Your browser doesn't support the features required by impress.js, so you are presented with a simplified version of this presentation.

For the best experience please use the latest Chrome, Safari or Firefox browser.

Section A

Potassium Channels

  • (2009) Ion channel blockers are important tools that reveal the function of potassium channels. How has the use of a potassium channel blockers furthered our understanding of potassium channel function and structure? Illustrate your answer with evidence from the use of at least two such blockers? BLOCKERS

  • (2010) Potassium channels are well adapted to conduct potassium rather than other ions. Discuss how the structure of the channel enables this process. SELECTIVITY

  • (2011) Both the amino terminal inactivation domain of the channel protein and quaternary ammonium salts interfere with the function of A-type potassium channels by acting from the cytoplasmic side of the channel. Describe the similarities and differences of their actions. BLOCKERS AND INACTIVATION

  • (2012) Voltage-gated ion channels are activated by changes in electrical potential differences near the channel. Describe the structural and dynamic properties leading to the opening of a voltage-gated potassium channel in response to voltage changes. Support your arguments by explaining at least two methods used to study these changes. VOLTAGE-GATING

  • (2013) Quaternary ammonium ions are important tools to postulate the existence of an intracellular potassium channel gate. Their intracellular application can lead to the “foot in the door” effect and the trapping of blockers behind closed activation gates. Discuss. BLOCKERS AND INACTIVATION

  • (2015) Critically discuss the following statement “Toxins bind to potassium channels via an induced fit mechanism”. BLOCKERS

Experimental Techniques

SCAM (substituted cysteine accessibility method) - Shows whether or not residues can be accessed by thiosulfonates which bind to the cysteine sulphide
Heterologous Expression (Xenopus oocytes)
Pharmacological blocking etc
Site directed mutagenesis

Channel Structure and General Information

Ion Channels

  • Integral membrane proteins forming aqueous pores through membrane
  • Allow ions to flow down electrochemical gradient (passive).

Role in Electrical Equivalent Circuit

  • EC gradients of ions equivalent to batteries in parallel
  • Ion channels ~ resistors in parallel
  • VGIC are variable resistors

These parallel resistors and their placement are the main determinants of neuron excitability and firing pattern

The presence of VG channels in the membrane allows transverse flow of charged ions in response to changes in membrane potential
Cells are able to generate rapid signals which can propagate and encode much greater information than equivalent molecular forms of communication (paracrine endocrine etc) would allow

Type-A K channels (voltage gated, like Shaker) are responsible for the generation of the action potential while others are responsible for setting and stabilising the resting membrane potential (such as IRK and Tandem-Pore channels)

Sequencing the first channels

Sodium Channels (and Ca2+)

  • Trypsinated and sequenced by Numa et al to produce a probe library, then eventually geenerated protein sequence. - Then used hydrophobicity polot to identify TM regions.
  • Cloned and expressed Na channel in xenopus oocytes. No confounding channels in membrane to contaminate results.
  • 4 homologous domains (~continuous tetramer)
  • 6 TM regions in each one
  • S1-S4 = voltage sensing region.

Potassium Channels

  • First variant examined = shaker; mutation in drosophila caused shaking of legs under anaesthesia, and more aberant behaviour.
  • Encoded protein 1/4 the size of Nav channel gene
  • Eventually understood to form a homotetramer (though can be hetero…..nothing wrong with that)
  • Homologous to a single sub-unit, showing the 6 TM domains.
    Functional domains
  • Pore structure (shown by Mackinnon, Kcsa crystallography)
    • 12 A length, 2.8 A diameter selectivity filter
    • Wide aqueous chamber, 10 A long
    • Long (20 A) narrow hydrophobic segment
  • Helix dipoles to stabilise to overcome electrostatic destabilisation of ion in pore at centre of bilayer
  • S4 = voltage sensor (along with S1-3 maybe, in paddle and other models)
    • Has positive AA ~ every 3 residues
  • P - loop = ion selectivity filer plus turret and pore helix
  • S6 = Pore-lining
  • Teepee architecture; packing of P-loop pore helix between “poles” of S5 and S6
  • Intracellular N-terminus = contains the ball, key for N-type (rapid) inactivation.
  • N.B. /Beta s.u.-s are extra-channel proteins which associate to the channel and modify function


Some potassium channels show inwardly rectifying currents

Attributable to polyamine block (e.g. spermine) and Mg block which both show strong voltage dependence
Minimal currents generated at depolarised potential rather than linear IV relationship
Structure also contributes to preference for inward flow of K+ ions, to reverse hyperpolarisation and restore the resting membrane potential


Calcium-Activated such as BK (CBTx sensitive)
A-type = voltage-gated transient outward-rectifier K+ channels

Mackinnon and crystallography of the K channel
4 subunits of a tetramer
in KcsA, 2 TM s.u.s with inner helix, outer helix, pore helix and selectivity filter
12 - narrow selectivity

Glutamic acids at channel mouth
then into wide hydration chamber
then into selectivity filter
then experiences negatively charged aspartic acids and enters the extracellular space

Permeation and Toxins


Charybdotoxin, TEA, Kappa-Conotoxin?

Scorpion Toxins

  • CBTx (Scorpion toxin)

    • is a pore occluder of Calcium activated K channels such as the BK channel
    • Used by MacKinnon and Miller to implicate P-loop in the pore, by using site-directed mutagenesis to modify affinity
    • Also used sub-unit mRNA mixture to form heteromers with a single CBTx sensitive unit
    • Generated 35% channels with sensitivity from 10% mRNA encoding sensitive sub-units
    • Based of random mixing; 1-(0.9)^4 = 0.34
    • Due to size cannot penetrate pore however
  • kappa-Conotoxin

    • Induced fit in shaker

Induced fit?

2002 Erikson

  • Computational modelling indicates at least two possible agiotoxin Shaker complexes, based on high temperature molecular dynamics and accounting for side-chain flexibility of AgTX
  • Blocks Shaker at external vestibule with high affinity
  • agioArg 24 binds to negative Aspartates
    Appears to dock in channel and then prevent opening through flexible side chain interactions??? wiki, unsupported

Lange 2006

  • Used high res solid state NMR spectroscopy to show that kaliotoxin binding to a chimaeric Kcsa-Kv1.3 channel is associated with significant structural rearrangements.
  • Previous studies using solution state NMR indicated rigid preformed sites

    Banerjee 2013 (crystallography of K channel with bound CBX)

  • Banerjee however continue to support rigid preformed sites for CBX binding.

  • Used x-ray crystallography
    • Lys27 extends and blocks the pore, rather than inducing a change which seals the pore
  • Binds to carbonyl backbone, behaving as a tethered surrogate cation
  • Highly rigid with concserved cysteine sequences generating sulphide bonds
  • Argument partly based on energetic; high affinity binding if energy requirements are significant seems unlikely
  • CTX binding has been shown to be voltage dependent, though this does not appear to be due to structural changes but rather destablization of toxin-receptor complex by permeant ions entering from the intracellular side.
  • Additionally crystal structure of paddle chimera-toxin complex and sole channel do not differ significantly
  • Key structural location of Cysteines and Lys are highly conserved between Pore-blocking scorpion toxins


  • Changes in NMR observed with kaliotoxin assumed to be due to changes in dehedral angles however could be due to electrostatic changes
  • High affinity toxin binding to channels as binding free energy is not ‘spent’ bringing about a protein conformational change.
  • Single mutations in the ‘toxin receptor’ region of the channel can drastically alter the affinity for the channel by disrupting the good fit

Tetraethylammonium (TEA)

  • Now commonly used for reversible K channel blockade
  • Quaternary ammonium, selectively blocks voltage sensitive K channels. AP termination due slowly to Na inactivation and leak currents. No after hyperpolarisation.
  • Beyond role in probing external and internal structure to show the P-loop double-passes the membrane, has also been used to probe oligomeric state of channel
  • Mixing subunits from two channel type with particularly different inactivation kinetics and TEA sensitivity produced mix; inactivation kinetics of one type and TEA sensitivity of the other (RCK4 and RCK1 respectively)
  • Support for tetramer model now confirmed by crystallography

Binds to internal side of selectivity filter within the aqueous chamber; blocks the dehydration transition site indicated by analysis indicated with caesium (has lowed dehydration energy and so has better stability at the transition site).

Internal sensitivity is voltage dependent due to block of intracellular side by ball-and-chain preventing TEA acccess

And Inactivation; Foot in the door

(and difference between Kcsa and Shaker??)

Importance of extracellular block mechanism; demonstrating C-type inactivation (studied often in Shaker or Kcsa chimeric channels with removed N-terminals to negate N-type inactivation)
Depends on same residues as TEA binding (477 Tyrosine or Phenylalanine aromatic residues
Thought to prevent “pinching” of pore through interaction with hydrophobic residues
TEA inhibits slow inactvation in Wt shaker channels



  • 4amines compete with inaactivation gate; bind to the same site
  • Both displaced by high ExC K+ mechanisms
  • Binding site identified in internal face of selectivity filter and in the aqeuous chamber

  • However “snaking of gate does not fit with quantitative studies identifying a binding site on the intracellular face of the pore

  • 3 states? active, preinactive and inactivated?


  • Internal binding appears to block the pore by producing a conformational change in the pore similar to that produced by low K concentration, likely due to loss permeant ions within the selectivity filters;
  • Normally in low ion concentration permeant ions alternate between the 1 and 4 carbonyl; when at 4 the lack of internal elements leads to a change in the external vestibule
  • Supported by the finding that high K+ concentrations antagonise the occurrence of C-type inactivation
  • Effect on c-type in-activation; internal binding accelerates
  • K+ loses bottom half of dehydration as it enters, through specific residue
  • Similar site on the exit of selectivity filter allow rehydration into the aqueous chamber?
  • Through Idea that TEA near internal end of the selectivity filter prevents transition of K+ ion into rehydrated state? retain at end? does not transition to top point to prevent closing?
  • Rigidity still supported; idea that “collapse” is only at the mouth


Ahern et al 2006 (questioning Kcsa orientation being applied to Shaker)

Tested for cation-pi in Shaker channels using in vivo nonsense suppression; increased fluoro of Phe side-chains at position 449
Minimal steric disturbance
Varying degrees of fluorination of aromatic side-chain to withdraw pi electrons and weaken cation-pi progressively, without altering hydrophobicity
Also progressively increased K(i) for TEA block of Shaker

In follow up 2009

Demonstrated that N-type inactivation of Shaker channels involves collapse of the outer edge of the external vestibule

  • T449 mutation to F; produced slowed C-type inactivation along with a spring in the door effect of TEA
  • Additional v438 mutation to A speeds inactivation; and makes TEA have foot-in-the-door
  • V438A has consequences on the selectivity filter and reorients aromatic ring?

