Cellular and Molecular Neuroscience - University of Wisconsin-Madison

Notes and sketchy diagrams for NTP 610, the graduate course on cellular and molecular neuroscience by the Neuroscience Training Program PhD at UW-Madison, supercharged with a few additions of my own. Computer graphics are not of my authorship, with the exception of this one.

• Cellular Excitability

Prof. Shing-Yan Chiu's lectures.

• Resting potential

action potential: all-or-none
action potential: fixed amplitude
resting potential: generated by K⁺ channels on normal cell

Non excitable cell (0 mV):

Intracellular: 140 mM K⁺, 140 mM Cl-: electrically neutral
Extracellular: 4 mM K⁺, 4 mM Cl-

Excitable (-90 mV):

Add permeable K channel, therefore K⁺ ions leave cell until equilibrium (Nernst, aka reversal potential)
velocity of decay/hyperpolarization depends on amount of channels.

• Single ion: Nernst

Three points about Nernst potential:

only minute amount of K needs to leave in order to create big potential, due to capacitance properties of membrane.
At Nernst equilibrium there's no further ion flow.
Energy conversion scheme: concentration gradient -> electric potential

From Gibbs' actual/free energy:

$Δ G = R T l n (o u t / i n)$ $(TeX formula: ΔG = RT \; ln(out/in) )$

but also ΔG is energy per Coulomb of transferred charnge:

$Δ G = z F E$ (TeX formula: ΔG = zFE )

therefore:

$\underset{c h e m i c a l e n e r g y}{\underset{⏟}{R T l n (\frac{K_{o u t}}{K_{i n}})}} = \underset{e l e c t r i c e n e r g y}{\underset{⏟}{z F E}}$ $(TeX formula: \underbrace{RT \; ln(\frac{K_{out}}{K_{in}})}_{chemical \; energy} = \underbrace{zFE}_{electric \; energy} )$

Where R is the constant for ideal gases, T is temperature, z is ion valence (either positive or negative), F is Faraday's constant, and E is the electric potential. K stands for "concentration" (not restricted to potassium). Therefore, the Nernst potential of an ion is:

$E_{K} = (\frac{R T}{z F}) l n (\frac{K_{o u t}}{K_{i n}}) ≅ (58 m V) l o g_{10} (\frac{K_{o u t}}{K_{i n}})$ $(TeX formula: E_K = (\frac{RT}{zF}) ln(\frac{K_{out}}{K_{in}}) ≅ (58 mV) \; log_{10} (\frac{K_{out}}{K_{in}}))$

substituting for K concentration levels:

$(58 m V) l o g_{10} (4 / 140) = - 90 m V$ $(TeX formula: (58 mV) log_{10}(4/140) = -90 mV )$

Increasing outside concentration depolarizes cell log-linearly

• Multiple ions: GHK

High Na²⁺ outside, low inside (opposite gradient).
Na²⁺ has its own Nernst potential
When both, resting potential lives somewhere in between (usually tilted towards potassium. P_K:P_Na = 1:0.03):

Resting potential with multiple ions is described by steady-state Goldman-Hodgkin-Katz equation:

$V_{r e s t} = (58 m V) l o g_{10} (\frac{Σ_{i = 1}^{n} K_{i_{o u t}}}{Σ_{i = 1}^{n} K_{i_{i n}}})$ $(TeX formula: V_{rest} = (58 mV) \; log_{10} \left( \frac{Σ_{i=1}^n K_{i_{out}}} {Σ_{i=1}^n K_{i_{in}}} \right) )$

No longer log-linear depolarization. Rapid increase, then log-linear:

There will be constant ion oscillations though, no steady state. That's why Na-K ATP-ase is needed to maintain steady state.

• Ion channels and patch-clamp

At rest, $P_{k} > > P_{N a}$ $(TeX formula: P_k >> P_{Na})$ . During action potential, $P_{N a} > > P_{k}$ $(TeX formula: P_{Na} >> P_k)$ .

2 hypotheses:

1 ion channel which changes selectivity from Na -> K
2 distinct channels

Early evidence: TEA blocks K, TTX blocks Na.

Definitive proof: patch clamp of single channel protein.

From Ohm's law:

$I_{i o n} = ɣ_{i o n} (V - E_{i o n})$ $(TeX formula: I_{ion} = ɣ_{ion} (V - E_{ion}) )$

where ɣ is conductance (1/R) and E_ion is the Nernst potential (its fingerprint), meanwhile V can be manipulated using clamping.

• characterizing ion channel with patch clamp

patch clamp modes:

ON-cell patch: pipette contiguous to cell, yet without breaking membrane
INSIDE-OUT patch: membrane patch is detached from cell, allowing scientist to change concentration on both sides.
WHOLE-cell clamp: pipette contiguous to cell, breaking membrane. Measures whole-cell properties.
OUTSIDE-OUT patch: like INSIDE-OUT, but membrane mosaic is reversed with respect to pipette.

• Ion channels

how do ion channels achieve such a high ion current (6 millions of ions / s)?

• Selectivity

In a typical protein, Radicals confer specificity, whereas backbone is featureless.

In a K channel, polypeptide chain resembles water-like polarity on the side that doesn't have the peptide Radicals (only Oxygen and Hydrogen). Radicals face outside, into the background.

"Featureless" hydrogens and carbonyl oxygens create specificity by creating a fake water tunnel.

Na isn't fooled, because channel is rigid and too wide for smaller Na atom to be surrounded by "fake water"

K channel: 6 transmembrane subunits. Pore depends on 5 or 6 peptides, common to all species and known channels.

Hypothesis confirmed by expressing crystals of this protein subset.

Screening domain at the opening of the Na channel doesn't need to span the whole pore.

• Gating

Voltage-gated state space:

        α
Closed ---> Open
       <--
        β

Movable gate lever is positively-charged particle called q.

Proposed mechanism (q+ is electrically attracted to be on either side (close or open), but not in the middle). Symmetrical energy barrier.

when voltage is applied on either side, q+ is "biased" to go to one side (the one that is more negatively charged). Energy barrier is tilted.

open/close rate constants (exponential growth/decay). Constant units are Hz.

$α = V_{0} e^{.5 q V / k T}$ $(TeX formula: α = V_0 \; e^{.5qV/kT} )$

$β = V_{0} e^{- .5 q V / k T}$ $(TeX formula: β = V_0 \; e^{-.5qV/kT} )$

Probability of being open at steady state is sigmoid with image [0, 1], perfectly odd, centered at 0.5.

$P_{o} (V) = \frac{α (V)}{β (V) + α (V)}$ $(TeX formula: P_{o}(V) = \frac{α(V)} {β(V) + α(V)} )$

Probability as function of time:

rising phase: $P_{o} (t) = 1 - e^{- t / τ}$ $(TeX formula: P_{o}(t) = 1 - e^{-t/τ} )$

decaying phase: $P_{o} (t) = e^{- t / τ}$ $(TeX formula: P_{o}(t) = e^{-t/τ} )$

With time constant

$τ = \frac{1}{(α + β)}$ $(TeX formula: τ = \frac{1}{(α + β)} )$

Meanwhile, the current observed in oscilloscope, encompassing N gates:

$I_{K} = N P_{o} (V - E_{K})$ (TeX formula: I_K = N P_o (V - E_K) )

• Single channel kinetics

Each individual voltage-dependent channel behaves as a memory-less stochastic square pulse train, yet I_K follows characteristic curve, because group behaves as I_K on average.

     ___    _______   _
____|   |__|       |_| |____
--->     -->       ->   ---->
dwell times (closed periods)

what is the relationship between opening rate α and dwell time? roughly speaking, the smaller the α, the longer the dwell time.

$α = v e l o c i t y e^{- Δ G / k T}$ $(TeX formula: α = velocity \; e^{-ΔG/kT} )$

where ΔG is the height of function.

histogram of dwell times (exponential distribution). In other words, most channels open early:

$P = α e^{- α t}$ $(TeX formula: P = αe^{-αt} )$

with mean dwell time = 1/α.

what if each channel has a different α? still an exponential PDF:

$P = (α_{1} + α_{2} + . . .) e^{- (α_{1} + α_{2} + . . .) t}$ $(TeX formula: P = (α_1 + α_2 + ...)e^{-(α_1 + α_2 + ...)t} )$

going back to both O and C, suppose 2-state model with rates:

        α = 100/s
Closed ---> Open
       <--
        β = 1000 /s

There will be 2 exponential PDFs of dwell times, one per state. substituting parameters (α for C, β for O)

• multiple states

real channels have much more states:

        α         β_2
Closed ---> Open ---> Closed2
       <--       <---
        β         α_2

PDF of dwell times in O will still be single exponential

$P (O) = (β_{1} + β_{2}) e^{- (β_{1} + β_{2}) t}$ $(TeX formula: P(O) = (β_1 + β_2) e^{-(β_1 + β_2) t} )$

same for Closed states, however, there's no way to tell which closed state the protein is in, therefore, measurement will show single plot with a sum of 2 exponentials.

$P (C) = A_{1} α_{1} e^{- α_{1} t} + A_{2} α_{2} e^{- α_{2} t}$ $(TeX formula: P(C) = A_1 α_1 e^{-α_1 t} + A_2 α_2 e^{-α_2 t} )$

Many possible state diagrams, even for same alphabet:

C_1 <-> O <-> C_2
or
C_1 <-> C_2 <-> O
or
      O
    /   \
  C_1 -  C_2

• Macroscopic currents

continuing with 2 state model, single channel (Ohm's law again, g is conductance, P_o is either 1 or 0):

$i = g P_{o} (V - E)$ (TeX formula: i = g P_o (V-E) )

meanwhile at the macroscopic level. (P_o ranges from [0, 1]):

$i = N g P_{o} (V - E)$ (TeX formula: i = N g P_o (V-E) )

substituting Ng for constant c:

$i = c P_{o} (V - E)$ (TeX formula: i = c P_o (V-E) )

macroscopic circuit (with time constant $τ = R C$ (TeX formula: τ = RC) , different from the microscopic τ presented before):

          Capacitor (phospholipid layer)
         __| |__
        |  | |  |
        |       |
inside -|       |- outside
        |       |
        |__/\/__|
           G_L (leakage conductance, channel)

Voltage source is added in series to resistor, to account for that ion's electrochemical force (Nernst potential).

