipid through several mechanisms: increased hepatic gluconeogenesis (cortisol-induced expression of glucose-6-phosphatase, PEP carboxykinase); peripheral lipolysis raising free fatty acid flux; visceral adipose deposition (cortisol preferentially supports visceral over subcutaneous adiposity); reduced insulin signaling at multiple steps. The intersection with Food Bachelor's Lesson 4 metabolic syndrome pathophysiology: ectopic lipid (intrahepatic, intramyocellular) drives DAG accumulation → PKCε/θ activation → IRS-1/2 serine phosphorylation → insulin resistance — and chronic cortisol amplifies this pathway through visceral adipose deposition and altered substrate flux. Reciprocally, metabolic syndrome and visceral obesity are associated with HPA alterations (flattened diurnal cortisol, blunted post-stress recovery). The systems are bidirectionally coupled. Clinical implication: depression and metabolic syndrome co-occur at rates substantially above chance and likely share mechanism.

Q13 — Schultz × Sutton-Barto unification. Schultz: phasic dopamine encodes reward prediction error — sharp increase at unexpected reward, shift to predictive cue across learning, decrease at omitted predicted reward. Sutton & Barto temporal-difference learning: agents learn state values through TD error δ = r_t + γV(s_{t+1}) − V(s_t); the bracketed quantity drives value updates. Mathematical match: Schultz's empirical signal has the form of Sutton-Barto's TD error. The unification (Schultz, Dayan, Montague 1996 Science) established that a specific neural population implements a specific computational algorithm with a specific function in reinforcement learning. Significance: one of the more complete cellular-to-computational mappings in neuroscience and a model for the kind of integration that upper-division work points toward.

Q20 — Receiver at Bachelor's depth. The brain receives inputs from every other modality through identifiable neural mechanisms at circuit-and-molecular level. Distinct from substrate (Food: amino acids, glucose, lipid, vitamin D — molecular inputs the brain depends on but is not itself); distinct from internal environment (Water: regulated chemical milieu in which neural function happens); distinct from synchronizer (Light: SCN-mediated timing information); distinct from consolidation (Sleep: the temporal pass that closes daily plasticity loops); distinct from active output (Move: kinetic signal of capacity that BDNF and exercise-driven plasticity emerge from); distinct from interface (Breath: voluntary-autonomic threshold engaged through prefrontal-brainstem connectivity); distinct from system probe and adaptive load (Cold, Hot: stress signals processed centrally). The receiver receives all of them and integrates into coherent cognition and behavior. At Bachelor's depth, the integration is through specific cellular machinery — the vagus-nucleus tractus solitarius pathway for some interoception, the retinohypothalamic pathway for light, the chemoreceptor pathway for breath-related blood gas, the proprioceptive thalamocortical pathways for movement, and so on.

Discussion Prompts

1. The Sorrells vs. Boldrini 2018 controversy on adult human hippocampal neurogenesis is unresolved. How should pre-clinical students hold this kind of contested literature? When is it appropriate to wait for resolution, and when is provisional integration into a broader framework appropriate?

2. The Schultz × Sutton-Barto unification is a striking example of computational neuroscience: a specific neural population implementing a specific algorithm. What does this success suggest about the prospects for similar integrations in other systems? Are there candidate areas where comparable unification might be on the horizon?

3. The Berridge wanting/liking framework explains how the brain's reward system can drive behavior without delivering corresponding pleasure. How should this neuroscience inform thinking about modern app design, the engagement economy, and policy questions around variable-reinforcement technologies?

4. The reproducibility crisis has prompted methodological reform across neuroscience and psychology. Identify a specific finding from this chapter (e.g., a fronto-striatal loop function, a dopamine RPE claim, an HPA-axis depression finding) and discuss how confident we should be in it given the methodological context.

5. The chapter discusses depression and addiction as research topics, not as conditions for students to diagnose. Why is this distinction important, and how should instructors handle student questions that move toward self-diagnosis or peer-diagnosis?