    • Similar mutation in Kcsa produces decrease in TEA affinity, increase in rate of slow inactivation, and spring in the door to foot in the door
  • Explained spring in the door in T449F mutants using progressive fluorination; enhanced rates of inactivation

    • Weakened electrostatic repulsion of four aromatic ring, so weakened rigidity of external face?
  • once xxx changed sign? TEA become foot in the door once again
  • Spring on the door relies on cation-pi interactions and explains foot in the door of Kcsa
  • Occurs through TEA attracting Phes into inactivate configuration
  • Lost not only through fluoro to remove cation pi but also V438A which may reorient en face -> edge on
    • Though there may also be other structutal changes.
  • Atomic structure of inactivated channel not yet shown?? though pinched structure proposed


MacKinnon and Miller 1989 - Shaker (A-type)

  • Charybdotoxin was known to be a pore-blocking inhibitor of potasium channels
  • Occludes pore of various K+ channels by binding to one of 4 independent, overlapping binding sites
  • Used site-directed mutagenesis of negative residues in Shaker channels to disrupt CBT block of K currents
  • Negative because CBT is highyl electrostatic
  • Replaced Glu422 with glutamine or lysine; lowered toxin affinity (no effect of Asp replacement)
  • As CBT is pore-blocking, this indicates that 422 is at the mouth of the pore
  • 422 is at the start of the S5-S6 linker region
  • Showed in later experiments that modifying other residues in linker affected ion permeation
  • Goldstein and Miller 1993 showed that block is through binding of one toxin molecule allowing the CBT Lys (basic) residue to enter the pore and coordinate with K ions or the Tyrosine residue backbone to stop further passage of K ions

Further Mutagenesis Identifies the P Loop

Yellon et al (MacKinnon), 1991

  • Alterations at various residues within the P-loop were found to weaken internal TEA block
  • Placed these residues on the internal side of the pore
  • Altering residues further along chain showed effect on external TEA blockade
  • Indicates that linker spans the membrane
  • Modified residues constitute internal and external binding sites for TEA
  • TEA block is voltage dependent, suggesting that these residues lie near the pore ends

Chimeric Mongrels

Hartmann et al 1991

  • S5-S6 linker (the P loop) was identified as a potential for being the pore
  • Generating a chimera, inserting the P loop from NGK2 into DRK1 should confer pore channel properties.
    • DRK1
      • Sensitive to internal TEA
      • Slightly IR IV curve
    • NGK2
      • Single Channel Conductance 3x that of DRK1
      • Sensitive to external TEA;
      • Not IR
  • Expressed in xenopus oocytes
  • Single channel recordings showed that chimera DRK1 showed NGK2-like properties; 3fold conductance and external TEA sensitivity


K Channel Selectivity

  • showed structure of selectivity filter in KcsA using crystallisation, stabilising channel with antibodies
  • Showed water-filled central cavity
  • Narrow end of pore is near extracellular mouth
  • Conserved TVGYG sequence in the P-loop is key to the K+ selectivity
  • Carbonyl backbone of AAs form surrogate water cages of K+ ions
  • Hold K+ in square antiprism
  • Spacing is perfect for K+ ions (but not smaller Na+) to shed hydration shell and be stabilised by ions.
  • Key of rigidity to prevent collapse of the filter to accomodate Na+
  • Pushed through the pore by subsequent K+ ions
  • Four sites of coordination; bind K+ alternately (1-K,2-Water,3-K,4-Water) or (1-Water,2-K,3-Water,4-K)

Zhou et al, (MacKinnon) 2001

  • Showed high electronegative extracellular face due to Glu and Asp side chains from turret
  • Showed high and low K+ filter structures
  • Loss of ion leads to twist of P-loop and occlusion of pathway by Glycine
  • Reorgansied hydrogen bond network, prevents conduction from extracellular face
  • Role in C-type inactivation evidenced by voltage independent Shaker inactivation; plus inactivation in Ball-and-chainless mutant

Data from Review

  • Recent data indicates that carbonyl atom atms fluctuate by 0.4A, (nearly the difference between Na and K ionic radii)
  • Snug-fit theory needs work
  • Non seletive channel with TVGDG
  • Mutagenesis to make TVGYG

Comparison to Ca channels

  • Rather than using steric element, developed an “affinity” mechanism
  • Selectivity is ion-concentration dependent
  • Sodium ions are able to pass through at low Ca concentrations
  • However binding site has higher affinity for Ca than Na
  • So requires another Ca to displace bond Ca


N-Type Inactivation

  • Classed due to rapid inactivation
  • Shown to occur intracellularly
    Zagotta, Hoshi and Aldrich (1990)
  • Applying trypsin (serine protease) to cytoplasmic side of patches with Shaker channels, using inside-out patch clamp
  • Inactivation of current following voltage step was depleted; open longer and more frequently.
    Did not occur with extracellular application of trypsin

Mutagenesis of the Ball and Chain

Round 1

  • 22 AA deletion at N-terminal
  • Indicated as contains many trypsin cleavage sites
  • Disrupted inactivation

Round 2 - Lets get selective

  • Within 22 AA domain deleted previously there was a series of charged residues
  • Changing to neutral residues reproduced changes in inactivation

  • Developed idea that this 22AA domain somehow plugged the intracellular face of the pore,

  • Charged residues were key to interact and bind with charged residues in the pore.

What was the purpose of the processding residues?

  • Deletion of some of the 63 residues following the ~20 AA ball region potentiated inactivation
  • Degree of potentiation proportional to number of residues deleted
  • Insertion lead to slowed inactivation.
  • Armstrong and Bezanilla developed Ball and Chain model supported by these findings.


Zagotta, Hoshi and Aldrich et al 1990 Paper 2

  • Chain theory well supported by insertion/deletion of residues
  • To confirm ball theory, they generated Shaker channel mutants missing the N-terminus
  • Inside-out patch clamp showed no fast inactivation
  • Restored inactivation following depolarising voltage steps using 20 AA peptide from N—terminal present in solution
  • Restored in a concentration dependent manner (akin to ligand and binding site dynamics)


  • Sodium channels with hinge and lid model;
    • in fact similar to ball and chain however blocking residues are in intracellular loop and not at the N-terminal
  • P-type involves movements of pore-lining structures near the extracellular end of the pore
  • C-type; follows P-types
    • Stabilization of S4 the voltage sensor
    • C-type inactivation requires strong hyperpolarization to revert.
    • Prevents reactivation

TEA and Slow inactivation

Choi et al 1991

  • Shaker channel shown previously to lose rapid inactivation following removao of N-terminal
  • However still inactivates during prolonged depoarization (seconds)
  • Slow inactivation is blocked in extracellular TEA
  • While fast inactivation is sensitive only to intracellular application
    • As if competing with the binding of the inactivation particle
  • Shows it is a distinct mechanism


  • Tiny outward current observed prior to Na current
  • Called the gating current
  • Due to conformational change in channel from charged residues moving through stationary electric field (the membrane)
  • Opening is voltage dependent (ratewise)
  • S4 has repeated 1/3 Arginine and only seen in VG-channels
  • Initial studies (Papazian) struggled as they could not determine whether S4 point mutations affected channel stability or voltage sensing
  • Lu et algenerated chimera of KcsA (not VG) with S1-S4 of a Kv channel; conferred strong voltage sensitivity
  • S1-S4 domains have alsio been identified on voltage sensitive proteins which are not channels (e.g. Ci-VSPhosphatase)
  • Mackinnon et al further examined S1-S4 domain; found conserved helix-turn-helix domain (S4-S3b) they called the voltage sensor paddle
  • Distortion in crystal structure showed flexibility between “paddle” and rest of channel.

The Paddle Model

  • Flex in crystal structure lead MacKinnon et al to believe that conserved domain (S4 and S3b) region moved through membrane like a paddle, pulling open the pore.
  • Used Fab to stabilise, allowing crystallisation
  • Also used Fabs to inhibit the channel, using intracellular vs extracellular exposure to identify location of S3b and S4
  • Extracellular inhibition required depolarisation, suggesting paddle moves from intracellular to extracellular side
  • Additionally used tarantula toxin (binds S3S4 residues) on extracellular side
  • rate of inhibition increased with frequency of membrane depolarization
  • Highly energetically costly, as moving positive S4 arginines through lipid domain would be unfavourable.
  • Main difference to other models
  • However there is research to show that Arg residues do not pass through a wholly proteic environment; must be some exposure to lipid environment.
  • Also there is evidence using SCAM to access S3 at different voltages to show that the S3 domain continues to be accessible to methanethiosulfonate (MTS) reagents (bind to cysteine sulfide)

FRET and Rotation

  • Spiral track of +ve Arg every 3 residues on S4
  • By labelling S4 and a nearby residue with fluorescent labels that require close proximity to transfer and shift from yellow to red (if close second label receives resonant energy from first).
  • Time course of fluorescence change parellels time course of the gating current
  • Verifies that S4 movement reflects channel gating
  • Also shows rotation of S4 bringing residues into proximity
  • Suggests a corkscrew path, helical twist with axial translation moving Arginines through accepting “groove”? Ratcheting mechanism?
  • Also possible without axial translation; rotation simply moves Args and exposes to membrane
  • Would explain the gating current
  • However axial translation potential explains extracellular Fab inactivation shown by Mackinnon et al

  • Additionally FRET with lipophilic acceptors in the membrane did not support the idea that S4 movement spans the full thickness of the bilayer



  • (2012) Explain concisely what is meant by a ‘voltage clamp’ and how it was applied by Hodgkin and Huxley to the study of the action potential of the squid giant axon. What are the essential observations which make the Hodgkin and Huxley model of the action potential plausible and how have their predictions been confirmed?
  • (2013) In the year 3315 it is discovered that plants on the planet Squalon contain cells that transmit all-or-nothing electrical signals along their processes. Electrophysiologists find that the resting potential (voltage inside the cell minus voltage outside the cell) is +60mV, and ion substitution experiments determine that it is produced by high resting permeability to Na+ ions. What does this tell you about the transmembrane concentration gradient for Na+? When all-or-nothing signals occur, the voltage inside the cell moves negative for few milliseconds by about 100mV, reaching -40mV. How might this negative voltage excursion occur? What experiments would you perform to determine the ion movements that underlie this curious negative voltage pulse?
  • (2014) In previously un-studied type of neuron, lh channels are found to be activated with an exponential time course (and time constant of -100 ms) when the cell membrane is hyperpolarized. Activation occurs over the voltage range from -45mV to -80mV. The channels conduct mixture of Na+ and K+ ions and have reversal potential of around -20mV. Sketch the currents you would expect to see in voltage-clamp experiment were you to hold the cell at -30mV and make hyperpolarizing voltage steps of 300 ms duration, starting at -40mV and increasing in 10mV increments down to -100mV. How would you adapt Hodgkin and Huxley’s formalism to describe such a current? Why might it be useful in neuron to have current with these properties?

Lateral Knowledge

The HCN Channel and others relevant to Qs

Hyperpolarization-activated cyclic nucleotide-gated channels are intermembrane proteins that serve as nonselective LGIC
Help to generate rhythmic activity in heart and brain cells
Responsible for Ih or If (for funny) current

History and Method

  • Intracellular recording of Action Potential associated currents;
  • Electrode inside and outside of squid’s giant axon
  • Measure potential difference across
  • Then use voltage clamp to set voltage steps, and kept constant using feedback current to negate the change in VM produced by ion flux
  • By recording the feedback required to maintain voltage step, it is possible to determine changes in membrane potential which HH showed were due to ion flux.

The Voltage Clamp

  • Use intracellular and extracellular electrodes to measure membrane potential
  • Sent to Feedback amplifier which also receives a command voltage from the signal generator
  • Subtracts differences which is amplified and sent into the axon to maintain voltage
  • Changes in current required to clamp voltage indicates changes in membrane conductance in response to change in potential

The Process

  • Series of 5 papers
  • Established function of neuron membrane at rest and experimental technique
  • Identified the effects of changes in sodium concentration on AP, and resolution into sodium and potassium currrents (Vstep to ENa).
  • Described response to voltage step on action potential and ionic conductance
  • Outlined the inactivation process terminating increase in sodium conductance
  • Integrated into a single mathematical model using terms which were later shown to be VGIC

The Assumptions

  • Gates on the channels are charged
  • Gate is either open or closed (rate of opening incredibly rapid)
  • Ion flow direction and gating are independent

Charged gate assumption was made to confer the voltage sensitive changes in ion conductance observed.


  • The second assumption was supported by Sigworth and Neher 1980
  • Used patch-clamp to record from single channels
  • Showed that changes in current are step like
  • Random due to random activation of channel (Depolarisation does not activate; increases P of opening)
  • Summation of these random openings when P(opening) was high leads to the whole cell current.

The Model

Gating and Energy

  • Changes in internal charge will make it energetically favourable for channel to open
  • Negatively charged h-gate closed by more positive intracellular medium
  • Positively charged m- and n-gates opened by positive (repelled by like charge)
  • Additionally activation energy differences could explain differences in activation time for rapid sodium and slow-activating/slow-inactivating potassium channels
  • Also key in HH theory was the observation of the threshold, indicating delayed reversal mechanisms

The basis

Extolled in the early papers

  • Depolarisation due to Na flux suggested by depolarisation past 0 mV towards ENa.
  • Also replaced sodium with impermeant ion to show that depolarisation no longer observed;
  • Also used voltage clamp to 55 mV (ENa) to negate driving force and showed that delayed hyper polarisation was all that was observed.