              C
         _____| |_____
        |     | |     |
        |             |
inside -|             |- outside
        |       |     |
        |__/\/__|  |---
                |
           G_L   E_k

Leakage I-V curve is straight slope (up is outflowing ions). More voltage, more positive (outgoing) current.

Then, extra parallel C+E_k lines are added for each ion in the full Hodgkin Huxley model (notice G_Leak is separate from ions)

Na channel I-V curve:

• Na-channel dynamics (depolarization)

Perturbations around resting potential fixed point (a stable fixed point, aka attractor), still below activation threshold.

Hyperpolarizing:

Depolarizing:

Perturbation above activation threshold (unstable fixed point, aka repulsor:)

Next stage is repolarization of action potential, due to the late opening of K-channels and inactivation of Na-channels.

• K channel disease

Molecular basis for diversity:

many genes
mix and matching subunits (monomeric subunits, polymeric)
splicing
posttranslational modification
beta subunits added, post translation

100s of genes, yet each individual K channel isn't redundant. Single mutation can lead to neuropathology:

Kv1.1 knock-out -> Episodic Ataxia (EA), uncontrollable peripheral (skeletal) motion after stress.

• Na inactivation (HH)

activation is much faster than inactivation.

activation >>> fast inactivation > slow inactivation

inactivation is not the reverse of de-activation. It needs a different molecular process

voltage-gated Na-channel has 2 gate subunits, m and h:

activation (m gate, fast): α_m, β_m
inactivation (h gate, slow): α_h, β_h

        outside
 ________    _________
       m  \
 ________\  h_________
          \
        inside
  ++++depolarized++++

inactivation (h) rate curves are the opposite of (m), but slower. current only flows when both of them are open, in other words, the product of their respective P_o.

fraction of occupancy curves are the same, but mirrored.

Units are independent stochastic processes. Random direct jumps from closed state to inactivated state (silent traces), were predicted by HH before they could be verified with single-channel patch clamp in the 80s:

C -> O -> Inactivated
\________/
   silent traces (1% or 2% of time)

Failed prediction: β_h is not a voltage-dependent particle inside membrane, but a ball-chain hanging outside, for fast inactivation. open times are constant, despite change in V.

• Na inactivation (Ball-chain)

blocking in squid's giant axon was already hinting at a separate ball-and-chain terminal dangling outside the protein and responsible for fast inactivation.

there are 3 tails to each side of the TM passes (3 N-terminals, 3-C terminals). both sides play a role in inactivation, but only one of them is well elucidated (N-terminals).

chain-ball is one of those tails. Very end (telomere) is hydrophobic. proximal part is hydrophilic, therefore it rolls like a ball.

ball-chain is not V-dependent. It swings randomly until plugging into channel by accident (thermal vibration).

   α    κ
C -> O -> Inactivation
  <-   <-
   β    λ

Mutations:

wrong ball size: bursting times unaffected (because tail is same length still), but inactive period is shorter because ball is quickly released.

wrong chain length: bursting duration affected, inactive period unaffected.

• Activation

steep P_o(V) of Na channel is key to physiology of activation. how do we increase steepness of sigmoid?

normal:

  α = q
C -> O
  <-
  β

steeper:

increase α = 4q. that is, charged particle in gate is 4 times more charged. normal curve of energy barrier becomes taller, but in time domain activation increases astronomically.

q: 1 ms
4q: 1 trillion years
12q: actual channel actually moves 12q when open. what gives?

channel is activated in steps via cooperativity, and actual sigmoid is compromise between q and 4q. only 1 subunit opens at a time, this creates some delay.

no delay in turning off currents, however. all it takes is 1 subunit to go back, and ion current will be off.

how can we measure opening of subunit? gating current: small pulse-like currents from ions already trapped within channel/membrane, as they flow in when single subunit opens.

mutations in Nav1.2 channel show Subunit 4 is the voltage sensor (same for K Shaker channel, Cav1.1, HCN1 Ih channel and CNG1 channel). neutralizing its peptide charges destroys voltage sensitivity.

whether mutation affects more intra or extra-cellular peptides biases P_o.

evidence for physical movement of S4:

biotin-avidin assay
Fluorescence Resonance Energy Transfer

• Presynaptic Mechanisms

Prof. Meyer B. Jackson

Synapse (σύναψις: conjunction, strong union), coined by Sherrington in 1897.

characteristic ultrastructure of synapse (as seen from EM):

lots of (pre) synaptic vesicles
synaptic cleft ~20 nm
post-synaptic density

• Classification

by contact:

electric (gap junctions): single step, bidirectional, fast
chemical (neurotransmitter+receptor): multiple steps, one direction, synaptic delay (~1 ms)

by effect (speaking of chemical):

excitatory: ligand-gated Na/K channels, fast (.5 ms - 5 ms)
inhibitory: ligand-gated Cl channels, somewhat slower (.5 ms - 5 ms)
modulatory: GPCRs acting downstream on channels, much slower (50 ms - 5 s)

• Transmitter apparatus

to use a particular neurotransmitter, neuron must have means to synthesize it. requirements:

synthesis
storage/packaging
release into cleft
receptor activation
removal/degradation/reuptake

with the exception of (3), all steps are transmitter-specific.

big synapses which allow for recordings:

squid's giant axon/synapse
Calyx of Held in auditory brain stem (chochlear -> trapezoid body), young animals.

ACh receptors were first to be found (~1981). found in neuromuscular junction as well as spinal cord, permeable to both K and Na.

• Quantal release

Axelrod: proved existence of vesicles using EM in rat pineal gland
Fatt & Katz, 1951, 1952: spontaneous blips at neuromuscular junction, small miniature random PSPs (minis), observed only at endplate.
Del Castillo & Katz: fluctations of Ach come in discrete values. number of observations as function of end-plate potential is a distribution with many peaks (multimodal).

summation of quanta (stochastic model):

N = total number of release sites
p = probability of release at each site
q = voltage change produced by one vesicle (quantal size)

PSP of magnitude v produced by k quanta has amplitude:

$v = k q$ (TeX formula: v = kq )

average number of vesicles released per trial is:

$m = N p$ (TeX formula: m = Np )

m is referred to as (quantal content)

therefore average EPSP is:

$\overline{v} = m q = N p q$ $(TeX formula: \bar{v} = mq = Npq )$

number of vesicles released, k, varies from trial to trial according to binomial dist:

$P (k) = (\binom{N}{k}) p^{k} {(1 - p)}^{N - k}$ $(TeX formula: P(k) = \binom{N}{k} p^k (1-p)^{N-k} )$

$= \frac{N!}{(N - k)! k!} p^{k} {(1 - p)}^{N - k}$ $(TeX formula: = \frac{N!}{(N-k)!k!} p^k (1-p)^{N-k} )$

with low Ca and high Mg (for large N and small p), the binomial dist approximates the Poisson dist:

$P (k) = \frac{m^{k} e^{- m}}{k!}$ $(TeX formula: P(k) = \frac{m^k e^{-m}}{k!} )$

for k=0,

$P (0) = e^{- m}$ $(TeX formula: P(0) = e^{-m} )$

which means m can be determined by looking at the number of failures (failure to release vesicles, despite stimulation), for any given stimulation magnitude.

$m = - l o g (P (0))$ (TeX formula: m = -log(P(0)) )

how big is a quantum?

neuromuscular Ach: 5000-10000 molecules, Kuffler & Yoshikami, 1975
glutamatergic: 1200-2000, Edwards, 1995

• pre vs post

m, quantal content (average # of vesicles) serves as index of presynaptic strength.

q, quantal size (voltage by one vesicle) serves as index of postsynaptic strength.

therefore, measuring m and q serves as basic method of finding locus of change in synaptic strength.

sources of quantal variability:

different vesicle volume
different degree of loading
different fractional release
different alignment (receptor density)

quantal release detected by other techniques:

Heuser and Reese designed machine to freeze tissue exactly after stimulation, allowing EM to photograph vesicles fusing into membrane.
capacitance measurement: capacitance is proportional to Area, and it turns out that membrane capacitance in patch clamp actually changes in discrete steps. there are 2 kinds of steps, corresponding to 2 types of vescicles: dense core and micro/fast.
amperometry: NE and serotonin can react upon contact with certain electrodes, releasing e- in the process which are measured as quantal spikes by electrode.
fluorescence staining

Dense-core vesicles are intracellular lipidic vesicles released by neurons for neuromodulation of synapses. They tend to contain neuropeptides (e.g. substance P, opioids), which are bigger molecules than catecholamine-based or single aminoacid-based trasmitters. This makes them denser to imaging techniques such as EM. In addition, they are bigger, and are released after considerable delays in presynaptic activity.

• Calcium

cells care about calcium

PMCA: plasma membrane Ca2+ ATPase pump
NCX: Nitrogen-Calcium exchangers
Ca2+ channels

extracellular Ca2+ concentration is higher. E_Ca = 120 mV.

19th century Ringer's solution: Ca2+ was found to be essential to keep frog's heart beating.
1964 Douglas & Poisner: Ca needed in bathing solution to get release of epinephrine from adrenal gland, and vassopresin from posterior pituitary (neurohypophysis).
1967 Katz & Miledi: Ca needs to be at nerve terminal before stimulation, Mg2+ inhibits the action of Ca2+.

Miledi & Katz's experiment on squid's giant axon:

   ___
__|   |__> I
     |\
   __| \
__/     \__> V
       ^
       due to voltage dependent Ca2+ current
   ^
   normal rising phase due to Na+

Fura-2: designed chelator molecule which changes fluorescence, invented by Tsien of GFP fame. Descendent: Fluo-5F

Natural Ca chelators

Calbindin
Calretinin
Parvalbumin

• Channels

Ca2+ channels aren't perfectly selective to Ca2+ only, other ions get in. Therefore I-V curve doesn't reverse exactly at E_Ca.