6. The HPA-metabolic intersection is the first Bachelor's-tier intra-tier lateral. How does this kind of cross-Coach integration shape the broader pre-clinical conversation about how mental and physical health relate? What other intersections do you anticipate in upcoming Bachelor's chapters (Sleep-Brain, Move-Brain, Cold-Brain)?

7. Optogenetics has transformed circuit-neuroscience research over the last 15 years. What can we expect from the next 15 years of methodological development? What questions, currently inaccessible, might become tractable?

Common Student Questions

Q: I'm interested in psychiatry. How should I use this chapter relative to medical-school coursework? A: The chapter covers research-grade pathophysiology that will overlap with medical neuroscience and psychiatry coursework but at less clinical depth. Medical school will add the clinical evaluation framework, the DSM diagnostic system, pharmacology in clinical context, and patient-care application. This chapter prepares the mechanistic ground; medical school adds the clinical layer. Both are needed.

Q: Should I be worried about my own mental health based on what I'm reading here? A: The neuroscience here is descriptive. Reading about depression, anxiety, or addiction in research terms is not the same as having those conditions. That said, the late-teens / early-twenties window is a real-incidence peak for several conditions, and if anything in the chapter resonates with patterns in your own life that you are working through alone when you do not need to be, asking for support is appropriate. Your college counseling center, your primary care provider, and the 988 Lifeline are real resources.

Q: What about nootropics? Microdosing? Off-label use of prescription stimulants? A: The chapter does not prescribe. Research on nootropics outside of clinically-validated indications is limited and often poorly controlled. Off-label use of prescription stimulants without ADHD diagnosis is associated with documented risks (cardiovascular, dependence, academic-integrity issues) and is illegal at most institutions. Microdosing psychedelics is an active research area with limited high-quality controlled human evidence. None of these are decisions to make on the strength of a neuroscience chapter; a healthcare provider who knows your medical history is the right resource.

Q: How do I evaluate a popular-press neuroscience claim? A: Use the five-point framework. (1) Is the mechanism plausible? (2) What study design produced the evidence, and what kind of question can it answer? (3) What is the effect size in context? (4) Has the finding replicated, and across what populations and methods? (5) Is the clinical or practical translation appropriate? Most popular neuroscience claims fail at points 2 or 5. Pre-health and neuroscience-major students benefit from training in this evaluation as a core literacy.

Q: I'm a neuroscience major. Will I cover the methods (optogenetics, fMRI, electrophysiology) at this depth in upper-division methods courses? A: Probably yes. This chapter introduces the methods at conceptual depth; specific methods courses extend each to procedural and analytical depth. The chapter is preparation; methods courses are application.

Q: The chapter mentions that the prefrontal cortex matures into the mid-twenties. Does this mean the mid-twenties are when you become an adult? A: Brain maturation continues into the third decade, but "becoming an adult" is not a single neuroscience phenomenon. Different brain systems mature at different rates; cognitive control, emotion regulation, and risk-evaluation continue to develop into the mid-to-late twenties; social cognition has its own developmental trajectory; and "adulthood" is a social and legal construct that does not map directly onto any single neural change. The neuroscience is real; the simple "brain is fully mature at X" framing oversimplifies what is a continuous developmental process.

Q: I am worried about a friend or roommate. What do I do? A: Talk with care, not judgment. Express that you have noticed something and you are concerned. Encourage them to engage with your college counseling center, health center, or a healthcare provider. The 988 Lifeline (call or text 988) and Crisis Text Line (text HOME to 741741) are available 24/7 for crisis situations. The National Alliance for Eating Disorders helpline (866-662-1235) is staffed weekdays by licensed therapists for eating-disorder-adjacent concerns. Stay in contact, do not promise secrecy if safety is involved, and bring in adult help. The older NEDA helpline (1-800-931-2237) is non-functional and should not be cited.

Parent / Adult Family Communication Template

(Optional for instructors whose course communicates with adult family members; many Bachelor's students are independent adults, so use at your discretion.)