The Formulae

Number open = n= α/(α + β)

At steady state (Vm is constant) dn/dt (change in number of open channels) is constant. Though there may be initial variation in ion flux this will eventually balance as ion channels inactivate etc.

α and β (rate constants of opening) are voltage sensitive.
In response to a voltage step,

n_t=1−e^((−t)/τ) τ=1/(α+β)

  • Predicted kinetics sugggests exponential currents produced in response to a voltage response
    HH found that the observed Ik varied with t4 and so assumed that there were 4 activating n-gates.
    • (importance of gate independence in the model)
  • This is incredibly close to the truth shown through structural and functional analysis;
  • the 4 Kv channel sub-units each have their voltage sensing gate. Though they do not function independently, it is still impressively close to the truth.

Sodium channel kinetics on the other hand are less straight forward due to rapid inactivation. Modelled using m3h (inactivating h gate).


The discussed m/h/n are dynamic variables which denotes the probability of “opening”; m and n are positively related to membrane potential while h decreases with depolarisation. h and n are both slow in their kinetics however. These factors are essential to the action potential.

These factors with their voltage dependence are used to express conductance in voltage dependent terms

The final equation is obtained by combining conductance questions with the equation for current (factoring in driving force)
Crucial to include static leak current
Gives an equation for change in membrane potential over time in response to external current


Explains the absolute refractory period observed;

  • h is slow to return from ~0 during depolarisation and so despite high m the Na channels will remain closed.

Also afterhyperpolarisation supports trains by reducing the threshold;

  • once Kv are closed, h still remains higher than at rest
  • Means that depolarisation, increasing m will lead to a greater increase in gna.


  • HH theory model explains Anode Break Excitation phenomenon where terminating a hyper polarising current to a neuron can trigger the generation of an action potential
  • Mass hyperpolarisation closes K channels, and h becomes ~ maximal.
    • M gates also closed so no Na flux.
  • When hyperpolarisation ends, the few m gates spontaneously activating will produce a greater increase in gNa, along with increased driving force leading to significant INa and crossing of threshold (while K channels remain inactivated).

Sodium Channel Inaccuracies

Independence of gates

  • Aldrich et al., 1983
  • Shown that Na acitvation/inactivation not independent
  • Used whole cell recording and divided currents by number of channels
  • Recorded latency of Na channel opening and closing
    T(activation) was such that in AP when inactivation begins, some channels are still activating.

Sodium channel inactivation

  • Crystal structure suggests 4 Na+ activation gates; within tetramer, sub-units are normally homologous; no immediately identifiable difference in one leading to generation of an inactivation gate
  • Fast inactivation is however as described by HH
  • b-inactivation shown by Raman and Bean 2001
  • Occurs after brief re-polarisation following large depolarisation
  • Involves transiently passing through the open state
    • Consistent with voltage-dependent block of channels by particles which can only enter and block open channels.


  • Voltage step not physiological; recreation?
  • Response to more complex inputs likely to require a more complex model

Method to Madness

NT Release

  • (2011B) Describe the pre- and postsynaptic mechanisms that enable a central synapse to signal with sub-millisecond precision. Some synaptic terminals are large and have many release sites while others are small and have few release sites. What are the functional implications of this heterogeneity?

  • (2013B) Explain how vesicular release probability and the size of the readily-releasable pool of vesicles sets the quantal content and short-term plasticity at synaptic connection. Describe the presynaptic mechanisms that determine whether synapse exhibits paired pulse facilitation or paired-pulse depression.

  • (2014B) Vesicular release from central synapses occurs with timescale of hundreds of microseconds. Discuss the properties of the presynaptic active zone that enable this exquisite temporal precision. Why is fast and precise synaptic signalling considered to be particularly important at synapses made by inhibitory interneurons.

  • (2015) Describe the key evidence leading to the initial proposal of the ‘SNARE hypothesis’ by James Rothman and coworkers. What essential evidence led to modifications of the original hypothesis to inform our current understanding of membrane fusion?

  • (2016) Outline the key questions posed but left unanswered by the work of Bernard Katz and his colleagues on the neuromuscular junction in the 1950s. Give details of how these questions have been answered with the elucidation of the molecular mechanisms underlying neurotransmitter release. (A Nobel Prize was awarded in 2013 for this effort!)

Fatz/Katz/del Castillo

Katz showed

  • Using frog NMJ and recording endplate potentials using electrode in muscle
  • The occurrence of mini EPPs occurring even when motor neuron not stimulated
  • High similarity to normal EPP
  • Concluded that minis due to spontaneous release of ACh
  • Next examined relationship between minis and EPP using Poisson distribution
  • Low calcium Ringer solution led to occurence of many small EPPs (indistinguishable from Mini), that varied between trials
  • Found that fluctuation in EPP was predicted by Poisson
  • Maxima in distribution of EPP amplitude = multiples of mean amplitude of minis.
  • Also explained existence of synaptic failures
  • Showed pool of quanta, with release synchronised to presynaptic stimulation

Katz’s questions
1) How does the synaptic vesicle fuse with the plasma membrane
2) How is this controlled by Ca2+
3) How is the rapid coupling of an AP and Ca2+ to NT release achieved?

Membrane Fusion

Theory of fusion; Stalk hypothesis

  • Treats bilayers as homogenous elastic sheets
  • Proximal monolayers merge and form a stalk, leading to hemifusion intermediate; go from double bilayer to central region with single bilayer;
  • Fusion pore can then be opened by combining remaining bilayer.
    • Apparently fusion pore flickers before irreversibly opening? Just a statement as to equilibria or is this significant?
  • However this is energetically demanding and likely requires proteins to bend the membrane
  • This does occur
  • But additionally, changes in hydrophobic-hydrophilic interactions of bilayers by amphiphilic proteins like NSF can support fusion


  • Microseconds to open; 2 nm in diameter
  • 10-20 ms of flicker period before irreversibly open
  • Membranes then must be recycled

Knowledge of fusion proteins found through viral fusion protein led to Jackknife model

  • Coiled trimer of coils forms
  • Orientation shifts and produces a near lateral movement, bringing membranes together while proteins bend away from future pore site
  • More like a reversed jackknife; spring closing mechanism, held in place by clamp (complexin; later). Removal of clamp through a calcium-dependent model leads to reorientation of all SNARE TM domains being located in the same membrane.

Fusion Apparatus; SNAREs

Developing the SNARE Hypothesis

Energy requirement for unfavourable membrane fusion immediately indicated the involvement of an ATPase
Rothmandeveloped cell free fusion assays of vesicles, and used poisons which blocked fusions
Used them to identify critical proteins
Identified NSF (NEM-sensitive fraction) which has ATPase activity, as well as SNAPs
Previously found by Schekman using radiation of yeast that mutation of the gene Sec18 blocked release. Turns out NSF and Sec18 are the same
Used bead purification column with brain homogenate and NSF with unhydrolysable ATP-gamma- to freeze the reaction
Bound proteins constituted key parts of release machinery
Identified SNAREs (R-SNARE Synaptobrevin/VAMP and the (usually) Q(t)-SNAREs syntaxin and SNAP-25)

Interestingly Botox (clostridial toxin) is a zinc protease which targets and cleaves SNAREs, disabling NT release.

SNARE hypothesis supported by using N-terminal labelling and Electron microscopy to show that N terminals come together. DUe to the fact taht SNARE C-terminals are always in the membrane though SNAP-25 has no TM domain; it is attached to the membrane by palmitoyl side-chains which are covalently bound to cysteines.



  • Interact (zipper) to form 4 helix from 3 proteins (SNAP-25 contributes two helices).
  • 16 “layers”, zero ionic middle layer has conserved 4 hydrophilic AAs
  • Q-SNAREs (often t-SNAREs) such as SNAP-25 and syntaxin contribute glutamine while R-Snare (synaptobrevin) contribute arginine. Q-a/b/c exist denoting the Q-SNAREs position in the bundle

  • Stiffness of these proteins to generate mechanical force is important, as substituting floppiness decreases efficiency of exocytosis.

  • All SNAREs include the “SNARE” motif of 60-70 AA with 8 heptad repeats.

  • When present in isolation SNAREs form helical bundles (Jahn et al 2003)

  • Form a trans-complex called a SNAREpin when on different membranes; this is the prefusion complex. Form by “zippering up from N to C terminals, avoiding a high activation energy. Additionally remaining in the trans-SNARE state is energetically unfavourable; this is however clamped by complexin. Therefore release or inactivation of complexin allows the energetically favoured transition to cis-SNARE formation to occur

  • This transition exerts mechanical force on the membrane by “pulling” on the transmembrane anchors.
  • Produces membrane fusion
  • NSF and co-factor alpha-SNAP then catalyse the separation of the cis-SNARE complex, “funded” by ATP hydrolysis

Alternative Fusion models

  • Potentially the amphiphilic regions of SNAREs can disturb hydrophilic/phobic boundary of memrane leading to formation of hemi-fusion pore state (stalk theory)
  • Supported by finding that SNARE TM domains can cause fusion of lysosomes

Accessory Proteins

SM proteins (Munc18)

  • Involved in formation of SNARE complex;
  • SM protein deletion prevents fusion
  • Three helix bundle at Munc-18 N terminus binds to at Habc of Qa-SNARE syntaxin and part of the SNARE domain (normally inner part of SNAREpin); produces “lock” in closed conformation
  • Though would expect to delay docking, Munc18 KO have impaired docking in chromafin cells but no effect on synaptic vesicle docking?
  • Remains associated with SNARE complex and has role in fusion;
  • Able to “lock syntaxin” but also bind and stabilise full SNAREpin


Cong Ma et al., 2011

  • Munc13 mediates transition by supporting transition of syntaxin to open state, which can then form ternary SNARE with SNAP-25 ro then bind synaptobrevin/VAMP
  • Zippering then occurs and munc18 stabilises the trans-SNARE complex
  • Also some evidence that Munc13 binds Qa/b/c-SNARE trimer to encourage R-SNARE zippering and SNAREpin formation

Rabs and Rims

RIMBP and RIM bind to Rab to bring vesicle to membrane
Complex also binds Munc13 which binds to syntaxin-Munc18 to open the Qa-SNARE
SNARE-pin assembles
Is stabilised by Munc18
Reliance on various proteins maintains focus in active zone, at calcium channels


  • Rabs are GTPases which can form effector complexes when bound to GTP
  • RIM on target membrane is the effector
  • Binds to Munc13 which primes the switch of syntaxin-1 from closed to open (role in Munc18)

Sudhof 2013

  • Binds to this RIM, which brings membrane together and supports formation of the SNAREpin
  • Leads to GTP hydrolysis, and Rab-GDP unbinds and dissociates
  • GTP hydrolysis can be controlled by GEP (Guanosine nucleotide Exchange Factor), a GTP-ase activating factor
  • GEP can therefore control the size of the readily releasable pool and may be involved in short term plasticity.


  • RIMs (Rab interacting molecules) KO shown to effect calckum channel clustering however no effect on vesicle to channel distribution
  • Along with RIM-BP and Munc13 form a tight complex at the active zone
  • Bind to intracellular Ca channel peptide “tag” to bind vesicles nearby into micro/nano domains

Proteins producing Ca-dependence and synchrony

Synaptotagmin - calcium sensor

  • KO Geppert
  • Buffers Augustine
  • Affinity mutation - Fernandez-Chacon
  • Complexin - clamp OR
  • Rabs and Rims - Positioning


  • Calcium triggering shown by Augustine 2001
  • Looked at Ca2+ gradients, as radial/micro/nano domains (increasing Ca2+ concentration)
  • Used BAPTA (rapidly binding chelator) to block NT release, through EGTA (slow acting) did not
  • Indicates nanodomain level of calcium concentration required
  • Response is to “PEAK” of Ca rather than baseline changes in []
  • Indicates a low affinity and fast sensor
  • Though this does vary, such as at the calyx of held, where both do
  • Suggests variation in calcium sensitivity
  • Supports observed difference in nano/micro domains such as changes in development.