4 subunits
pore loop very similar to Na channels
point mutation at pore loop converts it from Ca2+ to Na+ pore.

types:

L-type (1.1 - 1.4)
P/Q, N, R (2.1, 2.2, 2.3)
T-type (3.1 - 3.3)

by selectively blocking using toxins from spiders and cone snails, type 2 channels were shown to be the ones involved in presynaptic release.

• Release

Ca2+ triggered by exocistosis in cooperative mechanism. vescicle surface is full of proteins. most abundant is synaptobrevin.

key proteins in release:

3 SNAREs:
synaptovrebin (aka VAMP)
syntaxin
SNAP25
synaptotagmin: binds to phospholipids, SNAREs and itself; when Ca2+ is bound.

botulinum toxins (tetanus toxin, which cleaves synaptobrevin) block syntaptic transmission. tetanus toxin turned out to be a protease requiring Zn.

classification by location

on the surface of vescicles:
synaptobrevin
synaptotagmin
cell membrane proteins involved:
SNAP25
syntaxin

• SNARE complex

from yeast to humans, vesicle fusion from one membrane to another is highly conserved. synaptic release leverages homologue mechanism.

SNARES on vesicles are called v-SNAREs, SNAREs on target membrane are called t-SNARES.

SNARE motifs comprise ~65 amino acids with helical tendency. they assemble into trimeric complex (1 syntaxin, 1, synaptobrevin, 2 SNAP-25).

the complex functions as ropes which pull membranes together. demonstrated using fluorescence on glycoproteins (Weber at al 1998, Cell).

2 target proteins originally close in vescicle membrane.
fusion
fluorescence diminishes because proteins get further appart as their membrane fuses with target membrane.

however this doesn't require Ca2+. What's the Ca2+ sensor?

• synaptotagmin

different membrane protein is responsible for sensing Ca2+: synaptotagmin. actually a whole protein family: SYT[1-17], ESYTs and others.

ɑ-helix as TM.
C2A subunit in proximal ECD (relative to membrane)
C2B subunit in distal ECD.

several aspartate radicals bind Ca2+

unsolved puzzle: SYT works on the seconds timescale, it's too slow to constitute the crucial synaptic mechanism.

• GPCRs and G-proteins

ligand-gated vs G-protein divide:

transmitter	ionotropic	GPCR
Ach	Nic	Musc
glutamate	iGluRs	mGluRs
GABA	GABA_A	GABA_B
serotonin	5HT3	5HT1, 5HT2
other amines (NE)		y
hormones		y
peptides		y
photoreceptors		y
olfactants		y

many co-transmitters are GPCR-only as well.

• GPCRs

Cellular signaling can be more complex than ionotropic receptors. G-protein-coupled receptors were first discovered for glycogenolysis. Glycogenolysis increases blood flow and sugar in fight or flight responses:

  epinephrine -> liver -> glycogen breakdown to glucose

search for hormone receptors in liver. search for protein in tissue which binds to epinephrine. it started with β-adrenergic receptor, which is homologous to rodopsin.

hydrophobicity profile showed it has 7 TM domains.

• Classification

largely conserved. family tree goes as far back as prokaryotes. Human GPCRs can function in yeast. Low sequence conservation however, conservation is at the level of higher-order structure.

G-protein coupled receptors is the most abundant family of membrane receptors. they are involved in vision (rodopsins), olfaction (ORs), dopamine receptors, opioid receptors, histamine, vasopressin, adrenaline, etc. They are the target of 30% - 40% of all drugs.

class A
class B (secretin)
class C (glutamate)
adhesion
frizzled

opioid receptors:

mu
kappa
delta

most common endogenous opioid peptides in humans:

enkephalins
endorphins
dynorphins

structure-function:

allosteric receptors
most binding sites are extra-cellular
ligand-binding domain is called "cradle"

different ligands have different actions

full agonist (e.g. epinephrine)
partial agonist (many drugs)
neutral antagonist (few drugs)
inverse agonist (most antagonists)

• G-protein

1971: Martin Rodbell showed GTP was involved in intracellular signaling in activation of adenyl cyclase
1977-1979: cell line with proteins which bind GTP, which he dubbed G-proteins.

diversity not as wide as GPCRs.

transduction cycle:

unbound
binding: GDP (intra) still coupled to Gɑ
GTP replaces GDP, binding Gɑ
Gɑ and GβGɣ dissociate (Gɑ is released)
transduce signals: both released Gɑ and attached GβGɣ may produce cascades
Gɑ slowly hidrolizes bound GTP, complex reassembles again

Targets:

Gɑ -> adenylate cyclase -> cAMP -> PKA

G-protein -> phospholipase C --> DAG -> PKC
                             \-> IP3 -> Ca2+ release from endoplasmic reticulum

• kinases

cAMP-dependent
cGMP-dependent
Ca-calmodulin-dependent
Ca or diacylclycerol-dependent

G-protein is a signal amplifier

signal
1. G-protein/effector enzyme
  1. second messenger
    1. activated kinase
      1. activated enzyme
      2. activated enzyme
      3. activated enzyme
      4. activated enzyme
      5. etc...
  2. second messenger
    1. activated kinase
      1. etc...
2. G-protein/effector enzyme
  1. second messenger
    1. activated kinase
      1. etc....
3. ad nauseaum

• Plasticity

many presynaptic mechanisms of short term plasticity (targets). Pretty much anything affecting Ca levels will result in short-term changes in release, until Ca levels are restored.

action potential
Ca levels
Ca channels: because it regulates release
other channels: "
available vescicles
autoreceptors
GPCRs
pumps? (speculative)
mitochondria: they regulate cytocolic Ca
postsynaptic receptors: more expression, desensitization

• potential broadening

mossy fibers in hippocampus show broadened (slower action potential) after high-frequency pulse train. synapse is facilitated with burst in DG. Caused by K+ channel inactivation.

• Facilitation with Ca

NMJ of crayfish claw opener, due to residual Ca2+. if Ca is elevated after first spike, then why is there no release during the inter-stimulus interval? first of all, sensor complex in vesicles has limited binding sites. also, there is a distinct facilitation sensor (NCS-1) at the exocytocis mechanism.

• Depression with Ca

there are buffer calbindin proteins which must store Ca before delivering to vesicles.

Calbindin is a natural calcium chelator (surrounds Ca2+ ions). These proteins act as Ca2+ buffers at nerve terminals, limiting the availability of free intracellular Ca2+ in the presynaptic neuron before exocytosis. They underlie a form of paired-pulse facilitation, as shown by Blatow, et al. (2003) with calbindin-KO mice, for which this kind of short-term plasticity is lost.

• potential slowdown

slow responses in bullfrog sympathetic ganglions are mediated by different transmitters (Glu, Ach, Leutenizing Hormone Releasing Hormone), each with a different EPSP/IPSP timescale (from ms, to s, to min).

ligand-gated neuromodulators almost always trigger G-protein coupled signaling cascades. e.g., cholera toxin attacks G-protein ɑ subunit.

Also, K+ channels are modulated by G-protein directly, without need for 2nd messengers.

Ca2+ as second messenger can also interact with some subtypes of K+ channels (K_Ca channels, BK, SK, IK). they are responsible for hyperpolarization right after rising phase of action potential.

• habituation

case study: Kandel's aplysia and gill withdrawal reflex:

habituation: if repeated stimulation at siphon is not accompanied by adverse effect in tail, animal habituates and no gill withdrawal takes place. this is due to facilitatory interneuron which lacks nociceptive sensory signal from tail. This probably results in inactivation of presynaptic Ca2+ channels and depletion of vesicles.
facilitation: if stimulation does have adverse effect, gill withdrawal response is increased. This is because of facilitatory interneuron activates GPCRs in the presynaptic terminals, and the extra GTP+Gɑ activates the adenylate-cyclase → cAMP cascade, which increases Ca2+ influx. Also, PKA is activated, which phosphorylates and inhibits certain K+ channels, slowing repolarization.

• case study: optogenetics

• CRE-lox driver:

promoter specific to some kind of cell
stop codon
loxP site (short aminoacid sequence after stop codon)
GFP

CRE-recombinase splices the loxP, so that the GFP will also be expressed in that cell type.

• GECIs

Genetically Encoded Ca Sensors. synthetic fusion of:

GFP (circularly permuted)
calmodulin
M13 kinase

when Ca binds calmodulin, it changes conformation of GFP, which changes fluorescence, which is leveraged by optical imaging (e.g. 2-photon microscopy) to record intracellular Ca2+ transients (albeit at a relatively slow timescale). GCaMP is the most popular one. E.g., the GCaMP6s variant is one of the “slow variants” (the ‘s’ stands for slow), which often provide increased sensitivity compared to the variants with faster temporal resolution.

• GEVIs

Voltage sensors. faster than Ca sensors. also synthetic voltage-sensing molecules attached to GFP dye.

quantum dipole moment changes from ground state to excited state in presence of stronger electric field. absorbance spectrum shifts as a result, creating small but noticeable fluorescence change.

Siegel & Isacoff (1997) paper: early attempt in oocytes.

GEVIs currently available:

hVOS: very good (single-spike) time resolution.

• channelrhodopsin

channelrhodopsin-2 GPCR (like bacterial rhodopsin, but it opens channel upon lighting instead of closing it) discovered in Deisseroth (2005) paper. found in a species of green algae. core protein is 315 aminoacids. needs trans-retinal to function.

under voltage clamp, light produces current.
under current clamp, long light pulses produces spiking. short light pulses produce subthreshold PSPs or pulse-coupled spikes.

retinal (conserved in our retina) is isomerized upon reaction with photon, forcing conformation change with channel.

Opsin based optogenetics arsenal:

channelrhodopsins: Na, K, Ca, H influx
halorhodopsins: Cl influx
archearhodopsins: H outflow

• Postsynaptic Mechanisms

Prof. Mathew Jones lectures.

• Pharmacodynamics 101

how were receptors discovered?

1878 Langley was working with salivary glands of cat:

pilocarpine: agonist which mimics physiology of drug. atropid: antagonist

Langley proposed existence of receptive part in tissue, which certain chemicals can excite or inhibit.