Subject: Coach Brain — Bachelor's Level — Cellular and Cognitive Neuroscience

Dear Families,

This unit covers the Coach Brain chapter at the Bachelor's degree level of the CryoCove Library — the second chapter of the upper-division undergraduate tier. The chapter goes substantially deeper than Associates: cellular and molecular neuroscience at biophysics depth, neuroplasticity and memory at molecular cascade resolution, stress neurobiology and clinical pathophysiology of depression and anxiety, executive function and reward neuroscience including the dopamine reward-prediction-error work, and neuroscience research methods at upper-division depth.

Several notes you may want to know about:

• Clinical pathophysiology is covered at research-grade depth — depression, anxiety, addiction, ADHD, the schizophrenia spectrum. All content is descriptive (mechanism and recognition) rather than diagnostic; clinical evaluation is framed throughout as the work of licensed clinicians, not undergraduate study.
• Research methods are taught as core curriculum. Upper-division neuroscience means learning to read primary research with appropriate methodological discipline. The chapter engages with the reproducibility crisis, the limits of fMRI, and the optogenetic methodological revolution directly.
• Mental health awareness is sharpened for the pre-health and neuroscience-major population, where late-teens / early-twenties first-onset windows for several conditions are real. Verified crisis resources are included: 988 Lifeline (call/text 988), Crisis Text Line (text HOME to 741741), National Alliance for Eating Disorders (866-662-1235). Note: the older NEDA helpline (1-800-931-2237) is non-functional and is not used in our curriculum.

If your student has any specific medical or mental health context that intersects with the chapter content, please encourage them to review the material alongside their healthcare provider.

With respect, The CryoCove Library Team

Resource Verification Note for Instructors

Crisis resources change. Re-verify the active status of the 988 Lifeline, Crisis Text Line (text HOME to 741741), and National Alliance for Eating Disorders helpline (866-662-1235) before each term you teach this chapter. The NEDA helpline (1-800-931-2237) was discontinued in 2023 and remains non-functional; flag any student work that cites it and redirect.

Additionally, re-verify currency of cited primary literature before each term. Bachelor's-level chapters are calibrated to the literature available at writing; specific clinical findings (depression treatment trials, neuroimaging consensus statements, addiction-treatment literature) update periodically and should be cross-referenced against current sources for clinical-rotation-bound students.

---

Illustration Briefs

Lesson 1 — The Hodgkin-Huxley Model

• Placement: After "The Hodgkin-Huxley Model as Mathematical Object"
• Scene: A diagram with three panels. Left: schematic of squid giant axon with voltage clamp setup (intracellular and bath electrodes). Center: the four coupled equations of the H-H model in clean typography (V_m equation plus three gating-variable equations). Right: the predicted action-potential waveform from the model, with phases labeled (resting, threshold, depolarization, repolarization, after-hyperpolarization).
• Coach involvement: Coach Brain (Turtle) at the side, gesturing toward the equations with the note: "From observation to inference to prediction."
• Mood: Foundational, mathematical, anchored.
• Caption: "The action potential is not a metaphor."
• Aspect ratio: 16:9 web, 4:3 print

Lesson 2 — The LTP Molecular Cascade

• Placement: After "LTP at Molecular Depth"
• Scene: A schematic of a CA1 dendritic spine receiving CA3 input. Sequential labeled steps: (1) glutamate release; (2) AMPAR depolarization; (3) Mg²⁺ block relief and NMDAR Ca²⁺ influx; (4) CaMKII activation and autophosphorylation at Thr286; (5) GluA1 phosphorylation and AMPAR insertion; (6) CREB phosphorylation pathway leading to nuclear gene expression; (7) protein synthesis returning to the synapse via synaptic tagging and capture.
• Coach involvement: Coach Brain (Turtle) at the side, watching the cascade unfold with the note: "From milliseconds to hours, the same spine."
• Mood: Molecular, integrative, clear.
• Caption: "Synaptic strengthening is a cascade, not a moment."
• Aspect ratio: 16:9 web, 4:3 print