  • Synaptotagmin has two Ca binding domains (C2A - 3 calcium and C2B - 2 calcium)

  • Syt1 KO leads to near complete loss of potential driven (synchronous) release (Geppert 1994)
  • Evidence for role in fast exocytosis - Fernandez-Chacon 2001
    • R233Q mutant of Syt1; two fold decrease in Ca2+ affinity without changing structure
    • Patch-clamp from cultured hippocampal cells
    • EPSC smaller than Wt but showed significant STP
    • Suggests reduced Pr is changed variable, as n was unaffected
  • Nanodomains important as Syts are low affinity sensors; react to dynamic calcium concentration rather than background presence of calcium ions.

Ca-dependent Synaptotagmin Isoforms - 2/7/9

  • Syt7 is thought to be involved in asynchronous release
  • Syt1 KO left asynchronous release intact but Syt1/Syt7 double KO lead to removal of ALL exocytosis in chromaffin cell, as well as in hippocampal neurons.
  • Also shown by W Regehrs lab in 2016 that Syt7 is essential for facilitation
  • High affinity allows response to changes in Ca-res
  • 2 and 9 are the other calcium sensitive for synchronous release
  • Syt2 is the fastest and drives release at the calyx of held
  • Syt9 is the slowest and is involved in striatal release
  • Syt9 KO decreased EPSC amplitude


  • Due to discrepancies in timescale regarding Syt mediated response to Ca, alternative sensors have been proposed
  • VGCCs may act as sensors and are coupled to Syt1 and syntaxin-1
  • Conformational change of the channel could propagate allosterically to the primed SNARE complex and trigger release (Atlas, 2013).

Additional role

  • Identified as a possibly important player in vesicular positioning at perimeter of Ca clusters
  • Using a different region to calcium sensing domain

The Complexin Clamp

  • Cytoplasmic protein
  • Compete with Syt for binding of SNARE complex
  • Complexin alpha helix inserts into complex to prevent completion of zippering
  • However Syt/Ca2+ has higher affinity than complexin leading to its displacement, and completion of zippering and transition from trans to cis complex.

Questions and Answers

  • Katz’s questions addressed
  • But new ones now exist
  • What changes in this machinery produce STP/D and vice versa?
    • Could be phosphorylation of Synapsin which normally bind vesicles to cytoskeleton;
    • Phosphorylation by PKA weakens this binding allowing more vesicles to migrate to active zones
    • Additionally Munc18 is modulated by PKC also

Section B

Quantal Theory of Release

  • (2011) Describe the pre- and postsynaptic mechanisms that enable a central synapse to signal with sub-millisecond precision. Some synaptic terminals are large and have many release sites while others are small and have few release sites. What are the functional implications of this heterogeneity?
  • (2014) Vesicular release from central synapses occurs with timescale of hundreds of microseconds. Discuss the properties of the presynaptic active zone that enable this exquisite temporal precision. Why is fast and precise synaptic signalling considered to be particularly important at synapses made by inhibitory interneurons.
  • (2015) Explain how the experimental design and the use of Poisson statistics enabled Katz and his colleagues to show that transmitter release at the neuromuscular junction is quantal in nature. Why are different strategies required to estimate quantal parameters at most central synapses?


Synaptic transmission relies on three main elements

  1. Presynaptic calcium
    • Can be studied using calcium chelators and indicators
  2. Vesicle pool
    • Capacitance measurements (addition of vesicle to membrane) and imaging vesicle proteins with pH-sensitive dyes
  3. Postsynaptic receptors
    • Pharmacological manipulation using antagonists and allosteric modulators can show changes in postsynaptic receptor saturation


Quantifying the elements


R - postsynaptic response
N - maximum number of readily releasable vesicles/release sites
Q - average quantal response (normally univesicular)
P - mean probability of release

  • Product of pr (probability of vesicle fusion) and po (probability of vesicle being primed)
    • Po assumed to be unity at rest,

The product of N and po describes Nt (total number of readily releasable vesicles at t)


  • Release sites are independent
  • Quantal responses summate linearly
  • Synaptic conductance is directly proportional to amount of transmitter release

Best model so far (though has been improved by modifications since). However the principle failing is in that rarely is this able to include post-synaptic influences.

Additionally, quantal parameters are complicated by issues such as receptor saturation

  • MVR (multi vesicular release) usually produces saturation
  • Therefore different to univesicular release model
  • If univesciular release is saturating, then
    • P = probability that release has occurred
    • N = number of active zones
  • Multivesicular saturation indicates that responses relate non-linearly

The Role of Calcium Domains

  • Clustering of VGCC along with endogenous buffers (calretinin) maintain release in nano/micro domains
  • Varies through development e.g. young CoH = micro, mature = nano
  • High local [Ca] = fast binding to low-affinity sensors (e.g. Syt1/2)
  • Couple calcium transient to SNARE and synchronous release
  • Calcium dynamics and processes determine pr (though can also influence po)
  • Effect is non-linear
  • P variance across central synapses is very common, due to variation in
    • Distance of release sites (micro/nano)
    • VGCC density
    • variation in calcium buffer saturation point
  • RIM and Munc13 important in recruiting calcium to active zone (or inverse)
  • Presence of differnt isoforms could have a role in size of domains (and speed of replenishment of RRP)

There can also exist heterogeneity within a single synapse

  • e.g. CoH, has both fast and slow pool
  • Therefore R=q∙(N1∙p1∙N2∙p2)
  • Q could also vary depending on vesicle recruitment methods;
  • Different vesicles preferentially primed by variant machinery?

Cluster Perimeter Release Model

Nakamura et al 2015

VgCC clustering at active zones shown using immungold labelling of freeze fracture preparations

Cluster area (not density) increases with age

Accounts for

  • Changes in the release time course during development
  • Developmental changes in the sensitivity to [EGTA]
  • The mild effect of [EGTA] on release time course

Using simulation with structure of cluster indicated by above information, found that EGTA inhibition could only be reproduced in calcium sensory was at the edge of the cluster

Change in EGTA sensitivity due to increase in cluster along with reduced coupling distance leading to greater spike in [Ca] along with location closer to higher gradient.

So transition from microdomain to nanodomains due to increased clustering of VgCCs and tighter linking of vesicles to the perimeter of these clusters

Post-Synaptic Elements

Multi-Vesicular Release

  • Shown to occur in climbing fibres versus univesicular at neocortical synapses
  • Considered as a mechanism to avoid failure of synaptic transmission
  • Quantal size is determined postsynaptically, by size of response to quantal
  • Used antagonists
  • Low affinity fast antagonist competes, if high glutamate then will compete
  • Can be used to test for amount of glutamate being released
  • Change in release probability by changing calcium should not affect observed response to glutamate as uniquantal means stable amount of glu released
  • In MVR synapses, effect from γ-DGG is evident though not in univesicularly releasing synapses.

Additionally, glial ensheathment can isolate synapses, with high numbers of AA transporters to terminate glutamate signals rapidly

Also some AMPA-R isoforms recover rapidly from desensitisation

  • What effect would that have on STP?

Quantal Release and STP


  • Delta[Ca] can affect pr, producing STPlas.
  • Calcium and pr have a non-linear relationship

Residual calcium hypothesis

  • Build up following first action potential
  • Supported by reduction in facilitation by slow kinetic calcium chelator EGTA-AM; not affecting peaks
  • However does not consider kinetics of calcium dependent processes such as slow-kinetic high affinity calcium binding proteins in various machinery could allow facilitation to persist regardless of residual calcium
  • Therefore not necessarily via an increased baseline but through continued exposure to more calcium ions? More frequent peaks means these slow factors are more likely to bind ions and become activated
  • Recruitment of a low pr pool; would produce facilitation
  • Could be by these slow, high affinity sensors

Use-dependent replenishment

  • EGTA-AM can slow recovery in some synapses (dittman and regehr 1998).
  • Calcium sensor could be Munc13, sensitive to Calcium-Calmodulin; mutant mice with Munc13 insensitive to Ca-CaM had slowed replenishment at CoH (Lipstein et al 2013)
  • Study in hippocampus indicates that Ca-CaM and Syt7 regulate replenishment as a complex


  • (2012) Briefly describe the main presynaptic and postsynaptic mechanisms underlying short-term synaptic plasticity. Explain how they interact to set the frequency-dependence of synaptic signalling.
  • (2013) Explain how vesicular release probability and the size of the readily-releasable pool of vesicles sets the quantal content and short-term plasticity at synaptic connection. Describe the presynaptic mechanisms that determine whether synapse exhibits paired pulse facilitation or paired-pulse depression.
  • (2016) Some central synapses exhibit facilitation while others exhibit depression. Explain how such different behaviours arise and discuss the consequent functional implications for neuronal computation.


  • Defined as changes in synaptic strength of efficacy arising from the immediate history of the presynaptic axon, which persist for less than 30 minutes in duration
  • All synapses undergo both facilitation and depression; therefore it is a question of the balance
  • Modulating neuronal responsiveness reconfigures functional connectivity

For example, STD allows neurons to respond to relative changes

Depression allows detection of changes to low frequency inputs.

Therefore Synaptic STP allows performance of powerful computations through (at least) 2 main mechanisms

  • Gain modulation
  • Enabling discrimination between relative changes in activity

Computational importance of STPlas

Within a functional context

Gain Modulation

Gain modulation is the change in slope in I/O relationship
Important in terms of functional connectivity of neurons
Controls routing of information along a network
Model pyramidal neuron without STD shows that inhibition produces lateral movement of I/O firing rate relationship
Addition of STD enables gain modulation leading to more complex changes and inhibition producing a reduction in maximum firing rate
Neuronal activity is more information rich.

*Based primarily on theoretical considerations, facilitation is thought to influence both information transfer and network dynamics profoundly.

  • In the hippocampus, facilitating synapses function as high-pass filters.
  • May account for the burst firing in place cells that encode spatial information.
  • In the auditory pathway, facilitation may counteract short-term depression to maintain linear transmission of rate-coded sound intensity.
  • Facilitation could forms the basis of short-term memory.
    • Facilitating recurrent connections within cortical networks could support the persistent activity states associated with working memory

Measuring STPlas

Different methods

Paired-Pulse Ratio

Normally deliver second pulse at different intervals to assess plasticity time course

<1 the has undergone PPD, >1 then PPF

Post-Tetanic Ratio

More stimuli are able to induce additional, longer lasting forms of plasticity
Similarly, varying intervals of test pulse allows description of plasticity time course.

Steady State Ratio

Forms of Potentiation

  • Facilitation is the most rapid aspect of STPotentiation.
  • Can be induced with a single pulse
  • Best characterized using PPR
    • repeating PPR at various intervals then plotting PPR against time shows decay of facilitation
    • Common to observe two distinct time courses in facilitation (F1 fast and F2 slow) which still occur in ms range
    • Indicate two distinct facilitatory processes are occuring
  • In addition to facilitation, augmentation also exists (seconds - minutes time scale)
  • Focus on facilitation


  • Usually recovers with multiple time constants
  • Most often attributed to depletion of vesicular pool
  • Can be induced with a single pulse (PPR)
  • Often paired with facilitation
  • Presents as a facilitation depression sequence

Mechanisms of STPlas

Always multiple in a synapse
High release probability leading to depletion of the RRP will almost always lead to depression,

  • Unless prevented by high replenishment rate, such as in ribbon synapses in photoreceptor/ribbon synapses


  • Change in presynaptic AP duration
  • Change in calcium entry
    • Calcium induced calcium channel inhibition
  • Change AP induced calcium at release sensor
  • Change in amplitude and time course of residual calcium
  • Change in sensitivity of release machinery
  • Change in Readily Releasable Pool
  • Rate of clearance of vesicles from active zone


  • Throughout train in hippocampal mossy fibres cells
    • AP got broader
    • EPSC amplitude increases proportionally to half duration of AP
    • Mechanism; inactivation of VGK channels, reducing repolarisation of later APs, increasing duration and therefore Ca entry

Frequency Stimulation and Presynaptic Calcium channel

  • Early work using voltage patch increased acitvation of presynaptic calcium channels at high frequency
  • Indicated NTS1 cAMP Kinase??
  • Later research showed that low frequency (~10Hz) inactivates Ca channels

Katz et al first to propose that build-up of Ca in active zone boosts release

  • Intracellular calcium following AP has a slow decay
  • PPR and residual calcium change similar
  • Application of EGTA-AM (slow buffer) reduced residual calcium, and PPR in parallel with EGTA-AM concentration
  • However addition of relatively very low amounts of calcium unlikely to have strong effect on low affinity fast calcium sensor for release.
  • Either other high affinity calcium sensors or through effects on other parts of release machinery

A molecular mechanism

  • In Schaffer collateral, facilitation is abolished by Syt7 KO, similarly to in corticothalamic, mossy fibre and perforant path neurons
  • Syt7 is a high affinity Ca2+ sensor
    Viral expression of Syt7 restored facilitation
  • However not if mutated to remove calcium binding ability of C2A domain
  • Thought to have a role in Ca-dependent vesicle replenishment
  • Unclear mechanism facilitates p during train by interacting with Syt1
  • Binds residual calcium during train to produce effect after excitation

Calcium and Cadmium

Olesckevich and Walksey 2000

  • Used Cadmium (low calcium) or high calcium
  • in facilitatory synapse
  • Normal had some depression
  • High calcium produced large depression
  • Due to changes in release probability
  • So both mean release probability and -
    replenishment rate important for STP
  • Also high mean release probability means that cannot increase much more to account for changes in other quantal parameters
  • Good example of masking; whatever other molecular mechanisms, RRP depletion will be observed as depression.