1908 A.V. Hill, preparation of frog's leg muscle bathed in solution, muscle in turn was linear actuator to draw its own contractions over rolling paper.

nicotine application:

  ____
_|    |__

capacitor-like plot (due to law of mass action)

  ____
 /    |
/      \
-------> time

simplest differential equation, with S being concentration of "stuff":

$\frac{d N}{d t} = - k N$ $(TeX formula: \frac{dN}{dt} = -kN )$

2 hypothesis about muscle contraction:

Mechanichal hypothesis:

          k
 N_{out} ---> N_{in}
         <--
          k

$\frac{d N}{d t} = k N_{o u t} - k N_{i n}$ $(TeX formula: \frac{dN}{dt} = k N_{out} - k N_{in} )$

turning it into a difference equation for numerical integration (substituting rhs for constant F):

$\frac{N (t + 1) - N (t)}{Δ t} = k N_{o u t} - k N_{i n} = F$ $(TeX formula: \frac{N(t+1) - N(t)}{Δt} = k N_{out} - k N_{in} = F )$

therefore

$N (t + 1) = Δ t F + N (t)$ (TeX formula: N(t+1) = Δt F + N(t) )

Chemical hypothesis:

                    k_{on}
N    +     Receptor ---> N+Receptor
                    <--
                    k_{off}

$\frac{d N R}{d t} = k_{o n} N R - k_{o f f} N R$ $(TeX formula: \frac{d NR}{dt} = k_{on} NR - k_{off} NR )$

(product NR is due to joint probability of there being both N and R).

$N R = \frac{k_{o n} N}{k_{o n} N + k_{o f f} N}$ $(TeX formula: NR = \frac{k_{on} N}{k_{on} N + k_{off} N} )$

• EC₅₀ and K_D

EC₅₀ is the empirical concentration in Mol at which NR (response) is half of maximum.

on the other hand, the closely related affinity constant is given a priori by the state model:

$K_{D} = \frac{k_{o f f}}{k_{o n}}$ $(TeX formula: K_D = \frac{k_{off}} {k_{on}} )$

• Hill equation

then, Hill's equation is:

$N R = \frac{k_{o n} N}{N + K_{D}}$ $(TeX formula: NR = \frac{k_{on} N}{N + K_D} )$

dose-response curve:

NR (response)
|     ------> saturates
|   /
| /
|___|_______> log[N]
   EC_50

a drug with less affinity would require higher doses for the same response (horizontal shift):

NR (response)
|          ------> saturates
|        /
|      /
|____/
|_______|________> log[N]
      EC_50

not all change is due to affinity/binding. there's efficacy/efficiency too (e.g. channels become "more open"), which causes up-shift in dose-response curve, while EC₅₀ stays the same:

NR (response)
|       ------> saturates
|     /
|   /
| /
|/
|___|_______> log[N]
   EC_50

Evidence for opening/closing of channels after transmitter or drug binding.

1951: Fatt & Katz
1959: Takeuchi & Takeuchi

muscle end plate preparation (neuromuscular junction). current clamp, and letting V in muscle vary as function of I.

V
|/\/\/\/\
|_________> t
I
|/\/\/\/\
|_________> t

after application of Ach to preparation:

           _____
V         /     \
|/\/\/\/\/       \
|_________> t
I
|/\/\/\/\/\/\/\/\
|_________> t

if V increases even though I is stationary (quasi-constant), that means Ach is reducing resistance (Ohm's law).

Ach-R are mixed cation channels (they allow both K and N to go through). that is, opening those channels is driving reversal potential to ~0 mV.

using voltage clamp they found that reversal potential was neither that of Na nor K:

G
|__  _________
|  \/
|__  _________
|  \/
|------------- ~0 mV
|--/\________
|__/\_________
|________________________> t
(bumps are due to Ach application)

• Excitatory transmission

general rule:

excitatory potential: reversal potential of ion-provoking postsynaptic potential lies above threshold potential, regardless of whether potential is driven up or down.
inhibitory potential: reversal potential of ion-provoking postsynaptic potential lies below threshold potential, regardless of whether potential is driven up or down.

in other words, whether transmitter is excitatory or inhibitory for a given neuron is solely determined by the position of its ionic current's Nernst/GHK potential, relative to cell activation threshold.

• AMPA/Kainate vs NMDA

experimental history:

1984: Mayer & Westbrook
1984: Ascher

CA3 -(glutamatergic)-> CA1
    ^                   ^
    induce spikes       recording site

EPS currents:

I
|__/-----\____
|__/--\_______
|------------- ~0 mV
|--\__/-------
|--\_____/----
|________________________> t
   ^    ^
   fast ^
        ^
        slow phase before going back to rest

fast (I-V curve, perfect straight line):

      /
    /
  /
/

V-clamp and I-clamp parametric curves look the same as the previous experiment, similarly for I-V: reversal potential is ~0 mV. blocking either Na or K moves reversal potential to that of the other ion. i.e. similar to N-Ach receptor.

slow:

However, slow phase is due to different receptor. I-V curve is sensitive to Mg2+ (although it still reverses at ~0 mV, so it allows both Na and K in and out).

      /
     /
____/______
\  /
 \/

by blocking Mg2+, it becomes a perfect straight line. at hyperpolarization, Mg2+ is attracted towards cell due to mere electrostatic force (it's a cation, and cell is hiperpolarized), which in turn blocks NMDA pore. therefore I-V curve is slow to go up, momentarily.

additionally, NMDA receptor also allows Ca2+ to flow.

fast: AMPA (antagonists: DNQX, CNQX)
slow: NMDA (antagonists: APS, APV)

• Inhibitory transmission

IPS currents:

I
|__/-----\____
|__/--\_______
|------------- ~ -75 mV
|--\__/-------
|--\_____/----
|________________________> t

-75 mV is close to $E_{C l^{-}}$ $(TeX formula: E_{Cl^-})$ , and it was shown experimentally to be the involved ion.

many barbiturates (e.g. propofol) and sedatives (e.g. benzodiazepins) work by enhancing GABAr.

• Ligand-gated channels

• Pentameric LGICs

Pentameric ligand-gated ion channels. A.K.A. cys-loop receptors due to presence of a disulphide-bridged loop in the extracellular domain, although recently discovered bacterial homologues do not contain the disulphide bridge.

Superfamily of proteins responsible for fast response to the following:

acetylcholine (nAch)
serotonine (5-HT₃)
g-aminobutyric acid (GABA_A, GABA_C)
glycine

Multiple subunit isoforms and splicing contribute to variance.

Structure:

Most complete structure comes from nAch of Torpedo marmorata (electric ray). Ring of 5 subunits, each one with a β-sheet-rich extra cellular (EC) domain, a ɑ-helix-rich transmembrane domain (TM) and a poorly known intracellular domain.

Between 2 and 5 binding sites at ECDs, depending on type. cys-loops communicate ECDs and TMDs. Each TMD has 4 ɑ-helices, one of which (M2) contributes to channel itself.

Physiology:

Binding at interface of complementary ECD loops -> conformational change.
Coupling: loop-C capping and receptor priming "traps" agonist and transmits energy downstream.
1. Evidence of pre-activation and flip states, before fully open state. Whether agonist is partial or full might depend on how efficacious it is to move from flip to open state.
Gating of ions (different agonists induce it with different efficiencies, more agonist molecules induce it more because of extra affinity).

transmembrane domains of Cys-loop receptors, particularly GABA_A receptors, contain binding sites for a range of clinical and endogenous modulators. 15' residue of M2 is often a key determinant: water-filled external cavity seemingly capable of accomodating drugs.

• Tetrameric iGluRs

ionotropic glutamate receptors (AMPA, kainate and NMDA). Main mechanism for fast excitatory transmission at the CNS.

Genetic and posttranslational editing generate iGluR subtypes with diverse kinetics, Ca2+ ion permeability, and sensitivity to voltage-dependent ion channel block by extracellular magnesium (NMDAr) and cytoplasmic polyamines.

There's evidence of presynaptic iGluRs, but their function is less well characterized. Embryonic vs adult subunits can be strikingly different, although role of ontogeny is poorly understood. Glial cells also express some iGluRs (e.g. oligodendtrocytes).

3 major families (named after preferred agonists):

AMPA (GluA1–4 genes)
Kainate (GluK1–5 genes)
NMDA (GluN1 + GluN2A–D or GluN3A–B genes)
fourth family: GluD1 and GluD2 genes, widely expressed in cerebellum, yet unresponsive to glutamatergic or glycinergic activation. Poorly understood.

Structure:

The 18 vertebrate iGluR genes have a unique conserved structure. iGluRs also found in insects, nematodes and primitive eukaryots. Each monomer is divided into:

TMD (4 passes): pore-loop helix flanked by 3 α-helices, with characteristic SYTANLAAF motif in M3 helix.
ECD (dimer of dimers):
- Ligand-Binding Domain (LBD) subunit (proximal to membrane): 2 domains (A+D or B+C subsubunits).
- Amino-terminal Domain (ATD) subunit (distal to membrane): 2 domains (A+B or C+D subsubunits)

Plus auxiliary proteins that interact to regulate iGluRs: TARPs, CKAMP44, CNIH-2, SynDIG1 (AMPAr); NETO1/2 (KainateR); NETO1 (NMDAr).

Physiology:

Models tend to define 3 states: closed, open and desensitized. computational models suggest pore closure-opening is due to M3, which in turn is moved by LBD. Distal and proximal subunits serve different roles in activation.

In NMDAr, glycine binding to GluN1 LBDs is not enough for activaton, whereas in KainateR, binding to only proximal GluK5 is enough as a partial agonist. Partial Glu binding to GluK1 produces mild desensitization, strong one after binding to GluK5.

Assembly of channel is started by ATD. ATDs (GluN2A–D subunits) are also key to allosteric modulation. E.g., GluN2A has a P_o = 0.5, whereas for GluN2B it is only 0.12, and even lower for B and D. Probably due to ATD-LBD linker.

NMDA receptors are heteromeric coincidence detectors requiring binding of both glycine and glutamate and that all 4 LBDs be occupied . Unlike non-NMDA receptors, for which binding to fewer than four subunits can trigger partial activation.