Lesson 3 — The HPA Axis and the GR/MR Receptor System

• Placement: After "The HPA Axis at Receptor and Gene-Expression Depth"
• Scene: An anatomical-style diagram of the HPA axis (PVN, anterior pituitary, adrenal cortex) with hormone labels (CRH, ACTH, cortisol) on each arrow. To one side, an inset showing the GR/MR receptor distribution across brain regions (hippocampus prominent MR, amygdala/PFC prominent GR). Cortisol concentration curve below showing diurnal rhythm with stress-elevated peak.
• Coach involvement: Coach Brain (Turtle) gesturing toward the receptor inset with the note: "Same hormone, different rooms."
• Mood: Endocrine-anatomical, integrative.
• Caption: "Cortisol is one signal. The brain reads it through two receptors."
• Aspect ratio: 16:9 web, 4:3 print

Lesson 4 — The Schultz × Sutton-Barto Unification

• Placement: After "Schultz and the Reinforcement Learning Unification"
• Scene: A two-panel split. Top: Schultz's classic raster plots showing dopamine neuron firing in the three conditions (unexpected reward; cue-predicted reward; predicted reward omitted). Bottom: the corresponding TD-error equation δ_t = r_t + γV(s_{t+1}) − V(s_t) shown in clean typography, with the three terms boxed and labeled to match the experimental conditions. An arrow links the two panels with the label: "Same signal, same equation."
• Coach involvement: Coach Brain (Turtle) at the side, with the note: "Computation in cells."
• Mood: Integrative, foundational.
• Caption: "When neurons compute, the algorithm shows."
• Aspect ratio: 16:9 web, 4:3 print

Lesson 5 — The BOLD Signal and the Inferential Chain

• Placement: After "The BOLD Signal: Ogawa 1990 as Foundational Anchor"
• Scene: A four-step diagram. Step 1: neural activity (action potentials, synaptic activity) in a cortical region. Step 2: hemodynamic response — local oxygen consumption rises, blood flow rises more, deoxyhemoglobin decreases. Step 3: BOLD signal increase in T2*-weighted images. Step 4: statistical map of significant voxels with appropriate corrections. Arrows between steps with caveats labeled (neurovascular coupling assumed; spatial blurring; temporal blurring; multiple comparisons must be addressed).
• Coach involvement: Coach Brain (Turtle) at the side, watching the chain with the note: "Each step is an inference. Read accordingly."
• Mood: Methodological, careful.
• Caption: "fMRI does not measure neurons. It measures blood."
• Aspect ratio: 16:9 web, 4:3 print

---

Citations

1. Hodgkin AL, Huxley AF. (1952). A quantitative description of membrane current and its application to conduction and excitation in nerve. Journal of Physiology, 117(4), 500-544.

2. Catterall WA. (2017). Forty years of sodium channels: structure, function, pharmacology, and epilepsy. Neurochemical Research, 42(9), 2495-2504.

3. Hartline DK, Colman DR. (2007). Rapid conduction and the evolution of giant axons and myelinated fibers. Current Biology, 17(1), R29-R35.

4. Trapp BD, Nave KA. (2008). Multiple sclerosis: an immune or neurodegenerative disorder? Annual Review of Neuroscience, 31, 247-269.

5. Südhof TC. (2013). Neurotransmitter release: the last millisecond in the life of a synaptic vesicle. Neuron, 80(3), 675-690.

6. Traynelis SF, Wollmuth LP, McBain CJ, et al. (2010). Glutamate receptor ion channels: structure, regulation, and function. Pharmacological Reviews, 62(3), 405-496.

7. Sigel E, Steinmann ME. (2012). Structure, function, and modulation of GABA(A) receptors. Journal of Biological Chemistry, 287(48), 40224-40231.

8. Beaulieu JM, Gainetdinov RR. (2011). The physiology, signaling, and pharmacology of dopamine receptors. Pharmacological Reviews, 63(1), 182-217.

9. Aston-Jones G, Cohen JD. (2005). An integrative theory of locus coeruleus-norepinephrine function: adaptive gain and optimal performance. Annual Review of Neuroscience, 28, 403-450.