Fernandez-Chacon et al 2001

  • Inverse produced using Syt1 point mutagenesis in C2A domain to reduce Ca affinity
  • Reduces Pr
  • Transforms depression into facilitation


Presynaptic PPD largely though to be due to a reducion in RRP

Residual calcium can acccelate recovery from PPDepression

  • Shown in preps where faciliitation is a small component
    • Mediated by calcium dependent vesicle replenishment
    • Might involve CaCam binding to Munc13

Addressing assumptions

  • Literature assumes that transition of vesicles to RRP (docking and priming) is very slow and therefore rate-limiting step
  • Potentially due to work in immature synapses
  • Work by Hallerman and Silver in cerebellar synapse (high frequency firing etc)
  • Demonstrates rapid replenishment of the pool
  • Limited by diffusion of vesicles
  • Shows that not the case in at least some synapses
  • Follow up study showed that transition from releasable to RRP is limited by priming machinery


  • Receptor saturation
  • Receptor Desensitization
  • Polyamine unblock of AMPA receptors
  • Rapid diffusion of AMPARs out of cleft
    • Contribution likely to be negligible (check Silver N&V)

Will all modulate Q

Receptor Saturation

Masks Pre-synaptic Multi-vesicular release and alters PPR at climbing fibre synapse

  • Use of low competitive antagonists to reduce AMPAR affinity for glutamate
  • DGG incubation leads to a depression in peak EPSC produced by 0.5 mM Ca
  • DGG competes for
  • Application of high [Ca] leads to mass release of glutamate, displacing gamma-DGG
  • Shows a larger EPSC versus control
  • In control receptors are already bound and activited
  • Receptor saturation allows more rapid recovery, as less replenishment needed to return to full capacity

Receptor Desensitization

  • AMPARs can desensitise following activation by a neurotransmitter
  • Recovery can take up to hundreds of ms
  • leads to inefficacy of subsequent quantal release
  • Cyclothiazide (CTZ) reduces AMPAR desensitisation
    • However fairly unspecific
    • Can increase AMPAR affinity
    • But can also change waveform of transmitter release through and effect on potassium channels
  • Reduced Paired pulse depression with CTZ incubation shows that this is contributes to STD
  • This is more likely in synapses with high P e.g. multiple release sites
  • Leads to pooling of transmitter and accumulation of Glu in the cleft
  • More likely to produce desensitization

Interplay between P and Q


Polyamine block

  • Spermine = endogenous molecule that can block AMPA receptors
  • Block attributable to GluR1/3/4 (Q sub-units)
  • Depolarisation leads to PA displacement by Ca2+
  • Rebind very slowly
  • Thus repeat release of vesicle will encounter more unblocked channels
  • Leads to STP
  • Through masking of depression

Desensitized AMPA receptors can diffuse away from the cleft?
However data indicates that contribution would be very small

LT Plasticity and STDP??


  • (2011) In 1949, Donald Hebb published a postulate for how information storage should happen in the brain. Describe-the Hebbian postulate and the first experimental findings supporting it. What changes are necessary and what mechanisms are involved? Over the past decade or so, neuroscientists have realized that timing is critical for information storage in many brain regions. Describe this plasticity mechanism and why timing is useful for learning in the brain. Are the mechanistic underpinnings the same as in the Hebb postulate?
  • (2013) What is meant by the ‘Hebbian’ postulate? Briefly describe any relevant experimental evidence supporting the idea. How does the phenomenon of spike-timing dependent plasticity (STDP) affect the postulate? Provide in your answer the role that NMDA receptors play in the phenomenon.
  • (2014) Explain what is meant by long term potentiation (LTP) discussing its basic properties of cooperativity, input-specificity and associativity. How can spike timing-dependent plasticity (STDP) help to explain how these properties of LTP arise? Include in your answer discussion of the function of backpropagating action potentials and NMDA receptors in STDP.
  • (2015) Explain what is meant by LTP and discuss which synaptic activity patterns trigger LTP, focusing on the basic properties of LTP: cooperativity, input-specificity and associativity. How can spike timing-dependent plasticity (STDP), back propagating action potentials and NMDARs help to explain how these three properties of LTP arise?

Donald Hebb and the Hebbian Postulate

When an axon of cell A is near enough to excite a cell B and repeatedly or persistently
takes part in firing it, some growth process or metabolic change takes place in one or both
cells such that A’s efficiency, as one of the cells firing B, is increased.
Cells that fire together wire together

STDP factors in element of “Reliable contribution”


Physiological evidence to support the Hebbian postulate has accrued over the years

Bliss and Lomo 1973 - LTP

  • Stimulated perforant path fibres
  • Extracellular recording of granule cells
  • 100 Hz stimulation trans increased long-lasting increase in population EPSP
  • No further strengthening could be achieved

Dunwiddie and Lynch 1978 - LTD

  • Stimulated SC and commissural fibres
  • Recorded from CA1 pyramidal cells
  • Reduced stimulation frequency to 1 Hz
  • Depression of population-spike response which lasted ~ 60 minutes

Kirkwood et al 1993 - Input Pattern and cooperativity

  • Stimulation in rat SC, recording in CA1
  • Also stimulation in visual cortex, Layer IV and III, recording in layer III
  • Theta burst (100Hz) potentiated both
  • 1-3 Hz depressed both
  • In VC, showed that only stim at layer IV which projects to layer produced LTP/LTD; not due to global effect, input specific
  • Recording at same synapse, shows capable of both


Bliss and Collingridge 1993 - Support for the Hebbian Postulate

  • Two sets of afferent fibres are stimulated using different electrodes
  • Recording from common cell receiving connections
  • Stim intensities adjusted so S1 electrode acitvates fewer fibres than S2
  • S1 tetanus produces PTP and STP, no LTP
  • S2 tetanus produces PTP and LTP (not if S1 is stimulated afterwards; shows not transference to S1?)
  • Tetani to both results in LTP of EPSP to both S1 and S2

  • Shows Input Specificity of LTP; No LTP at S1 when S1 wasn’t stimulated

  • Also demonstrates Associativity; dual stim produced more LTP than individual

Three features of LTP

  1. Cooperativity
    • Induction by coincidenct activation of synapses conveying subthreshold inputs
  2. Associativity
    • Can potentiate weak input which it coincides with a strong input
  3. Input Specificity
    • Can only be induced at activating synapses
    • Due to compartmentalized (microdomain) increase in Ca2+ through unblock of NMDA-Rs

Importance of Coordination - STDP

Markram et al 1997 - Synaptic Efficacy controlled by coordination of EPSPs and APs

  • Recorded from 2 reciprocally connected layer 5 pyramidal neurons
  • If burst firing in post-synaptic neuron is triggered (current injection) during EPSP, then LTP of EPSP amplitude.
  • Does not require sustained postsynaptic depolarization

  • Found that if post-synaptic AP is triggered 10 ms before EPSP then LTD
  • If triggered 10 ms after then LTP
  • Lead to discovery of STDP
  • STDP - a temporally asymmetrical form of Hebbian learning.
  • Likely due to back-propagating APs, as synaptic contacts are too far from soma for passive spread of depolarisation
  • Also calcium produced by bAP likely to contribute to molecular mechanisms of LTP
  • Was dependent on train frequency; not observed below 10 Hz
  • Time window varies between neurons due to characteristics of back-propagation in post-synaptic neuron
  • However nearly all share a sharp temporal transition

  • Now also known that temporal requirement varies between brain area and cell types
  • Pre-leading-post can lead to LTD in anti-hebbian STDP, seen in distal dendrites of L2/3 and L5 cortical pyramidal neurons.

Functional Implications of STDP

  • Adds temporal sensitivity to Hebb’s postulate
  • Important for sequence dependent processing
  • Prevents run away excitation of reciprocally wired neurons
    • Cell A fires before B
    • A to B is strengthened
    • Because B lags behind A
    • B to A would be post-leading-pre
    • ALSO B to A is weakened

From Hebb to STDP

With hindsight, it is clear that a temporal element to the Hebbian Postulate is required
If two neurons fire simultaneously then the presynaptic neuron can not have contributed to the firing of the postsynaptic
Therefore a strengthening of the connection would have no computational value
This bidirectional element of STDP combines perfectly with the heart of the Hebbian Postulate
Of strengthening Cause and Effect; enhancing signal and reducing the noise.

Alternative simplification would maybe be
“Those who trigger each others firing, strengthen their wiring”

Perhaps Slightly more wordy, yet includes the causative elements of “firing together” that STDP demonstrates

Cellular and Molecular

All well and good stating that it occurs; but what are the mechanisms?

Evidence for the role of bAP

The bAP is relatively recently acknowledged phenomenon
Generation of action potential in the soma generates an “echo” which depolarizes parts of the dendritic tree
Activates VGNa and Ca Channels
Depolarizes the dendritic tree

Stuart and Sakmann 1994

  • Recorded from dendrite and soma of neocortical pyramidal cell
  • Found dendritic spikes following somatic current and AP
  • Blockable with TTX, but slow decay rather than immediate block
  • Implies that dendritic Na channels are only boosting the spike which originates elsewhere

bAP and the EPSP

  • Example; glutamate remains bound to NMDA receptors for ~300 ms
  • If Mg+ block is ongoing, then coincidence with a bAP may relieve the block
  • By depolarising the post-synaptic terminal
  • Allows cation flow
  • Sequence of AP after EPSP is important
  • Bliss and Collingridge 1993 also showed that relief of voltage dependent block of NMDAR is important for LTP

Implications for LTP are complex
Whole cell mechanism of potentiating inputs by reducing EPSC required to reach threshold potential
Additionally occurrence of calcium spikes is likely to activate post-synaptic mechanisms, potentially affecting receptor density etc.

Coincidence Detection in tLTP

Rodriguez-Moreno and Paulsen 2008 - Mechanisms of coincidence detection

  • Timing-dependent LTP and LTD requires NMDARs
    • Extracellular stimulation of layer 2/3 pyramidal cells in barrel cortex
    • Used AP-5 (daP?)
    • Both forms blocked by D-AP5
  • t-LTP depends on post-synaptic NMDARs
    • MK-801 (intracellular blocker) applied through patch-pipette into post-synaptic cell
    • Prevented t-LTP but not t-LTD
  • t-LTD depends on pre-synaptic
    • Recording from pre and post in whole-cell patch
    • Adding MK-801 to pre-synaptic prevented t-LTD but not t-LTP

Presynaptic Effects?

Mechanism of Presynaptic LTP: REVIEW (R-M et al 2010)

  • Source of glutamate activating pre-synaptic NMDARs “not understood”
  • Could be glial cells; autocrine signally; retrograde? Autoreceptor
  • If autoreceptor then why don’t LTP protocols also produce LTD???
    • Due to slow-calcium dependent process producing coincidence dependence
    • Retrograde signal could be cannabinoids produced via PLC in response to low levels of Ca2+
  • Varies; both tLTP and LTD are dependent on postsynaptic NMDARs
    • Astori et al 2010;
    • found that adding MK-801 to recording post-synaptic pipette blocked induction of both.

