Ca2+ Signalling

SOCE

Spatial Localisation

Decoding Ca2+ Signals

PLC Regulation/Inositol pathway

Protein Regulation

Ion Channels and Pumps

Membrane Contact Sites

Once decoded Ca2+ stimulates exocytosis, contraction, proliferation, fertilisation, hypertrophy and some metabolic pathways and genes' transcription

Can be just as effective as modifying biochemical properties in controlling flux through a pathway

Sensitivity - often you want a low Kd (greater sensitivity), fast k1 (for rapid onset) and slow k-1 (for sustained action - but means that recovery is slow)

Proteins allow for greater sensitivity, signalling throughput and modulation than other molecules

Spare receptors ensure that EC50 < Kd and therefore that the system is more responsive and can amplify signals more effectively

Proteins can function as switches (cooperative bonding) and logic gates (multiple binding sites), they can also remember through phosphorylation/other modifications (methylation of histones)

Original work was done by Sydney Ringer in 1883. In 1953 the 'PI response' was seen by Lowell and Mabel Hokin, which was identified as being causally linked to receptor-stimulated PLC by Michell in 1975. In 1980 Tsien developed the first Ca2+-stimulated fluorescent dye (fura 2). Then in 1983 work by Jim Putney and Gillian Burgess showed that Ca2+ was predominantly stored in the ER through murine experiments (it had previously been assumed to be in the mitochondria) and then Berridge determined that IP3 triggered it's release through blowfly experiments. This was key as it meant people finally understood how Ca2+ was released. In 1986 Putney released a theoretical mechanism for how this all fit together and in 1992 Hardie demonstrated that Ca2+-permeable TRP channels are essential for Drosophila phototransduction.

Different membranes have different concentrations of phosphoinositides which allow different proteins to bind and identify them (PI4P = Golgi, PI3P = early endosome, PI(3,4)P2 = multi-vesicular endosome) and this has been utilised with fluorescent probes for imaging purposes

Most ion channels (Trp are regulated by DAG) are thought to be regulated by PI(4,5)P2, as is f-actin

Some Trp channels have a bifunctional purpose as Ca2+ channels

PKC

DAG can regulate PKC too, as does Ca2+

Regulated by PI(4,5)P2

Action results in increased [IP3] and [DAG] but a decrease in [PIP2]

Inositol passes through SMIT1 and is incorporated into PtdIns on the ER where PITP shuttles it to the PM where it recombines with DAG

PI 4-kinase families are stimulated by PKC (reverese reaction by sac1) and PIP 5-kinase families are stimulated by phosphatidic acid (reverse reaction by PIP2 5-phosphatases)

Reduced [] is seen at nerve terminals in Downs syndrome model which results in learning deficits - can however be corrected by ablation of the third gene of synaptojanin 1 (Voronov, Frere, Glovedi et al)

PA switches off the reactions and initiates recycling

All InsP is converted into inositol eventually

Bipolar disorder is treated with Li+ and valproate

Li+ is thought to inhibit substrate-bound enzyme to stabilise the [inositol] in brain cells and keep it low as the BBB is impermeable to Ins, although there are concerns about this theory (Balla, 2013)

Reduced inositol results in reduced PIP and PIP2 and hence DAG and IP3, resulting in lower signalling output (inositol depletion hypothesis)

PITPs (PI Transfer Proteins) are the transporters used, but only at membrane contact sites (10-20nm apart) and they are used to ferry PA and PI between the ER and PM

Nir2 is important in humans (Kim, 2015, Dev Cell), the drosophila homolog is Rdg

The NTD binds PI/PA through a hydrophobic binding pocket (no energy input required) and the CTD binds to DAG/PA in the PM to tether it to the PM. The FFAT sequence binds VAP in the ER for a similar function

These are maintained by extended synaptotagmins (E-Syt1-3) which under Ca2+ binding to their C2 domains reach out to the PM and draw the ER closer to it (Chang, 2013) although some have argued that PIP2 is required as well (Giordano, 2013)

Catalytic activity requires tightly bound Ca2+ in the active site to withdraw the negative charge of IP3
Additionally, the XY linker has to be moved as it ordinarily tonically inhibits the enzyme as a pseudosubstrate

PLC subtypes

Beta1-4 are activated by GPCRs and stimulate GTPase activity of Gq's alpha subunit

Gamma1-2 are the only ones with SH domains, which allows them to interact with RTyrKs

Delta1,3,4 are the simplest and stimulated to amplify activation in some cells

Zeta is only expressed in sperm and is essential for fertilisation

Epsilon is the largest and activates rap1 via its Ras GEF domain

Challenges are that it has to sever a kinetically stable phosphodiester bond, and needs to cut very close to the PM without damaging the PM

GTP-Gqa binds the C2 domain and Rac1-GTP & Gbg bind the PH domain to hold it in a specific orientation

PIP2 binds the PH domain (at a non-catalytic site) and Ca2+ at the PM binds the C2 domain

Doing this pushes the pseudosubstrate out of the catalytic site due to the -ve charges at the PM - this is called interfacial regulation