• Shaping of synaptic potential

membrane capacitor + single leakage channel:

$I = I_{C} + I_{L}$ (TeX formula: I = I_C + I_L )

$I = C \frac{d V}{d t} + G_{L} V_{L}$ $(TeX formula: I = C \frac{dV}{dt} + G_L V_L )$

$\frac{I - (G_{L} V_{L})}{C} = \frac{d V}{d t}$ $(TeX formula: \frac{I - (G_L V_L)}{C} = \frac{dV}{dt} )$

more generally, for any number of different neurotransmitter receptors:

$\frac{d V}{d t} = \frac{- Σ I}{C}$ $(TeX formula: \frac{dV}{dt} = \frac{-Σ I}{C} )$

$\frac{d V}{d t} = \frac{- [I_{s t e a d y s t a t e} + G (V - E_{L}) + G_{s y n} (V - E_{s y n})]}{C}$ $(TeX formula: \frac{dV}{dt} = \frac{ -[I_{steady\;state} + G(V-E_L) + G_{syn}(V-E_{syn})] } {C} )$

algebraic trick to solve for V as steady state as a function of equilibrium potentials V(E), assuming C=1 and dV/dt = 0:

chord conductance (steady state trick):

$V_{S S} = \frac{G_{k} E_{L}}{G_{t o t}} + \frac{G_{k} E_{L}}{G_{t o t}} + . . . + \frac{G_{N a} E_{N a}}{G_{t o t}}$ $(TeX formula: V_{SS} = \frac{G_k E_L}{G_{tot}} + \frac{G_k E_L}{G_{tot}} + ... + \frac{G_{Na} E_{Na}}{G_{tot}} )$

• Dendritic integration

• Temporal summation

in real synapses in CNS, injected I_syn isn't a perfect Dirac pulse nor a square pulse, because of neurotransmitter diffusion at synaptic cleft:

EPSC:

I_syn
|
| |\   |\     |\
|_| \__| \____| \__
|__________________> t

meanwhile for potential, when many EPS currents come in train:

V------------------ V_thr
|       __
|   /\_/  \___  /\
|__/          \/
|_______________> t

There would be equivalent pair of I_syn and IPSP but in the opposite direction. Total PSP is the sum of EPSP and IPSP.

If new receptors are opened (new E+R parallel path in membrane circuit), by Ohm's law that means resistance decreases, therefore time constant decreases, therefore V spikes are shorter (faster decay), therefore there's less summation.

$Δ τ = R (t + 1) C - R (t) C < 0$ (TeX formula: Δτ = R(t+1) C - R(t) C < 0 )

And vice versa.

• Spatial summation (cable theory)

compartmentalize dendtrites into simple patches of iso-electric properties.

suppose 3 compartments:

dendtrite 1
dentrite 2
soma, where final decision is made

    _____| |_____
   |     | |     |
   |             |   |
 --|             |- -||| outside   <- dendrite 1
|  |       |     |   |
|  |__/\/__|  |---
|          |
|
/
\ intracellular Resistance due to difference in V from compartment to another
/
|
|   _____| |_____
|  |     | |     |
|  |             |   |
 --|             |- -||| outside    <- soma
|  |       |     |   |
|  |__/\/__|  |---
|          |
|
/
\ intracellular Resistance due to difference in V from compartment to another
/
|   _____| |_____
|  |     | |     |
|  |             |   |
 --|             |- -||| outside    <- dendtrite 2
   |       |     |   |
   |__/\/__|  |---
           |

space constant:

$λ \propto \sqrt{\frac{R_{m e m b r a n e}}{R_{i n t r a c e l l u l a r}}}$ $(TeX formula: λ ∝ \sqrt{\frac{R_{membrane}}{R_{intracellular}}} )$

If Δλ < 0, there's less spatial summation, and vice versa.

• Firing patterns and oscillations

• single spike

-------|-------> t

I_L-V curve:

  /
 /
/

I_K-V curve (above I_L):

  /
 /
/

I_Na-V curve (below I_L):

    /
\  /
 \/

Σ I:

based on differential equation,

$\frac{d V}{d t} = \frac{- Σ I}{C}$ $(TeX formula: \frac{dV}{dt} = \frac{-Σ I}{C} )$

we can determine sign of derivative and whether 3 fixed points are stable or unstable:

         /
  /\    /
-/--\--/----
/    \/
-><- --><---
 s   u  s
        ^
        steady state after spike
     ^
     V_threshold
 ^
 V_rest

however, V doesn't stay at after-spike steady state, because high depolarization opens K channels and inactivates Na channels with ball-chain. therefore individual curves and total I curve changes.

after repolarizing, curves change back to normal. rinse and repeat.

• bursting

• pacemaker cell

--|--|--|--|--

due to I_L being way lower, therefore V_rest moves below x-axis, it's no longer a solution. cell is always tending to depolarization (before spike).

• Plateau

plateau potential, with addition of I_Ca2+ T-type:

  ______
__|     |___ V(t)
--\     /--- I(t)
   \___/

this results in attenuation of spikes (adaptive habituation):

 |
 | |
_|_|_|__

• LTP/LTD, pattern storage & recall

1890: Pavlov
William James, Principles of Psychology: Elementary Law of Association "When two elementary brain processes have been active together or in immediate succession, one of them, on reoccurrence, tends to propagate its excitation to the other"
1940s: Donald Hebb, Hebbian Synaptic Plasticity (rephrases James). "Coincident Pre- and Postsynaptic activity will lead to strenghening of the synapse."
1971: David Marr, Theory of Archicortex (hippocampus). Autoassociative network, relies on hebbian plasticity.

at that point it was all theory. first experiment came from Bliss & Lomo, 1973:

Rabbit Dentate Gyrus (close to Marr's hippocampus)
In vivo, extracellular recording of LFP EPSPs after estimulation of pre-synaptic neurons.
recordings of normal EPSPs in time:

_/\_ _/\_ _/\_

By serendipity, they gave tetanic (high freq) stimulation for 10 s
EPSP was stronger, and lasted for weeks

   /\      /\
 _/  \_  _/  \_

If normalizing peak amplitude against an original denominator, plot of EPSP points over time look like this:

EPSP amp
|          ** *   *
|        **  * ***
|       *
|      *
| *   *
|*  **
|-------------------> t

requires NMDAr activation (coincidence detector: because of Mg+ blocker, both post and pre must be active for it to allow K/Na/Ca ions flow)
NMDAr required for induction, not expression
requires postsynaptic Ca2+ influx (through NMDAr)

People disagreed on whether plasticity is due to presynaptic or postsynaptic. Consensus is on postsynaptic protein synthesis.

Resolution: originally the synapse is "silent" (AMPA receptors aren't originally there, by the action of NMDAr, LTP inserts them at the postsynaptic dendrite).

• LTD (reverse process)

EPSP amp
|** *   *
| **  * ***
|          *
|           *
|             * **
|              *  **
|-------------------> t

requires NMDAr activation (as with LTP).
NMDAr receptors required for induction, not expression (as with LTP).
requires postsynaptic Ca2+ influx (through NMDAr) (as with LTP)

Difference lies in that stimulation is low frequency, around 1 Hz, for 1 min. Removal and insertion of AMPAr depends on different enzymes. Insertion enzymes have low Ca2+ affinity, removal enzymes have high Ca2+ affinity.

• Cascades and Transcription

Prof. Avtar Roopra's lectures

translation is required for induction of long term potentiation, but not for maintenance. anysomicin/actinomycin-D application before or after stimulation doesn't impair potentiation, only during stimulation.

• Transcription and splicing

• Transcription

element: short functional sequence of DNA.

In essence, RNA polymerase binds to promoter and starts transcription to mRNA.

eukaryotic promoters:

TATA Box (TATATT, etc.)
in neuron genes, GGGGG... is another common motif

Full process actually requires a pre-initiation complex with 6 steps, starting with TBP, helicase to un-braid DNA and several transcription factors.

• Splicing

gene could have contained introns (non-Coding Sequences), which are discarded later on by alternative splicing. splicing at junctions is performed by destruction complex.

anatomy of mRNA:

   |CAP|UTR|AUG(start)|CDS|...|CDS|UTR|AAA.. (poly-A tail)|
mRNA
↑
| splicing
|
5'   ...UTR|AUG(start)|CDS|nCDS|...|CDS|nCDS|stop|UTR... 3'
heterogeneous mRNA (pre-mRNA)
↑
| RNA polymerase (transcription)
|
5'   ...UTR|AUG(start)|CDS|nCDS|...|CDS|nCDS|stop|UTR... 3'
DNA

• Enhancers

suppose:

5'|10000 bp|promoter|300bp|
^RNA Polymerase II + transcription factors = RNAP complex

add sequence to tube filled with salt and ATP buffers: no transcription.

add TBP (TATA Binding Protein) + TAF_n factors = TFIID complex: moderate transcription.

add bunch of other things: maximal transcription (that something else enhances transcription).

remove preceding 10000 bp: back to moderate transcription.

that means something in the discarded base pairs together with extra proteins (extra transcription factors) enhances transcription: epigenetically targeted sequences. The extra proteins are known as enhancers or activators.

In the case of c-fos, it turned out to be the CREB protein and corresponding cis sequence (CRE sequence: cAMP Response Element Binding protein).

Mark Ptashne's model of transcription factors:

() <- activation domain
() <- DNA binding domain
____________________________________ <- DNA

activation domain was hypothesized to interact with epigenetic signal. to demonstrate it, AD is replaced by some sham protein, which is heterologous to another piece which now contains a synthetic activation domain. this restores expression:

moreover, activation domain was shown to be enough to bring TAF complex, or even whole RNAPII.

many genes have more than 1 promoter.

• Chromatin packing and gene regulation

late 1800s Germany: nucleus divided in 2 regions:

heterochromatin: darker patches, denser after centrifugation
euchromatin: rest of nucleus, lighter after centrifugation

active vs inactive genes:

euchromatin: active genes (as shown by reporters/immunohistochemistry.
heterochromatin: inactive genes.

histones (basic, positively charged):

H2A|H2B      <- upper cover
_______
H3|H4|H3|H4  <- roll
__________
H2A|H2B      <- bottom cover

in addition, each subunit carries a tail approx 33-34 aminoacids long:

e.g:

H3 -----------------------N
     K27    K9       K4     <- lysines

147 bp around each histone protein.

in heterochromatin, lysines are positively charged, whereas the phosphate backbone is negative. in euchromatin/unrolled DNA, lysines are acetylated.