Signal Transduction

Post-synaptic calcium seems key; conventional STP model involves Ca gradient model
LFS leads to low rise in [Ca]
HFS produces large increase
Produce LTP and LTD respectively
Mg block removal by bAP likely contributes to a similar mechanism in STDP


CaMKII (Calcium/Calmodulin-dependent Kinase II)

  • Autophosphorylates
  • Also activation of PKA (either through CaM dependent adenylyl cyclase or through GPC events) can boost CaMKII activity
  • translocates to the synapse, where it binds to the NMDA-type glutamate receptors
  • produces potentiation by phosphorylating principal and auxiliary subunits of AMPA-type glutamate receptors

  • ErK/MAPK has also been implicated

  • Src kinase implicated in enhancement of NMDA-R function during LTP
  • Also PKC isozyme PKM(Zeta) has been shown to be rapidly expressed upon LTP induction
  • Has been implicated in maintenance of LTP in vivo
    L.J. Volk et al 2013
  • However KO of PKM(zeta) had no effect.
    Pharmacological inhibitor ZIP still had effect suggesting that previous data was due to impact on another pathway

Recylcing of AMPARs

  • Main mechanism of LTP is increased postsynaptic AMPA-R density
  • Reserve pool mobilized through process involving Rab11a
  • Exocytozed then diffuse and integrated into membrane by TARPs
  • Evidence suggest that phosphorylation of TARPs by PKC or CaMKII is a possible mechanism


  • Lower calcium levels differentially activate protein phosphatase cascade involving calcineurin. Leads to dephosphorylation of PKA substrates
  • Reduced PKA activity could have an effect on AMPA-R as PKA is known to phosphorylate residues on the GluR1 subunit which may be involved in its trafficking
  • Also mechanisms involving activation of postsynaptic mGluRs (PLC coupled) leads to generation and release of 2-AG, released to activate pre-synaptic CB1R (endocannabinoids potentially released as retrograde signal from post-synaptic neuron)
  • Mechanism does appear to involve presynaptic NMDARs; coactivation with CB1Rs drive the LTD
  • 2AG may also activate CB1Rs on astrocytes


  • STDP biased towards LTD powerfully depresses uncorrelated inputs
  • Likely important in development, balancing convergent inputs
  • Also helps to prevent runaway excitation as mentioned above, due to temporal asymetry

Computational implication from a Model

  • Provides stability through rational weakening
  • Reduces latencies
  • Sharpens response duration
  • Extracts temporal correlations

Section C


  • (2011) For 50 years, epileptologists have debated whether focal epilepsy is due to abnormal neurons (so-called “epileptic” neurons) or a derangement of neural circuits. Explain why these ideas are not mutually exclusive and how both may be correct.
  • (/2012/ - chanelopathies) You find a mutation in an ion channel in a large family with epilepsy. Segregation is not perfect, but 3 of 5 affected individuals have the mutation and only 1 of 16 unaffected has it. You have not sequenced the gene in any other individuals, and nobody has reported this gene plays a role in epilepsy. You write a grant to follow up your work. What would be the three main aims of the grant and what is the hypothesis you wish to test? Please justify your approach.
  • (2014) Seizures can be induced in healthy brain, but epilepsy is the propensity to have spontaneous seizures and, it is claimed, results from changes in small scale networks. Describe the cellular and (small scale) network changes that can lead to temporal lobe epilepsy.

RL focus on tonic GABA-A

Aetiology of Epilepsy

  • Heterogenous group of neurological disorders
  • Characterised by propensity to seizures (synchronous and excessive discharges in the cortex)
  • Partial seizures affect initially only one part of the brain
    • Often produce abnormal sensations, may not be recognised as a seizure
  • Generalized impair consciousness
    • Distort electrical activity of whole/larger (both hemispheres)
  • Partial = 60% of all adult cases
  • Temporal Lobe Epilepsy is the most common form of partial
    • Medial (hippocampus, parahippocampal gyrus, amygdala)
    • Or Lateral (neocortex at surface onf temporal lobe).
  • Partial seizures
    • Simple
      • Mnestic sensations (déjà vu, jamais vu, amnesia, or recall of a memory/set)
      • Gustatory; olfactory; auditory; visual; sensory (surface or visceral)
      • Can also produce dysphoric, euphoric and generally emotional senatons
      • Called “auras” by some that think they precede a full seizure;
    • Complex
      • Impair consciousness (alter ability to interact normally with environment)
      • Begin as simple and then spread, normally to both hemispheres
      • Can cause motionless staring, unusual speech and behaviour, automatic movements
  • Frequency of seizures increase following initial seizure
    • As partial seizures are sometimes unnoticed, this makes not only the original insult but also the original seizure hard to detect


  • Genetic predisposition
  • Environmental factors
    • Brain Injury
    • Inflammation
    • Febrile convulsions
    • Hypoxia/ischaemia
    • Origin of insult often unknown due to delay between occurrence and emergence.
  • Neuronal loss
    • TLE often shows loss of CA1 and CA3 neurons
    • Damage to mossy cells and inhibitory interneurons of hippocampus, plus granule cells of the dentate gyrus
    • Animal models show neuronal loss during seizures
    • In humans loss predates first seizure (how is that known? Partial seizures often not recognised).
    • Loss of inhibitory interneurons may increase hyperexcitability of hippocampal neurons leading to seizure propensity
  • Granule cell dispersion
    • Structure of layer changes; no longer closely packed and dendrites rearrange
    • New connections may lead to hyperexcitability?
  • Mossy fibre sprouting
    • Loss of granule cell connections leads to mossy fibre sprouting
    • TLE shows larger mossy fibres than in normal brain
    • Sprouting continues from 1-2 weeks after injury
    • May create excitatory feedback circuits
    • However may also synapse with inhibitory basket cells (GABA and NPY)
    • Granule cell excitability in animal models also detectable prior to sprouting

Network Changes underlying propensity to seizure

  • Epilepsy has often been attributed to a shift of excitation/inhibition balance
  • However seizures are infrequent
  • Rather than shift in balance there is a shift which is compensated for
  • This compensation results in reduced stability
    Homeostatic mechanisms are the likely culprits (though also non-homeostatic)


  • Granule cell dispersion of dendrites and mossy fibre sprouting may be due to inflammatory response
    • Astrocytes normally secrete signalling factors that contribute to injury repair and damage control
    • However in epilepsy also contribute to formation of feedback loops
    • Factors increasing glutamate sensitivity are also realised
    • Blood brain barrier is also disrupted by inflammatory response leading to entry of proteins such as albumin which trigger further hyperexcitation and inflammation
  • The kernel theory of gene change
    • Pre-ictal activity in seizure triggering kernel leads to loss of phasic IPSPs
    • Enhanced by reduced Cl extrusion due to changes in KCC2

Network Level

Modify Gain leading to increased responsiveness?
Steeper slop in I/O relationship

  • Loss of interneurons leading a decrease in phasic inhibition increases gain without affecting actviation threshold. This increase in neuron excitability is likely to be epileptogenic
  • Increase in tonic inhibition raises the activation threshold, suppressing network excitability
  • However gain is unaffected
  • Matches with epileptic model of shift in E/I compensated to maintain function yet leading to an unstable network.

Loss of Interneurons Changes Inhibitory Pattern

  • In vitro models in rodent slices have shown breakdown in feedforward inhibition which normally restrains the excitatory drive
  • Human brain slices have shown loss of GABA-ergic function prior to seizure
  • Usually seen as a loss of interneurons, producing a reduction of the number of GABAergic synapses onto principal neurons (Pavlov 2011)
  • Many changes occur in the hippocampus following injury, but general consensus is that net outcome is the loss of phasic inhibition
  • More specifically, in pyramidal neurons it was seen that spontaneous GABAergic inhibition increased at the soma but decreased at dendrites
  • Resulted from hyperactivity of somatic projections of interneurons (increased frequency of spontaneous glutamatergic currents in these interneurons)
  • Alongside increased excitatory input from increased and aberrant axonal sprouting of CA1 pyramidal cells
  • Reduced dendritic inhibition likely due to degeneration of an interneuron subpopulation
    • O-LM cells? synapse primarily onto distal dendrites
    • Parvalbumin positive hippocampal interneurons which shape rhythmicity, and control synaptic plasticity in the hippocampus
  • Could reduce seizure threshold by potentiating excitatory inputs
  • Increased somatic inhibition would support the infrequency of seizures in partial epilepsy.

Kobayashi and Buckmaster 2003

  • Pronlonged EPSP in Rat granule cells; loss of feed forward
  • Showed reduced spontaneous mIPSC frequency and monosynaptic IPSC conductance in dentate granule cells
  • Reduction in number of parvalbumin and somatostatin positive interneurons;
  • And reduction in both fast and slow rising GABAa currents
  • Indicates reduction in dendritic and somatic

Increased Tonic GABAA

Role of Tonic inhibition in epilepsy indicated by mutations in genes encoding extrasynaptic GABA-ARs associated with epilepsy
Loss of phasic dendritic inhibition as explained above

Pavlov and Walker Review/Scimemi 2006

  • Correlation between tonic and phasic inhibition in hippocampal neurons
  • Synaptic release of GABA important source
  • Loss of interneurons/reduction in phasic GABAergic activity should decrease tonic inhibition
  • Experimental models (induced status epilepticus) show during and beyond epileptogenic period, tonic GABAAR currents are maintained in hippocampal neurons (some studies show increase)
  • Large tonic GABAAR currents also reported in neocortical and dentate neurons in humans with TLE


  • In hippocampus of controls, α5 and δ subunits main contributors to tonic conductance’s in vitro
  • Also confirmed pharmacologically
  • In animal models of status epilepticus, δ is downregulated and apparently compensated by α4
  • γ2 sub-unit decreased synaptically and increased Extrasynaptically , leads to reduction in affinity (GABA EC50 5x greater)
  • Also desensitize more rapidly??
  • Synaptic decrease associated with decrease in phasic
  • Further conflicting regarding lower affintiy and
  • Increase in extrasynaptic GABA concentration?
    • Tonic currents appear to be largely due to glial GABA released through the calcium activated channel Bestrophin1;
  • Change in Inhibitory Amino Acid Transporter in neurons or glia?
  • However these s.u. changes are not seen in other insults such as brain trauma

Changes in tonic inhibition along with decrease in phasic inhibition leads to a network with reduced excitability as well as a narrower dynamic range, combined with decreased stability (tonic inhibition does not affect neuronal gain, mainly subthreshold effects).

However increased tonic GABA-A activation could be pro-convulsive due to excessive load on Cl- extrusion combined with decreased KCC2 expression
Pushes reversal potential to more depolarised values


  • Low Ca and high Mg incubation blocks synaptic responses
  • Results in increased spontaneous firing of principal neurons
  • Rhythmic and synchronous burdting developed
  • Restricted to CA1 population
  • Depression of after hyperpolarization and increased excitability
  • Synchronization attributed to electrical interactions between pyramidal cells
    Fast oscillations
  • EEG oscillation preceding the onset of focal seizures appear to be dependent on gap junctions
  • Leads to very fast population oscillations

Cell Intrinsic

Bursters *

Sanabria 2001

  • Paired intra/extracellular recording in rat CA1
  • Orthodromic activation in control neurons evokes a single spikes
  • Epileptic slices produces all or none high-frequency burst of 3-6 spikes
  • All sampled cells showed same response (control and sham control had ~ no difference)
  • ~10% of cels also burst fired spontaneously affter suppression of synaptic activity (TTX/CNQX)
  • Blocked by Ca channel blocker or removal of calcium from ACSF but not by intracellular BAPTA
  • Therefore must be changes in post-synaptic Ca current dynamics
  • Act as pacemakers for initiating epileptiform bursts
  • Additionally bursts lead to generation of LTP in CA1 synapses?
  • Spontaneous bursters then likely to have strong synaptic connections with other neurons including conditional and spontaneous bursters
  • Combination with reduced gabaergic activity?

  • Upreg of certain calcium channels shown in hippcampus of human TLE (T-Type shown in animal models?)