Gbg & rac1-GTP may stimulate activation by interfacial activation as the structure (Lyon, 2013) does not show movement of the pseudosubstrate

Gqa binding stimulates allosteric activation by sequestering Ha2/3 that previously locked in the psuedosubstrate

SH2 - Tyr-P binding results in phosphorylaiton on Y783 and this then binds cSH2 and moves the pseudosubstrate

Ca2+ sensors

EF hands exist in pairs and consist of seven -ve residues and bind Ca2+ in a pentagonal bipyramidal conformation

CaM has two pairs (and is transcribed from three separate genes)
Ca2+ binding causes an unstructured region to form a helix and allow binding to another protein

C2 domains have a beta sheet structure and Ca2+ binds between the top end of the domain and the PM to localise proteins there

PLCd, PKC and synaptotagmin all have these domains to ensure that they function correctly

Annexin domains are functionally similar as C2 domains but each repeat binds two Ca2+ ions and are found in annexin proteins and in the same way are vital to localisation of proteins

C2 and annexin domains drive the localisation of proteins to the PM in response to Ca2+ and ensure that only there are these proteins activated

Motifs allow binding of more than one Ca2+ ion, allowing a graduated response to [Ca2+] and most proteins respond to much higher [Ca2+] than is experienced globally in the cell

Ca2+ buffers (by large proteins or organic molecules) ensure that diffusion occurs slowly and that pools exist only very locally - up to 99% of cytosolic Ca2+ is bound to such molecules

Signals are switched off through transporters as cells do not have the power to force nuclear fission/fusion

Pumps have alternating gates and are much slower than channels which have only one
Pumps are driven by secondary transport whereas channels are limited by ion diffusion

Two main routes of transport Ca2+ out of the cell - PMCA and NCX

PMCA has a lower capacity but higher affinity

NCX dominates in rapid response cells and PMCA in cells that have slower changes in Ca2+ flux as NCX can respond to a wider range of and larger peak [Ca2+]

Ca2+ stimulates both of these channels

Only bacterial NCX has been crystalised (Liao, 2012, Science) and it shows that the steps are driven by diffusion, with electrostatic repulsion driving Na+ out of the cell

Overall this is rapid for a pump though as there is little conformational change

PMCA is stimulated by CaM-Ca2+ which binds at two sites (one high affinity = basal, and one low affinity = graduated response)

This system allows for memory as the 1st site becomes saturated at high [Ca2+] and is then slow to dissociate

Some Ca2+ is stored in the mitochondria but this is strictly limited

Most Ca2+ is stored in the ER/SR and transport into it is driven by SERCA (selectively inhibited by thapsigargin)

SERCA is related to PMCA and pumps 2xCa2+ for each ATP molecule used (PMCA only transports one)

SERCA has been crystalised in multiple conformational states and has EF hand-like structures in the M10 region

ATP hydrolysis is coupled with the protein moving from high to low Ca2+-affinity states (in the ER facing state Ca2+ affinity is low and it is replaced by H+)

SERCA is regulated by phospholamban, which traps it in the E2 state (ER facing)

NCLX exchanges Na+ (in) for Ca2+ (out) and MCU imports Ca2+ directly

The large H+ gradient that is present drives both ATP generation and Ca2+ transport into the mitochondria

Mitochondrial Ca2+ uptake complex utilises this H+ gradient to drive the transfer of Ca2+ ions
Membrane potential controls the gating of this (it is activated once [Ca2+] = 1miM in the cytosol

MCU is the pore-forming subunits of the MCUC and is regulated by MICU1&2(&3 in the brain) which confer Ca2+ regulation, which they sense through EF hands. EMRE scaffolds these two together and is required for activity of MCU units

MICU1 binds to EMRE and MICU2 to MICU1 and inhibits MCU at basal [Ca2+], when [Ca2+] rises MICU2 is removed and MICU1 now interacts to stimulate MCU (both deprepression and activation)

IP3 receptors and ryanodine receptors trigger Ca2+ release from the ER/SR

RyR(1-3) form homo-tetramers to respond to membrane depolarisation

IP3R(1-3) form homo- and hetero-tetramers to respond to IP3 from PLCs

They provide the Ca2+ required for muscle contraction

RyR1 is present in skeletal muscle and is localised to transverse tubules and is stimulated by a conformaitonal coupling to open L-type CaV channels
DHPR are present in 4 per tetrad here

RyR2 is present in cardiac muscle and is opened by a chemical coupling to Ca2+ ions
DHPR appear to be randomly distributed

IP3Rs are opened by sequential binding of IP3 (primes the channel) and then Ca2+

This allows for coincidence detection where Ca2+ release activated surrounding primed channels

Highly expressed in Purkinje cells in the cerebellum

Coincidence detection has been observed in these cells

Each is innervated by parallel fibres and a climbing fibre, which when stimulated result in long term depression
LTD is blocked by KO of IP3R1, mGluR1 (and restored by cell-specific expression), lack of ER in spine apparatus and inhibition of IP3-evoked Ca2+ release also inhibit LTD

IP3Rs are inhibited by large [Ca2+]cyt, but this has a slower response time than the activation of the IP3Rs
Without this there would likely be explosive feedforward activation