2 catalyzed reactions for going back and forth between positive and neutral tails:

hetero -> eu: uses acetyl-CoA, esterification leaving CoA residual. GCN5 enzyme
eu -> hetero: uses H20, hydrolizes with acetyl-group residual. RPD3 enzyme.

catalized in the order of minutes:

                      acetyl-CoA           CoA
                            \HAT(e.g. CBP) /
transcription OFF H3-K-NH3^+ ---------> H3-K-N-OAc transcription ON
                             <---------
                               / HDAC \
                vinegar(acetate)        H2O

glossary:

HAT: histone acetyl transferase
bromo domains: domain at CBP-like proteins to recognize lysines (and then acetylate them using KAT domain).
HDAC: histone de-acetylase

CREB Binding Protein (CBP, which binds to CREB's active domain) was shown to share sequence with GCN5, meaning that gene activation or silencing is mostly done through mechanical unrolling of DNA, to allow transcription machinery to bind to genes.

CBP
(CREB) <- activation domain
(CREB) <- DNA binding domain
____________________________________ <- DNA
CRE

opposite process (wrapping/making histone tails positive. E2F can work both to activate or deactivate):

RDP3
Rb (retinoblastoma protein: named after its dysfunction in cancers)
(E2F) <- activation domain
(E2F) <- DNA binding domain
____________________________________ <- DNA
e-box

thus separate fields of histone packaging and gene regulation got reconciled.

• Chromatin packing and gene regulation II

Soviet scientists discovered the story wasn't as crystal clear. euchromatin still had some lysines which were methylated, and positively charged. why? do they serve a function?

       CH3
        |
  CH3 - N+ - CH3
        |
K27     K9        K4
--------------------- N

Kouzarides at Cambridge came up with simple experiment. he hypothesized there was some protein which bound to them.

western blots and gene characterization showed that protein was HP1: Heterochromatin Protein 1:

\ /
/ \
  ^ chromoshadow domain (dimerizes with other HP1's chromoshadow)
^ chromo domain (binds all 3 methyl residues)

HP1 proteins compact DNA by cross-linking histones (or rather, whole nucleosomes):

      |                  |
      |---HP1--><--HP1---|
      |                  |
---histone1----DNA----histone---

HP1 always binds K9. other proteins with chromo domains were later found which bind specific lysines. E.g. unmodified K4: NuRD complex + HDACs + ATPase.

• More suppression mechanisms

yet more reconciliation between DNA folding and gene expression:

HP1, and other histone modification 'binders' are sometimes called 'Readers' (they read chromatin modifications)

SUV39H, and other histone modifying enzymes are sometimes called 'Writers' (they write modifications).

        t               t+1
   (SAM)(**HP1**)        (**HP1**)(SuV39h)
  (SuV39h)(HDAC)         |
        Rb               |
      (E2F)              |
      (E2F)              |
______________________histone____
e-box

how does a gene like BDNF know that it should be expressed in neurons only (or mostly)?

     HDACs                     histone methylase
      SIN3                   G9a
N Repressor Domain     C Repressor Domain
                  \   /
                   DBD
5'  _____Represor Element 1 (RE1)__|BDNF|_____ 3'

sometimes the DNA itself is methylated. these CpG items tend to be close to promoters, and amount of methylation correlates with suppression of nearby genes.

          CH3
           |
5' -------CpG-------- 3'

• Regulation by neuronal activity

        glu glu     Ca2+ Ca2+
          glu          Ca2+
         GPCR          NMDAr
--------|    |---------| |-----
--------|    |---------| |-----
        GDP+ɑβɣ        Ca2+
          ↓             ↓
         GTP+ɑ         ryanodine receptors in Endoplasmic Ret.
          ↓             ↓
cAMP (2nd messenger)    Ca2+      Ca2+    Ca2+
          ↓             ↓                 ↓
         PKA             Cam-Kinase IV     PKA
______________________________________________________________________<- nucleus
         CBP <- **can only bind to CREB after phosphorilation by PKA and CamKIV**
        (CREB)
  _______CRE_______|early response genes (c-fos,BDNF,Arc,AMPAN,Homer)|

patients born with Rubenstein-Taybi syndrome have single allele of CBP, severe cognitive decline.

also, ketone bodies inhibit HDAC (therefore promote more extended gene expression): keto diet.

consider LTP experiment

EPSP
|   ________
|  |
|  |
|__|
_________________ t

after rapamycin application, which inhibits mTOR:

{growth factors, neurotransm} -> {PI3K, AKT, MAPK} -> {TSC1, TSC2} -| {mTOR} -> CAP-dependent translation
                                                          _
                                                          |
                                                         AMPK (energy sensor, if ATP levels are low
                                                         (AMP is high), it inhibits cascade)

EPSP

|   _
|  | \
|  |  |
|__|   |____
_________________ t

• Sensation

Prof. Raunak Sinha's lectures.

• GPCRs

Nobel Prize in Physiology 1994, Nobel Prize in Chemistry 2012.

largest cell surface protein family. 4% of human genome
activated by diverse stimuli: light, odorants, nucleotides, peptides, neurotransmitters, chemokines, protein hormones
most ubiquitous drug target

basic structure of β2 adrenergic receptor (homologous with rhodopsin):

7 TM domains
amino NH3 exterior terminal (N terminal)
carboxil (COO-) terminal in cytosol (C terminal)

location of agonist binding site:

class 1: TM domains (3-6) (Ach, DP, E, 5HT, photon, olfactant)
class ii: N terminal (peptide hormones, somatostatin, met-enkephalin, orexin)
class iii: inside ball hanging from N terminal (mGluR, GABA_B)
class iv: (pheromones)
class v: frizzled and smoothened

2nd messenger pathways

Gs: Gɑ+GTP -> (adenylate cyclase) ATP -> cAMP -> PKA activation
Gi: Gɑ+GTP -> (adenylate cyclase) ATP -> cAMP -> PKA inhibition
Gq: (phospholipase C) PIP2 -> DAG + IP3 ->
- IP3 -> smooth ER IP3 receptor -> ER Ca2+ -> Calmodulin-dependent PK
- DAG -> PKC

evidence for G-protein pathway:

G-protein activation/deactivation absolutely requires GTP
non-hydrolyzable GTP analogs: GPT+Gɣ activates, GTP+Gβ inhibits
cholera toxin irreversibly activates Gɑ subunit
pertussis toxin inhibits

conformational change:

when receptor is activated, Gɑ opens and releases GDP, GTP enters cavity and takes its place.

regulation by exchange factors (facilitators). 2 kinds of G proteins carry GDP at rest:

monomeric G proteins: similar to Gɑ, hydrolize GTP -> GDP in cytosol.
hetero-trimeric G proteins (ɑ, β, ɣ subunits)

GPCR and Guanine-Exchange Factors catalyze GTP exchange to activate Gɑ.

G-protein has GTPase activity to terminate active state (GTP -> GDP + P).

cAMP-dependent protein kinase:

inactive PKA: 2 catalytic subunits are blocked by autoinhibitory domains of R (regulatory) subunits.

phospholipase C (PLC):

phosphilipid in membrane (PIP2)
DAG (still hydrophobic) -> PKC
IP3 (phospholipid head) diffuses and reaches IP3 receptors at endoplasmic reticulum -> Ca/calmoduline PK, etc.

diversity of G-proteins:

β2 adrenergic: Gs
Dopamine: Gi
mGluR, type 1 mAch: Gq

more diversity of 2nd messenger mechanisms:

in Ach muscarinic type 2, Gɣ actually has a role. it opens K channel directly. not as fast as ionotropic channels, but much faster than 2nd messenger cascade.

signaling to nucleus.

• photoreceptors

retina has more cellular diversity than the whole cortex.

Schultze's 1866 Duplex theory: rods and cones: proportion of cones/rods between species must mean one mediates dim vs bright vision.

very specialized neurons. 3 segmens:

outer: all about phototransduction, disks made out of plasma membrane foldings. non-selective cation (Na/Ca channel open by default). also, Na/Ca-K pump.
inner: normal intracellular matrix , nucleus and organelles
synaptic terminal: constantly depolarized and releasing neurotransmitter in dark

• opsin GPCR

chromophore (11-cis retinol, derived from vitamin A) accross TM in the middle of the GPCR/opsin's 7 passes, extremely thermostable.
photon (E=hv) transforms 11-cis retinal to all-trans retinal (straightens the molecule)

nomenctlature for visual receptors:

receptor: GPCR, opsin
ligand: chromophore, pigment, retinal
receptor+ligand: rhodopsin (rods), cone opsin (cones)

evolutionary diversity:

mouse: bichromats (blue, green)
primates: trichromats (+red)
zebrafish: zebrafish (+UV)
mantis shrimp has ~33 opsin transcripts

• phototransduction

each outer fold in cell has many opsin GPCRs.

2nd messenger isn't ATP->cAMP but cGMP->GMP.

photon -> Rh -> Rh* -> transducin (Gɑ + GTP) -> ...(neither AC nor PLC, but) ...
-> (Gɑ + GTP) + GAP complex + PDE protein (cGMP-phosphodiesterase: cGMP hydrolizer) -> ...
-> **cGMP** -> **GMP** -> ...

this depletes cGMP bound to cation channel, replacing it for GMP -> closed state.

Na/Ca channels (mostly Na) are ligand-gated. can be regulated by phosphorilation or calmoduline.

cones have faster kinetics.

• inactivation

recovering opsin and G-protein:

GRK1 (rodophsin kinase 1) phosphorilates rhodopsin
arrestin binds to phosphorilated rhodopsin, preventing light from hitting it
all-trans -> 11-cis

also, changes of PDE protein accelerates disgregation of PDE+GAP+Gɑ

• adaptation

turning on guanylate cyclase: opposite of channel opening by cGMP->GTP exchange

GTP -> GC+GCAP+Ca+ -> cGMP

when less Ca+ is available, concentration of cGMP increases. one of the main mechanisms for light adaptation.