10. Berger M, Gray JA, Roth BL. (2009). The expanded biology of serotonin. Annual Review of Medicine, 60, 355-366.

11. Hasselmo ME, Sarter M. (2011). Modes and models of forebrain cholinergic neuromodulation of cognition. Neuropsychopharmacology, 36(1), 52-73.

12. Lisman J, Yasuda R, Raghavachari S. (2012). Mechanisms of CaMKII action in long-term potentiation. Nature Reviews Neuroscience, 13(3), 169-182.

13. Bliss TV, Lømo T. (1973). Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. Journal of Physiology, 232(2), 331-356.

14. Kandel ER. (2001). The molecular biology of memory storage: a dialogue between genes and synapses. Science, 294(5544), 1030-1038.

15. Bailey CH, Kandel ER, Harris KM. (2015). Structural components of synaptic plasticity and memory consolidation. Cold Spring Harbor Perspectives in Biology, 7(7), a021758.

16. Sacktor TC. (2011). How does PKMζ maintain long-term memory? Nature Reviews Neuroscience, 12(1), 9-15.

17. Volk LJ, Bachman JL, Johnson R, Yu Y, Huganir RL. (2013). PKM-ζ is not required for hippocampal synaptic plasticity, learning and memory. Nature, 493(7432), 420-423.

18. Collingridge GL, Peineau S, Howland JG, Wang YT. (2010). Long-term depression in the CNS. Nature Reviews Neuroscience, 11(7), 459-473.

19. Paoletti P, Bellone C, Zhou Q. (2013). NMDA receptor subunit diversity: impact on receptor properties, synaptic plasticity and disease. Nature Reviews Neuroscience, 14(6), 383-400.

20. Kandel ER, Dudai Y, Mayford MR. (2014). The molecular and systems biology of memory. Cell, 157(1), 163-186.

21. Bartsch D, Ghirardi M, Skehel PA, et al. (1995). Aplysia CREB2 represses long-term facilitation: relief of repression converts transient facilitation into long-term functional and structural change. Cell, 83(6), 979-992.

22. Erickson KI, Voss MW, Prakash RS, et al. (2011). Exercise training increases size of hippocampus and improves memory. Proceedings of the National Academy of Sciences USA, 108(7), 3017-3022.

23. Castrén E, Monteggia LM. (2021). Brain-derived neurotrophic factor signaling in depression and antidepressant action. Biological Psychiatry, 90(2), 128-136.

24. Egan MF, Kojima M, Callicott JH, et al. (2003). The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell, 112(2), 257-269.

25. Kempermann G, Song H, Gage FH. (2015). Neurogenesis in the adult hippocampus. Cold Spring Harbor Perspectives in Biology, 7(9), a018812.

26. Sorrells SF, Paredes MF, Cebrian-Silla A, et al. (2018). Human hippocampal neurogenesis drops sharply in children to undetectable levels in adults. Nature, 555(7696), 377-381.

27. Boldrini M, Fulmore CA, Tartt AN, et al. (2018). Human hippocampal neurogenesis persists throughout aging. Cell Stem Cell, 22(4), 589-599.e5.

28. Moreno-Jiménez EP, Flor-García M, Terreros-Roncal J, et al. (2019). Adult hippocampal neurogenesis is abundant in neurologically healthy subjects and drops sharply in patients with Alzheimer's disease. Nature Medicine, 25(4), 554-560.

29. Spalding KL, Bergmann O, Alkass K, et al. (2013). Dynamics of hippocampal neurogenesis in adult humans. Cell, 153(6), 1219-1227.

30. O'Keefe J, Dostrovsky J. (1971). The hippocampus as a spatial map. Preliminary evidence from unit activity in the freely-moving rat. Brain Research, 34(1), 171-175.

31. O'Keefe J, Nadel L. (1978). The Hippocampus as a Cognitive Map. Oxford University Press.

32. Hafting T, Fyhn M, Molden S, Moser MB, Moser EI. (2005). Microstructure of a spatial map in the entorhinal cortex. Nature, 436(7052), 801-806.

33. Kunz L, Schröder TN, Lee H, et al. (2015). Reduced grid-cell-like representations in adults at genetic risk for Alzheimer's disease. Science, 350(6259), 430-433.