Enhanced spontaneous activity

Bernard 2004

  • CA1 pyramidal neuron excitability increased in TLE
  • Due to decreased A-type potassium channel presence
  • Reduced transcription as well as increased channel phosphorylation by ERK mechanisms
  • ERK inhibition partly reversed increased dendritic excitability
  • Reduced A-type mediated K channel current will contribute to epileptogenesis by
    • increasing amplitude of bAP
    • Boosts calcium influx in response to X
  • Increasing amplitude will boost STDP

Shah 2004

  • 24 hrs after seizure (latent period, enhanced spontaneous activity in entorhinal cortex (in vivo and in vitro)
  • Accompanied by reduction in Ih, and decline in HCN1 and HCN2 subunits
  • Though membrane potential was hyperpolarized, dendritic excitability was enhanced
  • resulting in increased neuronal firing

  • Reduction in Ih and HCN proteins also shown in epileptic neurons during the latent period, following experimental induction of status epilepticus

  • Loss of Ih shunting effect, which provides hyperpolarising drive, and decreases EPSP amplitude
  • Slowed decay of EPSP previously shown in epileptic neurons
  • Leads to increased spontaneous acitvity, increased dendritic EPSP summation enhancing EPSP-spike coupling (increased STDP through slowed decay of EPSP)

Reversing excitability with Kv1.1 and Seizures with halorhodopsin

Wykes 2012 - The epileptic neuron

  • Induced experimental epilepsy using tetanus toxin
    • Cleaves SNARE synaptobrevin; likely leads to tarnscriptional effects on various channels
  • Detected enhanced intrinsic excitability in focal pyramidal neurons.

    • Had less hyperpolarised membrane voltage
    • Reduced current threshold
  • Optogenetic inhibition using with halorhodopsin, lentivirally delivery to epileptic focus, attenuated EEG detected seizures

  • Lentiviral expression of Kv1.1 reduced intrinsic excitability of transduced pyuramidal neurons; prevented the development of epilepsy if delivered prior to tetanus toxin
  • Administration to epileptic focus progressively suppressed epileptic actiity over several week

Cell Synaptic

Increased Glu release from astrocytes

Tian 2005

  • Rat hippocampal slices
  • Showed paroxysmal depolarisation can be triggered by extrasynaptic glutamate
  • As well as uncaging Ca in astrocytes
  • In vivo two photon microscopy showed that anti epileptics (valproate etc) reduced calcium signal transduction in astrocytes, which probably reduces glutamate release

AMPA/Kainate Changes

Epsztein 2005

  • Aberrant synapses from sprouting mossy fibres onto original granule cells.
  • In epileptic animals (not controls) Kainate receptors are involved
  • Produce 50% of non-NMDA excitatory drive
  • Slower kinetics than AMPA; role in LTP?
  • Large effect on computational properties of affected DG granule cells.
  • Role of kainate in recurrent transmission suggests usefullness of kainate blocker in preventing synchronized activities.

Enhanced NMDAR

Lieberman and Mody 1999 - Enhanced NMDAR

  • Calcineurin is a Ca-CaM dependent protein phosphatase
  • Inhibitory action on NMDA receptors; reduces opening duration and desensitizes receptor
  • However in hippocampal slices from human TLE, calcineurin inhibitors failed to prolong openings
  • Indicates decreased calcineurin activity
  • Factor involved in immune response and activation of T-cells

  • Increased PKC and ERK mechanisms observed in epilepsy? Tie into bursting phenomenom triggering LTP processes?

Scimemi 2006 - Enhanced release and NMDA

  • Reduced PPR also seen in perforant path evoked EPSC
  • Change persistent after blocking various tonic presynaptic modulators (opiate, adenosine, GABA)
  • Indicates increased release probability
  • Also seen enhanced NMDA mediated EPSC
  • Not explained by s.u. changes or decreased glutamate transporter activity
  • Indicates change in NMDAR kinetics
  • Synaptic cross-talk also seen (lateral stimulation produces use-dependent block of receptors in medial synapses, even in MK801
  • Increased Pr, enhanced NMDAR current and increased cross talk could all promote seizure generation through enhancing excitability

Depolarising GABA-A; Return to Immaturity

  • Change in driving force due to changes in NKCC1 (hippocampal tissue from TLE patients more sensitive to bumetanide, NKCC1 blocker). NKCC1 mediates Na driven Cl uptake
  • Subpopulation of pyramidal neurons with low KCC2 seen in human TLE hippocampal tissue
  • Changes in ionic gradient make GABA depolarising
  • Implications in combination with increased tonic inhibition?
  • This is immature form seen during development


Mechanosensation, Pain and Sodium Channels

  • (2011) By describing the molecular basis of mechanosensation in the periphery, discuss how tissue damage is detected by the central nervous system.
  • (2012) What molecular mechanisms contribute to activation of peripheral pain pathways?
  • (2013) Discuss our current understanding of mechano-reception in the peripheral somatosensory system
  • (2015) Discuss the role of voltage-gated sodium channels in modality-specific pain pathways

Pain Transmission and Peripheral Nerves

Pain transmission; A delta fibres

  • Myelinated
  • 2-30 m/s
  • First (sharp) phase of pain
  • Unmyelinated
  • 2 m/s
  • Thought to be polymodal based on electrophysiological data though gene deletion shows modality specific effects
  • Emery et al 2016 showed modality specificity (pressure, heat and cold specific activation of DRG cells)
    Prolonged activation can lead to wind-up

Blocking Pain Transmission blocks Pain

Haroutounian et al 2014

  • Efficacy of peripheral nerve conduction block in neuropathic pain shown

Varo et al 2014

  • Additionally in Phantom Limb Pain
  • Block of DRG using intrathecal or intraforamen application of sodium channel blockers (lidocaine)
  • Removed phantom limb pain as well as non-painful PLS
  • PLS previously thought to be due to cortical malfunctions, lack of plasticity leading to persistent interoceptive dissonance
  • In the exception of cases where damage is to thalamus or insular cortex (such as in stroke), appears that neuropathic pain can be prevented by peripheral nerve block

Modality Specificity of Primary Afferents

  • Emery et al 2016
  • Showed using GCaMP6 (which has single AP sensitivity) the activation of specific neurons
  • Activation in response to mechanical (pinch), or noxious cold, or noxious heat
  • modality specific response seen in vivo
  • Brush neurons which do not respond to noxious stimuli e.g. heat but do to brush
  • Discriminating nature of Mechanosensation; over stimulus does not lead to activation
  • Polymodality produced in vivo
  • PGE2 unmasks silent nociceptors
    • One of the mechanisms behind allodynia?
  • Co-expression of transducers in nociceptors? Would this not lead to polymodality?
  • Needs to be resolved
  • May modulate each other? so TRPA1 in TRPV1 neurons has a role in response to inflammation but no longer to cold?
  • Likely to be modality preference over modality specificity
  • However if not casts doubt on a number of KOs as compensatory function may lead to dominant activity of transduction element responding to a different modality

Karolinska institute bank account is hench

Supported by Usoskin finding distinct classes of sensory neurons
Unbiased classification to produce mocelular map
11 unique types of sensory neurons
Vary in terms of receptors, channels, trafficking proteins and peptides
all indicate different roles

Conduction of Sensation

Sodium Channel Morphology

Structure of Na Channels


  • S4 voltage sensory selectivity filter between
  • Isoleucine, phenylalanine and methionine and threonine, inactivation?
  • Beta subunits and AnkG for localization
  • Protein kinase A and C affect phosphates on intracellular loop serine residues between alpha units 1 and 2

Alpha Subtypes

Alpha subtypes

  • Nav1.7/8/9 are main peripheral Na channels
    • Nav1.1 Loss leads to epilepsy; mainly present in inhibitory neurons so counterintuitive effect
  • 1.2?
  • 1.3 link to autism? Fetal brain role and development?
  • 1.4 in skeletal muscle,
  • 1.5 in cardiac tissue
    • Brugarda syndrome? Sudden death?
    • Right next to 1.8 in genome???
    • TTX resistant??
    • 1.8 implicated in Brougarda
  • 1.6 key in nodes of ranvier KO = lethal
  • 1.7 Gale Mandel, found in PNS AND CNS
  • 1.8 Sensory Specific
  • 1.9 also but creates assistive current
    • Also found in motorneurons
  • Nax - key for Sodium homeostasis, conduit for sodium absorption if extracellular sodium exceeds range.
  • Variety key for different biophysical properties
    • Some very rapidly adaptive etc and modified by different intracellular messengers e.g. PKA, GTP
    • 1.9 KO reduces activity, without blocking it by removing this assistive current

Sodium Channel Alpha Subtype Distribution

  • Expressed in peripheral nociceptive neurons and visceral afferents
  • Threshold channel
    • Has extremely slow inactivation kinetics and window current;
    • Active at -80 mV in humans
    • Has role in DRG neuron excitability
    • Nav1.9 current lowers AP threshold and drives repetitive and spontaneous firing
    • Shown using Nav1.9 activator ATP-gamma-s
  • Huang et al 2014 (human)
    • Gain of function missesnse mutations can cause painful peripheral neuropathy
    • Depolarized membrane potential of DRG neurons
    • Enhanced spontaneous and evoked firing of DRG neurons
  • Kurth et al 2013 (human)
    • Loss of pain produced by Nav1.9 gain of function
    • Excessive actviity at RP leads to sustained depolarization of nociceptors
    • Effect on mEPSC
    • Imparied action potential generation
    • So produces loss of function in sensory conduction mechanism
  • Also shown to have a role in muscle growth in motor neurons
  • Nav1.9 LoF can often lead to muscular weakness
  • Expressed exclusively in primary sensory neurons
  • Key in generating AP; Nav1.8 KO DRG had no change in membrane resistance, RP, or thresholds but showed significantly impaired electrogenesis of AP.
  • Depolarized voltage dependency but slowly inactivating and rapid recovery from desensitization, key for contribution to generating inward AP current

  • No human mutations found or related to pain until 2012 (Waxman group)

  • GoF leading to enhanced channel response contributing to hyperexcitability
  • Likely to contribute to neuropathy

Abrahamsen et al 2008 - Mouse

  • Cre-lox system and DTX receptor used to KO all Nav1.8 expressing neurons
  • Retained low threshold mechanical sensation
  • Showed loss of
    • Noxious cold
    • Noxious mechanical
    • Inflammatory thermal and mechanical pain
    • Spontaneous pain triggered by inflammation
    • Neuropathic pain unaltered
    • Noxious heat unaltered

Role in Sensitization

Laird and Cervero 2002

  • Nav1.8 KO mice
  • Normal response to acute noxious impact of intracolonic saline
  • However weak pain, no hyperalgesia to intracolonic capsaicin
  • Blunted pain to intracolonic mustard oil
  • No difference in mutants and wt in tissue damage or inflammatory response

Also shown separately that; increased trfficking of Nav1.8 caused by PGE2 contributes to inflammatory sensitization

  • Unexpectedly, Nav1.8 levels are reduced after nerve injury
  • However Nav1.8 block of function pharmacologically or genetically (antisense oligonucleotide applciation) alleviates/reverses pain related behaviour in rats
  • Highly expressed in nociceptive neurons (DRG and trigeminal ganglion), sympathetic ganglion neurons, myenteric ganglion neurons
  • Vasylyev et al 2014 (SG Waxman’s group)
    • Showed that Nav1.7 has important role in generation and conduction of APs; rapidly activation and inactivation, generating current which amplifies AP generator potentials
    • Supports Nav1.8 by amplifying those APs to higher nav1.8 threshold, and activate window current
  • Boosts subthreshold stimuli, increasing P of AP generation

  • Also slow reprining; not well suited to neurons with high firing frequency; C-fibres are low so goody two shoes

  • However can also produce resurgent currents (triggered by repolarisation after strong depolarisation), leading to bursting in DRG
    • Nav1.7 KO leads to massively reduced current density in DRG neurons
  • Nav1.7 in the hypothalamus, modulation allows change of bodyweight in mice.
  • Demonstrated a persistent current, allowing integrating current which allows synaptic integration to break threshold
    Nav1.7 KO leads to loss of summation.