There's a 2nd mechanism for adaptation: calmodulin acting directly on channel.

• rod vs cone

although overall mechanism is the same, rods and cones occupy opposite ends of the speed-sensitivity tradeoff:

rods: dim light (scotopic), sensitivity to scarce photons, slower response
cones: opposite

thermal stability is critical to reduce baseline noise

GPCR allows single photon detection to amplify signal. 1 rodopsin activates many surrounding G-proteins.

membrane proteins are densely packed in 2D membrane, membrane is high in unsaturated fatty acids, increasing probaility of reaction even further.

amplification:

1 rhodopsin -> ~100 G-transducin and PDE -> 1x10^5 GMP -> ~1000 closed channels

single-photon response is very reproducible. lifetime of molecule phosphorilation (inactivation) has very low-variability.

• Retinal transmission

if photoreceptors are constantly depolarized and releasing at rest (even if analog), how is neurotransmitter not depleted?

because of...

• ribbon synapse

thethered vesicles along a ribbon, 10-fold increase.
allows graded transmission of electric signals
support a continuous and high-hroughput mode of release
well-suited for rods and cones, which continuosuly release Glu
rod: single ribbon
cone: many ribbons

functional specializations of ribbon synapse

multivesicular release (not quantal minis).
L-type Ca channels. slower activation and deactivation kinectics to match multivesicular release. depolarized resting state allows these channels to work continuously

• triad synapse

signal transfer from photoreceptors to bipolars:

1 rod -> {2 horizontal, 2 bipolar}
1 cone -> {2 horizontal, 1 ON bipolar, 3 OFF bipolar}

There are ON bipolar and OFF bipolar cells. receptive-field wise, ON bipolars detect bright center, OFF bipolars detect dark center (aided by lateral inhibition to create edge detection, see below).

metabotropic signaling in inverting ON bipolar:

glu -> mGluR6 -> G-protein -> Gɑ+GTP -> inhibition of non-selective TRPM1 Na/Ca channel

ionotropic signaling in conserving OFF bipolar and horizontal:

Glu -> AMPA or kainate -> depolarization

• pathways

tight correlation between how deep the bipolar laminates into retina and function.

• rod

rod pathway features a lot of convergence. single RBC pools 20-100 rods

non-linear threshold (see computational neuroscience lecture notes). single-photon responses larger than 1.3 times average are retained. almost 100% of rod single-photon responses are ignored. nonlinear threshold increases SNR by 350%.

molecular mechanism: mGluR cascade. In darkness most of the cascade leading to TRPM1 is saturated. single-photon event is usually not enough to desaturate it.

• cone

fovea is very injective, whereas periphery displays degeneracy.

• horizontal cells

horizontal cell creates lateral inhibition: center ON + surround OFF, and vice versa.

downstream target: photoreceptors themselves (feedback from other photoreceptors in the surround area). delta-I Ca channel response curve is shifted to left.

proposed molecular mechanisms of feedback inhibition:

GABA hypothesis (doesn't involve Ca channel, acts directly in GABAr)
ephaptic hypothesis (involves Ca channel as target)
pH hypothesis (also involves Ca channel as target)

• color vision

horizontal cells also play a crucial role mixing cone signals for color vision.

color sensitivity curves are in fact very broad, there's a big overlap. color vision also depends on opponency of spectrally-different, mixed cone signals.

red,green opponents (red+ green- and viceversa)
blue,yellow opponents

• gap junctions

there's also electric coupling between different cells. this is evidenced by linear I-V curves during simultaneous recording. site of contact is axon terminals.

rod-rod coupling: alternative pathway for conveying rod signals at dusk/dawn for efficient coding of faintly backlit objects.
cone-cone coupling: improves the signal-to-noise ratio thus improving visual resolution.
rod-cone coupling: might be important for integrating rod signals into the cone pathway under mesopic light conditions.
horizontal-horizontal coupling: allows the dynamic enlargement of receptive fields up to several millimeters.

• smell (chemosensation 1)

humans can detect between 10k and 100k different odors.

• olfactory transduction

regular Gs pathway. endpoint is CNG channel (nonselective cation channel, as in phototrasnduction, but polarity is normal. Odor results in channel opening).

as in phototransduction, influx of Ca has an inhibitory modulatory effect in conjunction with CaM.

important difference: GABA/glycine is depolarizing, Cl- leaves cell (similar to brain during early development). this amplifies odor-evoked response. positive feedback.

Ca-activated chloride current

• Ca dual role

excitatory: rise in intracellular Ca concentration opens Ca-activated Cl channels in ciliary membrane.
inhibitory: canonical negative feedback on CNG channel (and other components of cascade, e.g. AC). adaptation: both short and long-term, both dependent on Ca.

• GPCRs

Buck and Axel, (Cell 1991) used degenerate PCR to amplify many sequences from olfactory epithelium and discovered a whole family of GPCRs to detect odorants: olfactory receptors (ORs). Nobel Prize in Physiology and Medicine 2004.

high aminoacid variability in transmembrane passes. Odorant binds to cavity as retinol in rhodopsin.

• ORs

highly distributed anatomical expression in olfactory epithelium.

single olfactory sensory neuron expresses single OR gene.

combinatorial response:

odor <-> OR mapping isn't injective though. there are more detectable odorants than ORs. each odorant depends on a different combination.

Valence roughly depends on whether compound is alcoholic or acidic (carboxilic acid). hexanol, heptanol, octanol, nonanol are perceived as pleasurable, whereas hexanoic acid, heptanoic acid, octanoic acid and nonanoic acid are disgusting.

coding in Olfactory Bulb:

spatially distributed epithelium neurons coding for same OR project to the same glomerulus in the OB.

• cortex

tufted cells and mitral cells project to layer-IV pyramidal neurons. many cortical targets:

AON
OT
PC
amygdala
Enthorinal cortex

cortical processing is poorly understood.

• vomeronasal organ and pheromones

TODO: bring info from UNAM lectures.

• taste (chemosensation 2)

papillae (bumps) contain taste buds, which in turn contain taste receptor cells.

types of taste receptors:

type I: ~50% of total. narrow morphology, they ensheath other taste cells and seem to have glia-like functions.
type II: ~33% of total. more spherical morphology. main mechanism of chemosensation, they use GPCRs too.
type III: differential expression accross tongue, scarce on anterior papillae, more numerous posteriorly. slender morphology.

primary taste qualia:

sweet: sugars (fructose, sucrose, maltose), proteins (monellin), aspartame, saccharin, sucralose.
bitter: Ions (KCl, Mg2+, Chinin, caffeine, denatonium)
salty: Salts.
sour: Acids.
umami: Proteins, monosodium glutamate, etc.

bitter threshold is lowest, to allow potentially dangerous substances to be detected.

taste	substance	absolute threshold
bitter	quinine	.000008 M
umami	glutamate	.0007 M
sour	HCl	.0009 M
salty	NaCl	.01 M
sweet	sucrose	.01 M

• salty

Beidler, 1953 argued that response of cat tongue was fast enough to use ionotropic receptors.

search for epithelial sodium channel (ENaC): Reverse-Transcriptomics PCR showed that subunits of ENaC, alpha, beta, gamma.

very straightforward mechanism: Na-gated ionotropic in membrane -> depolarization.

yet a portion of response (~20%) wasn't sensitive to ameloride. rest might be due to TRPV1 (metabotropic).

• sour

stimulus is acidification (pH decrease) inside the taste cell.

many channels have been implicated:

hyperpolarization-activated cyclic-nucleotide-gated (HCN)
Acid-sensing ion channels (ASICs)
Potassium (K2P) channels
H+-gated Ca channels
Acidic inactivation of K+ channels

most solid evidence: KO transgenic Otop1 gene. Hardly any response after KO.

when proton enters cell through Otop1, ΔpH blocks K channel (GIRK: K inward rectifying), which in turn leads to depolarization.

• sweet, umami, bitter

rest of flavors are due to GPCRs.

subunits first identified by screening for mRNAs preferentially expressed in mouse taste buds. novel GPCR expressed in small subset of receptor cells. taste receptor type 1 and 2 (T1R and T2R), hypothezised to combine as heteromeric polymers.

T1R distantly related to metabotropic glutamate receptors (mGluRs), vomeronasal receptors and GABA_B receptors. T1R forms heteromeric dimer with subunits T1R1, T1R2 and T1R3.

T2Rs are a family of ∼30 taste-specific GPCRs distantly related to opsins. short amino termini and may function as monomers or dimers.

evidence from matrix of subunits KO times flavor. responses are abolished.

• signaling cascades

main one: PLC patchway (Gq) -> Trpm5 (Na permeable) and Panx1 (ATP permeable) channels. Although not released using vesicle exocytosis, ATP is used as neurotransmitter.

others:

T1Rs (sweet, umami): AC -> cAMP -> PKA
T2Rs (bitter): NOS -> NO -> sGC -> cGMP

• coding

popular misunderstanding: there's no tongue map. responsiveness to all 5 main flavours is distributed across tongue, as with the olfactory epithelium.

individual cells are mostly tuned to single flavours, however. yet, over 90% of cells are at least partially responsive to more than 1 basic flavor.

specificity of whole cell buds is still debated. competing models:

labelled-line: each cell is tuned to single taste.
cross-fibre models
cells detect many flavors
cells detect single flavors, but they get mixed at level of afferent fibers.

combinatorial models say there are both generalist taste bud cells and specialist cells, so picture may be even more complicated.

primary sensory cortex is the posterior insula.

• somatosensation (mechanotransduction 1)

single dorsal ganglion root cell is responsible for both sensing and transmission to CNS.

myelination and thicker myelin sheet increases velocity of transmission (because capacitance is increased). Thermal receptors and nociceptors have small-diameter axons and unmyelinated (type IV) or thinly-myelinated (type III) fibers. Opposite case for touch and proprioception.

epidermis: nothing, neurally speaking.
dermis:
1. Merkel's disk SA1
2. free nerve ending
3. Meissner's corpuscle RA1
4. Hair follicle receptor
5. Pacini's corpuscle RA2
6. Ruffini's ending SA2

RA, SA: rapidly adapting, slowly adapting

All of them have low threshold for activation.

classification by location of innervation:

type 1: superficial. single fiber branches out into multiple receptors.
type 2: deeper dermis. single fiber innervates single receptor.