34. Liu X, Ramirez S, Pang PT, et al. (2012). Optogenetic stimulation of a hippocampal engram activates fear memory recall. Nature, 484(7394), 381-385.

35. de Kloet ER, Joëls M, Holsboer F. (2005). Stress and the brain: from adaptation to disease. Nature Reviews Neuroscience, 6(6), 463-475.

36. Reul JM, de Kloet ER. (1985). Two receptor systems for corticosterone in rat brain: microdistribution and differential occupation. Endocrinology, 117(6), 2505-2511.

37. Pariante CM, Lightman SL. (2008). The HPA axis in major depression: classical theories and new developments. Trends in Neurosciences, 31(9), 464-468.

38. McEwen BS, Stellar E. (1993). Stress and the individual. Mechanisms leading to disease. Archives of Internal Medicine, 153(18), 2093-2101.

39. McEwen BS. (2007). Physiology and neurobiology of stress and adaptation: central role of the brain. Physiological Reviews, 87(3), 873-904.

40. Lupien SJ, McEwen BS, Gunnar MR, Heim C. (2009). Effects of stress throughout the lifespan on the brain, behaviour and cognition. Nature Reviews Neuroscience, 10(6), 434-445.

41. Sapolsky RM. (2005). The influence of social hierarchy on primate health. Science, 308(5722), 648-652.

42. Krishnan V, Nestler EJ. (2008). The molecular neurobiology of depression. Nature, 455(7215), 894-902.

43. Miller AH, Raison CL. (2016). The role of inflammation in depression: from evolutionary imperative to modern treatment target. Nature Reviews Immunology, 16(1), 22-34.

44. Berman RM, Cappiello A, Anand A, et al. (2000). Antidepressant effects of ketamine in depressed patients. Biological Psychiatry, 47(4), 351-354.

45. Duman RS, Aghajanian GK, Sanacora G, Krystal JH. (2016). Synaptic plasticity and depression: new insights from stress and rapid-acting antidepressants. Nature Medicine, 22(3), 238-249.

46. Daly EJ, Singh JB, Fedgchin M, et al. (2018). Efficacy and safety of intranasal esketamine adjunctive to oral antidepressant therapy in treatment-resistant depression. JAMA Psychiatry, 75(2), 139-148.

47. Holsboer F. (2000). The corticosteroid receptor hypothesis of depression. Neuropsychopharmacology, 23(5), 477-501.

48. Tovote P, Fadok JP, Lüthi A. (2015). Neuronal circuits for fear and anxiety. Nature Reviews Neuroscience, 16(6), 317-331.

49. Fuster JM, Alexander GE. (1971). Neuron activity related to short-term memory. Science, 173(3997), 652-654.

50. Goldman-Rakic PS. (1995). Cellular basis of working memory. Neuron, 14(3), 477-485.

51. Curtis CE, D'Esposito M. (2003). Persistent activity in the prefrontal cortex during working memory. Trends in Cognitive Sciences, 7(9), 415-423.

52. Stokes MG. (2015). "Activity-silent" working memory in prefrontal cortex: a dynamic coding framework. Trends in Cognitive Sciences, 19(7), 394-405.

53. Alexander GE, DeLong MR, Strick PL. (1986). Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annual Review of Neuroscience, 9, 357-381.

54. Schultz W. (1998). Predictive reward signal of dopamine neurons. Journal of Neurophysiology, 80(1), 1-27.

55. Sutton RS, Barto AG. (1998). Reinforcement Learning: An Introduction. MIT Press.

56. Schultz W, Dayan P, Montague PR. (1997). A neural substrate of prediction and reward. Science, 275(5306), 1593-1599.

57. Berridge KC, Robinson TE. (1998). What is the role of dopamine in reward: hedonic impact, reward learning, or incentive salience? Brain Research Reviews, 28(3), 309-369.

58. Berridge KC. (2007). The debate over dopamine's role in reward: the case for incentive salience. Psychopharmacology, 191(3), 391-431.