Fertleman 2008 - PEPD (familial rectal pain)

  • Mutation in inactivating portion of Na channel resulting in extreme pain from mechanical pressure
  • Mutation in inactivating region of Nav1.7
  • No inflammatory pain because there is not a dysregulation of activation
  • But continuous activity
  • Channel does not go through the origin

Primary Erythromelalgia

First channelopathy linked to chronic neuropathic pain (2004) Similarly to PEPD, flushing and episodes of intense pain
Burning pain specifically (TRPV1 expressing neurons?

KO showing function

Global Nav1.7 deletion leads to early death in mice

  • Role in olfactory processing (Weiss et al 2011)
  • Mice couldn’t smell milk so died!
  • However also appears to have developmental roles whose loss leads to observed fatal KO
  • However Nav1.7 disruption leads to loss of inflammatory pain

Nassar et al 2004

  • Made use of sensory neuron specific Nav1.7 KO (Advillin-cre, floxed Nav1.7)
  • Von Frey hair mechanical testing showed reduction in dorsal horn responses
  • Reduced noxious thermal sensitivity (Hargreaves test)
  • Induced inflammation using CFA had no effect

Minett et al, 2012

  • Compared advillin-Cre/Nav1.7 KO to Wnt1 linked Nav1.7 K
  • Advillin present in all sensory neurons?
  • Wnt1 expressed in embryonic precursors of sensory and sympathetic neurons
  • Nav1.8 expressed in DRG, but also in LowThreshMechR neurons
  • Advillin KO = all sensory neurons
    • AdvCre versus Nav1.8-Cre
    • Thermal sensitivity persists in Nav1.8KO (Hot plate and acetone test)
    • Shows that Nav1.8-negative neurons are important in thermo-nociception
  • Advillin KO also abolished wind-up and substance P release
    • Increase in spiking not observed in KO mice following electrical stimulation of periperhal inputs to spinal cord
    • Yes seen in Wt
  • Wnt1 KO = sensory and sympathetic neurons

    • Sensory KO; still sensitive to noxious heat
    • Required KO in sympathetic also to produce full pain imperviousness
  • However more importantly, required pan DRG AND symp KO to prevent neuropathic pain produced by L5 spinal nerve transection (SNT)

  • Directly implicates sympathetic nervous system in neuropathic pain;
  • Old literature acknowledged but fell out of fashion when not a rewarding analgesic target

  • 2014 Paper showed that Wnt1 knockout prevented sprouting on ipsilateral spinal cord which was observed in all other knockout cases

  • Has been previously shown that sympathetic neurons develop increased sympathetic sensitivity through upregulation of alpha2
  • Additionally, lack of mechanical allodynia following SNTin Wnt1 Nav1.7 KO was negated by plantar injection of Noradrenaline

Failure in Cancer Pain

Minett and Falk, 2013

  • Mouse osteosarcoma behaviour unaffected
  • Oxaliplatin induced pain also Nav1.7 independent
  • However data indicates significant difference in mechanisms (though there is overlap)
    • Inflammatory models show increased substance P and CGRP in spinal cord
    • Absent in cancer models
    • Additionally scoring mechanisms differ significantly
    • Von Frey hairs etc not used as largely inconsistent in cancer pain
    • Rather limb use scoring and weight bearing ratios
    • Suggests profound difference in pain sensation
  • Nav1.7 blocker Protoxin-II has minimal efficacy as analgesic
    • Case with all selective blockers; in fact affinity and efficacy inversely proportional
  • Though adult gene ablation effective in adult floxed mice (Minett et al 2014)
  • Used Tamoxifen induced Cre-recombinase expression
    • Fusion to Estrogen R
    • Tamoxifen binding displaces heat shock proteins, allowing translocation from cytoplasm to nucleus and activating Cre expression (and removal of floxed SCN9A)
  • Reversed reduction in withdrawal threshold produced by neuropathic pain
  • Appears to function as both a channel and transcriptional regulator
    • Few hundred proteins found which bind to Nav1.7
    • Synaptotagmin1 and 2
  • GPRIN? Modulate opioid receptors?
    • Opioid actions blocked by Na??
    • Sodium in some way inactivates the Rs?
    • Nav1.7 KO shown to have strengthened response to fentanyl
  • 10+ gene transcripts dysregulated in KO mice
  • Sodium affecting PENK1 expression; increased in KO

    • How is gene transcription effected?
    • Is sodium a second messenger in sensory neurons?
    • NFAT5???
      • Gene transcription activated by osmotic stress, potentially role for Na? Through salt-dependent kinase.
    • Enhanced in Nav1.7 nuclei
      • 5 consensus binding sites for Penk??
      • Na as a secondary messenger?
      • SiRNA for NFAT5 removes enhanced Penk expression
  • PENK (enkephalin precursor)

  • Metenkephalin immunoreactivity increased in DRG + dorsal horn in Nav1.7 KO mice
  • Tested using Monensin (sodium ionophore which tracks Na through the membrane with channel activity)
  • Produced significant decrease in PENK mRNA levels in Wt females but not males
  • Increased enkephalin production in women thought to underlie increased pain tolerance
  • However produced significant decrease in both genders of Nav1.7 KO
  • However no effect on PENK expression with Nav1.8 KO
  • Shown with mutations in other sodium channels? What makes Nav1.7 Na so special?
  • Analgesia due to increased opioid expression
  • KO analgesia reversed by Naloxone in both mice and humans

SCN9A LoF leads to peripheral enkephalin production along with attenuation of electrical excitability
Together lead to loss of pain


  • Sodium channel blocker and opiate synergy shown before using lidocaine and morphine (Kolesnikov et al., 2000)
  • Nav1.7 blocker and opiate synergy
  • Combine Protoxin-II with thiorphan shows significant increase in withdrawal latency? Time? Maybe Randall-Selitto test
  • Thiorphan = enkephalinase inhibitor

Nociceptors and the Action Potential

Lucas 1909 showed all or nothing signalling using current input and muscle movement
Cole and Curtis 1939 showed Action Potentials linked to increases in conductance
Hodgkin and Huxley used voltage clamp to detect changes in current, attributed to Na and K ions mainly linking potential change to equilibrium potentials for major ions Na, Ca, K and Cl. Developed gate theory of channel activation and inactivation which determine the kinetics of the action potential.

Voltage Clamp, allows study of ionic changes in media
Excitable tissue

At rest, K+ permeable, Na impermeable though large driving force
Raising extracellular K+ depolarises membrane potential. Increasing excitability of the neuron

Axon Hillock; site of AP initiation (

  1. Lower threshold for AP initiation due to high density of Nav (20-200 x more than Soma/dendrites)
  2. small diameter , reducing current required to drive membrane threshold
  3. Different Nav channel properties

Propagates, as distal Na channels activate, proximal Na channels inactivate and late K channels activate
This active current generation is essential for the all or nothing current generation behind the action potential

Pharma tools in ion channels

  • TTX and Saxotoxin; Na channel blockers
  • Tetra ethylammonium (TEA) blocks K current revealing Na current
  • Axon diameter determines conduction speed (effect on Tm, capacitance etc and resistance.

History of Na channel comprehension

  • cloning of Na channels
  • Identified ion selectivity models, and Horn 1995 carried out mutagenesis studies to identify the S4 region as the voltage sensor
  • Variety of alpha sub-units (the pore-forming channel proteins rather than the associated Beta sub-units)

Treating Pain

  • Suppressing nociceptor activity is obvious answer to treating pain
  • Attempts to block transduction mechanisms failed due to high levels of redundancy
  • However blocking conduction appears to work in all cases
    • Efficacy in treating phantom limb pain for example (Haroutounian/Varo)
  • New focus is on using specific sodium channel blockers to prevent transmission
    of nociceptive information

Molecular Aspect of Sensing


  • Variety in mechanosensory channels means that different currents are produced depending on the stimulus and the differential activation of mechanically sensitive channels
  • Drew et al 2007 were first to distinguish MA currents pharmacologically
  • Used NMB-1, which did not affect voltage gated or ligand gated channels
  • Inhibited MA currents in a subset of sensory neurons
  • These MA currents where unaffected by tarantula toxin which inhibits stretch activated ion channels
  • NMB-1 also blocks FM1-43 loading in cochlear hair cells
  • NMB-1 inhibits high intensity mechancial stimulation with no effect on low threshold or thermo sensation via TRPV1
  • In randall-selitto test, increases rat pain threshold by 40% with no effect on response to heat stimuli
  • Shows that distinct channels respond to different types/degrees of stimuli.


Can be measured experimentally through different methods of inducing stretch of neuron membrane

  • Behavioural
    • Randall-Selitto (ramping force)
    • Von Frey
    • Tactile Acuity
  • Single Fibre
    • Skin-nerve prep
  • Cell based
    • Osmotic stretch (hypo-osmotic solution causes swelling)
    • Focal pressure (with a glass probe)
    • Substrate stretch (grow neurons on an elastic membrane)
  • FM1-43 - blocking dye
    • Labels cell membranes
    • Permeates mechanosensitive channels (blocking MA currents)

Mechano-Transducers; Piezo channels

  • Cation channels
  • 24-36 TM domains
  • Only Piezo 2 found in sensory neurons
  • No clear pore-forming sub-units so not clear if channels or if they confer mechanosensitivity to other conducting proteins
  • However P1 found to contain ion conducting pore
  • Purified Mouse Piezo found to assemble as homo-oligomer with no other associated proteins
  • Not yet shown for P2 however very likely


(Coste et al 2010)

  • Showed Piezo1 mediated MA current in neuroblastoma cell line
  • Overexpression increased
  • Piezo1 and Piezo2 produce currents with distinct kinetics
  • Piezo 2 responsible for rapidly adapting MA currents in DRC
  • Showed rapid kinetics
  • Then used piezo2 siRNA
  • Stopped firing only in neurons with MA currents inactivating in <10ms
    (Eijkelkamp et al 2013)
  • DRG EPAC1 (cAMP sensor) mRNA increases in neuropathic pain
  • cAMP analogue with EPAC1 specificity 8-pCPT enhanes
  • Interaction with cAMP sensor EPAC1 could produce allodynia
  • Piezo2 MA currents are enhanced by EPAC1
  • Long-lasting allodynia is prevented by EPAC1 kd and prevented in Piezo2 KO mice

Patapoutian Group showed that Piezo2 is crucial in innocuous touch but not in noxious mechanical sensation though inverse in drosophila;

Side-note; TRPV1 activation inhibits Piezo receptors through depletion of PIP and PIP2

Surprise; TRPA1

  • Previously thought to be responsible for thermosensation, also mechanically sensitive (Kwan et al 2006)
  • TRPA1 mutation causes Familial Episodic Pain
    Mutation in S4 segment increases inward current 5-fold (Kremeyer et al, 2010)

More Surprises TRPM8

Molecular Sensitization; Allodynia and Hyperalgesia

Galeotti et al, 2004 - Protein Kinase C

  • Activation by increased IP3 via activation of PLC pathway (Gq GPCRs)
  • H1 receptors involved in hyperalgesia (inflammation and sensitization)
  • FMPH produces hyperalgesia which is blocked by PLC and PKC inhibitors (unaffected by IP3-R antagonists)

Di Castro et al., 2006 - NGF and PKC

  • Modulates TRPV1 reaction to heat stimuli at various steps
    • Transcription
    • Translation
    • Post-translation
  • Increases transcription of mechanosensitive ion channels
  • PMA (PKC activator) enhances MA currents
  • Removal from rat neurons culture removes large-amplitude MA currents produced by PMA
  • Indicates that NGF and PKC contribute to same pathway, activated by PMA
  • Thought that NGF increases transcription of PKC-sensitive mechanically activated channel proteins
  • Incubation of NGF -> 350% increase in MA currents
  • Incubated with NGF and transcription or translation inhibitors blocks this effect
  • PKC activation also increased MA currents, inhibited by tetanus toxin
  • Thought to control insertion/trafficking of the new mechanically sensitive channels generated due to NGF effect
  • Induces exocytosis
  • Labelling of vesicles with FM1-43 showed that PKC increased rate of vesicular trafficking

Anything with a banging essay

  • Olfaction?
  • Interneurons?
  • Hippocampus
    ….. or nought