• functional properties

timing:

      ____________
 _____|           |____ <- stimulation
      |||         |||   <- FA (phasic response, Meissner, Pacini)
      ||||| | | | || |  <- SA (tonic response, Merkel, Ruffini)

intensity and location of receptive fields:

size of receptive field is proportional to depth. size of corpuscle is also increased at deeper levels.

| Merkel (SA1)
| Meissner (RA1)
| Pacini (RA2)
| Ruffini (SA2)
|
˅
(bigger receptive field)

• molecular basis

Coste el al. Science (2010) found the Na-Ca ion channel which is opened after mechanical deformation of cell membrane, after testing hundreds of cloned genes: the Piezo1 channel. also found in many internal organs, and even roots of plants. Nobel Prize in Physiology 2021.

Piezo1 (in HEK cells) is extremely long: 24 to 36 transmembrane domains.
Piezo2 is expressed in RA dorsal ganglion neurons.

other channels implicated in mechanotransduction (C. elegans, etc.): DEG/ENaC, TRP, TMC1, TMHS, TMIE, K2P.

Piezo1 structure and gating:

3 blades
each blade/subunit has intra latch, beam, clasp and the piezo repeats
gate is probably deformed mechanically when membrane is deformed.

• temperature

• nociception

• proprioception

TODO: bring info from UNAM lectures

• hearing (mechanotransduction 2)

• hair cells

outer hair cells differ from inner hair cell. inner cell is involved in boosting/amplification.

separated by tunnel of Corti.

supporting cells:

Deiters' (below outer and inner hair cells).
Pillar cells (make Corti tunnel)

endolymphatic vs perilimphatic

E. Potential = +100 mV, high in K+.
P. Potential = normal extracellular potential, right outside hair cells.

receptor potentials are perfectly amplitude-locked and phase-locked to some frequency component of heard stimulus.

• mechanotransduction

each cells can have 50 to 100 stereocilia (modified microvili). tallest one is called knocilium.

cross-section:

Outer Hair Cell:

|
||
|||
cell
body

Inner Hair Cell:

|
|
|||

cohesion when bending is produced by inter-ciliary links of different kinds.

• MET channel

there are also tip links, at the cilia tips, making HC extremely sensitive to deflection (~1 nm). tip link opens channel (MET channels).

When opened, K+ from endolymphatic fluid enters the cell.

location of MET channels:

Beurg at al., Nat Neuro (2009) showed that Ca fluoresence mostly increases for smaller rows of stereocilia, not the tallest one, suggesting there are no channels facing on the direction of negative (closing) deflection.

(diagram is missing tip links):

       /
   /* /
/*/  /  ---> positive deflecton
(channels = *)

identity of MET channel:

TMC2 dominates in early stages of development, then TMC1 takes over. KO studies show transmission is obliterated without TMCs.

Ca adaptation:

tuning curves (I) as function of deflection (Δx) can shift to avoid saturation, based on Ca influx, creating greater dynamic range.

MET complex:

TMC1/2
TMHS
TMIE
PCDH15 (adapter between MET and tip link)

mutations in hair bundle proteins can cause deafness (e.g. Usher's syndrome).

• frequency selectivity

basilar membrane has 2 key properties to allow it to perform Fourier decomposition, depositing the energy from the relevant frequency at the right place as the impulse travels along the cochlea (in the oval window -> round window direction).

width: wider towards apex -> lower frequencies towards apex
stiffness: more flacid towards apex -> lower frequencies towards apex

basilar membrane is like a flipper for swimming!

• tonotopic map

as with subjective musical pitch, frequency is an exponential function of distance in the cochlea.

passive harmonic oscillator model:

k: stiffness
m: mass
c: damper

resonant frequency depends on ratio k/m.

there are 2 pivot points involved: one at the tectorial membrane and one at the basilar membrane.

typical HC tuning curve:

dB SPL
|___      ____
|   \     |
|    -_  /
|      ||
|______________ kHz

• tuning

the mechanical (velocity and displacement) properties of the basilar membrane predict almost perfectly the neural response of specific fibers. that is, tonotopy is also preserved by the inner HC action potentials.

again, not only are hair cells at a given distance uniquely selective to some frequency, their firing rate also follows frequency and phase of their Fourier component:

mechanical wave:
       /\
  / \ /  \ / \
---------------> t
firing rate
|
|      |
|      |
|  |   ||  |
|  ||  ||  ||
---------------> t

• Outer Hair cells

it's believed that more numerous array of OHCs play active role by amplifying local vibration of basilar membrane.

destruction of OHCs or impairment with antibiotics shifts and lower amplitude of tuning curves of IHCs.

current model (notice feedback loop by OHC):

sound stim -> basilar vibration -> IHC excitation
             |               |
         force by OHC   OHC excitation
             |               |
             <- OHC modulation

transgenic studies have implicated protein called prestin in the amplification mechanism of OHCs. Cl- anions bind and unbind to prestin depending on voltage.

In this view, OHCs not only sense the basilar motion, they add motion. cell length has been shown to change with OHC depolarization, like muscles acting on the membrane.

                       |
                      ||
    |                |||
   ||                ---
  |||               |   |
  ___               |   |
 |   |              |   |
_|Cl |_ <-prestin ->|   |
_|   |_             |   |
 |   |              |   |
  ___                ___

• ribbon synapse

presynaptic contains many ribbons, each contacted by postsynaptic afferents.

HCs have simple round morphology without neurites.

IHCs don't fire action potentials either. it is thought that analog voltage allows for greater bandwidth.

MET -> K+ influx -> depolarization -> Ca2+ channels -> Ca2+ influx -> ribbon Glu

big vesicle pool allows for wide dynamic range. only mild adaptation necessary.

no synaptotagmin Ca2+ sensor. Otoferlin instead. there may not even be a SNARE complex for vesicle fusion.

Cav1.3 channel (type L): long lasting, high-voltage activated.

• vestibular

TODO: bring info from UNAM lectures

• Extra: Epilepsy

• Physiology and anatomy

1/26 people will have seizures in their lifetime, more than Alzheimer's Parkinson's and Huntington's summed together.

not a single disease, but a syndrome of disorders, described by numerous symptoms. seizures are the defining symptom.

space of seizures:

psychological dimension: simple vs complex
anatomical dimension: focal/partial vs generalized

	simple	complex
partial/focal	simple partial	partial complex
generalized		tonic/clonic (grand mal), absence (petit mal)

• EEG

absence: high amplitude, 3 Hz on all channels. stereotypical spike-and-wave discharges
etc.

• etiology

initial insult
genetic mutations (channelopathy)
epigenetic changes (channel expression, or channel regulation)
unknown

• proposed mechanisms

imbalance between excitation and inhibition
timing issues?
bastardization of normal rhythmic mechanisms?

• Genetics

epileptogenesis: process by which seizure propensity increases over time.

458 genes were differentially expressed on day 1 post-Status epilepticus across all models
MAGIC: predict transcription factors and regulators
master regulators downregulation (in order): EZH2, RAD21, SUZ12, CTCF, GATA2, CTBP2, SMC3, POU5F1, NANOG

EZH2 in dentate gyrus' granular cells increases levels for ~20 days across status epilepticus. EZH2 works by shutting down other genes, it's a methylase.

• EZH2's role

EZH2 keeps seizure threshold high, reducing probability of seizure. EZH2 is protective.

• EZH2 targets

WT + no seizure: normal threshold WT + seizure: lowered threshold MT + no seizure: normal threshold MT + seizure: hyperlowered threshold

genes were clustered to discover similar genes. regulatory network seemingly centered around STAT1/2/3 and IRF1. cluster turned out to represent genes associated with neuroinflammation. STAT is a transcription factor, but it lives in cytoplasm of neurons:

membrane receptors -> JAK -> STAT

• STAT3-EZH2 pathway

EZH2 controls STAT3 phosphorilation by inhibiting JAK1.

membrane receptors -> JAK -> STAT
                       T
                      EZH2

treatment could be attempted using drug CP690550, an already FDA-approved drug (for different purposes).

Cellular and Molecular Neuroscience - University of Wisconsin-Madison

• Cellular Excitability

• Resting potential

• Single ion: Nernst

• Multiple ions: GHK

• Ion channels and patch-clamp

• characterizing ion channel with patch clamp

• Ion channels

• Selectivity

• Gating

• Single channel kinetics

• multiple states

• Macroscopic currents

• Na-channel dynamics (depolarization)

• K channel disease

• Na inactivation (HH)

• Na inactivation (Ball-chain)

• Activation

• Presynaptic Mechanisms

• Classification

• Transmitter apparatus

• Quantal release

• pre vs post

• Calcium

• Channels

• Release

• SNARE complex

• synaptotagmin

• GPCRs and G-proteins

• GPCRs

• Classification

• G-protein

• kinases

• Plasticity

• potential broadening

• Facilitation with Ca

• Depression with Ca

• potential slowdown

• habituation

• case study: optogenetics

• CRE-lox driver:

• GECIs

• GEVIs

• channelrhodopsin

• Postsynaptic Mechanisms

• Pharmacodynamics 101

• EC50 and KD

• Hill equation

• Excitatory transmission

• AMPA/Kainate vs NMDA

• Inhibitory transmission

• Ligand-gated channels

• Pentameric LGICs

• Tetrameric iGluRs

• Shaping of synaptic potential

• Dendritic integration

• Temporal summation

• Spatial summation (cable theory)

• Firing patterns and oscillations

• single spike

• bursting

• pacemaker cell

• Plateau

• LTP/LTD, pattern storage & recall

• LTD (reverse process)

• Cascades and Transcription

• Transcription and splicing

• Transcription

• Splicing

• Enhancers

• Chromatin packing and gene regulation

• Chromatin packing and gene regulation II

• More suppression mechanisms

• Regulation by neuronal activity

• Sensation

• GPCRs

• photoreceptors

• opsin GPCR

• phototransduction

• inactivation

• EC₅₀ and K_D