59. Nestler EJ. (2008). Review. Transcriptional mechanisms of addiction: role of ΔFosB. Philosophical Transactions of the Royal Society B, 363(1507), 3245-3255.

60. Ogawa S, Lee TM, Kay AR, Tank DW. (1990). Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proceedings of the National Academy of Sciences USA, 87(24), 9868-9872. [Foundational anchor — BOLD signal.]

61. Logothetis NK. (2008). What we can do and what we cannot do with fMRI. Nature, 453(7197), 869-878.

62. Bennett CM, Baird AA, Miller MB, Wolford GL. (2009). Neural correlates of interspecies perspective taking in the post-mortem Atlantic Salmon: an argument for proper multiple comparisons correction. Journal of Serendipitous and Unexpected Results, 1, 1-5.

63. Eklund A, Nichols TE, Knutsson H. (2016). Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates. Proceedings of the National Academy of Sciences USA, 113(28), 7900-7905.

64. Buzsáki G. (2004). Large-scale recording of neuronal ensembles. Nature Neuroscience, 7(5), 446-451.

65. Boyden ES, Zhang F, Bamberg E, Nagel G, Deisseroth K. (2005). Millisecond-timescale, genetically targeted optical control of neural activity. Nature Neuroscience, 8(9), 1263-1268.

66. Armbruster BN, Li X, Pausch MH, Herlitze S, Roth BL. (2007). Evolving the lock to fit the key to create a family of G protein-coupled receptors potently activated by an inert ligand. Proceedings of the National Academy of Sciences USA, 104(12), 5163-5168.

67. Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), aac4716.

68. Button KS, Ioannidis JP, Mokrysz C, et al. (2013). Power failure: why small sample size undermines the reliability of neuroscience. Nature Reviews Neuroscience, 14(5), 365-376.

69. Munafò MR, Nosek BA, Bishop DV, et al. (2017). A manifesto for reproducible science. Nature Human Behaviour, 1, 0021.

70. Sahel JA, Boulanger-Scemama E, Pagot C, et al. (2021). Partial recovery of visual function in a blind patient after optogenetic therapy. Nature Medicine, 27(7), 1223-1229.

71. Tonegawa S, Pignatelli M, Roy DS, Ryan TJ. (2015). Memory engram storage and retrieval. Current Opinion in Neurobiology, 35, 101-109.

72. Tank DW, Hopfield JJ. (1986). Simple "neural" optimization networks: an A/D converter, signal decision circuit, and a linear programming circuit. IEEE Transactions on Circuits and Systems, 33(5), 533-541.

73. Niv Y. (2009). Reinforcement learning in the brain. Journal of Mathematical Psychology, 53(3), 139-154.

74. Wise RA. (2004). Dopamine, learning and motivation. Nature Reviews Neuroscience, 5(6), 483-494.

75. Volkow ND, Koob GF, McLellan AT. (2016). Neurobiologic advances from the brain disease model of addiction. New England Journal of Medicine, 374(4), 363-371.

76. Koob GF, Volkow ND. (2016). Neurobiology of addiction: a neurocircuitry analysis. Lancet Psychiatry, 3(8), 760-773.

77. Lupien SJ, Maheu F, Tu M, Fiocco A, Schramek TE. (2007). The effects of stress and stress hormones on human cognition: implications for the field of brain and cognition. Brain and Cognition, 65(3), 209-237.

78. Belmaker RH, Agam G. (2008). Major depressive disorder. New England Journal of Medicine, 358(1), 55-68.

79. Diamond A. (2013). Executive functions. Annual Review of Psychology, 64, 135-168.

80. Casey BJ, Jones RM, Hare TA. (2008). The adolescent brain. Annals of the New York Academy of Sciences, 1124, 111-126.

81. Poldrack RA, Baker CI, Durnez J, et al. (2017). Scanning the horizon: towards transparent and reproducible neuroimaging research. Nature Reviews Neuroscience, 18(2), 115-126.

82. Wise SP. (2008). Forward frontal fields: phylogeny and fundamental function. Trends in Neurosciences, 31(12), 599-608.