Slide 86 is the current best model for how this functions

IP3 binds 7nm away from the pore and the two phosphate groups draw in the alpha and beta units (clam shell) to prime the channel (Seo, 2012, Nautre)

Information is transduced through the loops in the TMDs, LNK domain is particularly important. Ca2+ binding sites are on the ARM3-LNK and ARM2-ARM1 interfaces (Fan, 2015 - well worth looking at before commenting on the regulation pathway)

Ion channels at the PM must be much more selective than those on the SR/ER because there are many more molecules/ions that could pass through

For comparison IP3R/RyR moves 1M Ca2+/s whereas SERCA moves 10/s

XY linker is targeted to PI(4,5)P2 containing vesicles

Nesures large amount of IP3 to trigger large Ca2+ release

Has a different regulation mechanism with a positively charged tail

PLCs were shown to encode digital signals in hepatocytes by Woods and Cuthbertson (Nature, 1986)

Inhibiting calcium channels with Gd3+ does not prevent spikes but slows the time between them being turned on and off

Two models were that there was that [IP3] was either oscillating (in line with Ca2+ oscillations) or staying stable

mGluR5 has an inhibitory phosphorylation site on it, this is not present on mGluR1

Using IRIS fluorescent biomarkers it was determined Matdsura (2000 or 2008) that [IP3] was in fact largely constant after an initial peak (conc increased rapidly and then oscillated at a high level before rapidly decaying)

Similarly injecting a non phosphorylatable/dephosphorylatable IP3 analogue into a Xenopus oocyte triggers [Ca2+] oscillations

Often done whilst trapping the IP3 analogue in a photosensitive cage and imaging with TIRF microscopy

Using these biosensors has allowed imaging of blips and puffs

Some IP3Rs appear to be preferentially activated

At higher [Ca2+] the nature of the signal changes from blips -> puffs -> waves

Imaging by Keebler and Taylor has shown that puffs repeatedly occur at the same sites

Thought to be due to the action of actin-associated KRAP, which is hypothesised to make IP3Rs competent at interacting with IP3 (licensing) - Taylor's lab hypothesises that this leads to a hierarchy of IP3Rs where only some can be primed

These respond to the spatial organisation of Ca2+ waves

Using an aequorin fluorescent probe that is targeted to the mitochondria showed:
1 - That local [Ca2+] reaches 10s miM
2 - Adding a bolus of Ca2+ does not induce the same response as a histamine-induced Ca2+ spike

Membrane contact site between ER and mitochondria allow for this high local [Ca2+], which triggers the MCU complex

The response is to activate dehydrogenases and immobilise the mitochondria to ensure that enough energy is being produced and that ATP generation is close to the stimulus

Only MCU is quick enough to oppose IP3R

Immobilisation occurs as a result of Ca2+ binding Miro, which ordinarily binds to kinesin (Ca2+ binding breaks this interaction to leave the mitochondria unattached to transport machinery)

Slides 114-116 depict various Ca2+ synapses

Responses to toxic Ca2+ concentrations are mediated by NFAT

Calcineurin (phosphatase) is activated by CaM-Ca2+ and dephosphorylates NFAT to activate it

NFAT is ordinarily hyperphosphorylated to inhibit it

CaN is inhibited by Tacrolimus, which is used to prevent organ transplant rejections

NFAT then forms a trimer TF

Forms this with various different TFs; can form a complex with AP-1 or other bZIP proteins to stimulate transcription in immune cells

Translocation to the nucleus is stimulated by oscillatory SOCE-mediated Ca2+ oscillations

The same sort of oscillations can be seen by using La3+ to inhibit the Ca2+ transporters in the membrane but without Ca2+ to move across the membrane NFAT does not translocate to the nucleus

Spikes every 2 minutes is the optimum frequency for translocation

The recovery is slow so that NFAT can accumulate in response to large and sustained Ca2+ influxes (if it was a rapid offset then no memory would be achieved)

Phos/dephos cycle occurs over ~2 minutes
IKB phosphorylates NFAT and therefore inhibits it - degradation/synthesis cycles of this occur over ~30 minutes

Ca2+-CaM-dependent protein kinase II (CaMKII) is widely expressed, particularly in the brain and has many target

Different isoforms selectively target specific Ca2+ channels (IP3Rs, RyRs, CaVs etc)

Selectively activated by local Ca2+ signals and has a major role in remodelling of synapses (e.g. by regulating spine volume and AMPAR activity during hippocampal LTP)

Hub domain allows it to form a dimer-hexamers (12 subunits overall)
Linker is the most variable domain and the length of it affects inhibition

Regulation is controlled by a pseudosubstrate that is removed from the kinase domain by Ca2+-CaM binding - this region is buried deep inside (hub or reg domain - not quite clear from diagram)

As a result of the masked CaM binding site, CaM only binds after CaMKII has exposed its binding site (which is more likely with a longer linker - hence longer linker = higher sensitivity)

The process of binding site exposure is called 'breathing' - longer linkers cause more 'breathing'

Regulation is also controlled by phosphorylation at Thr286

Phosphorylation slows CaM-Ca2+ dissociation and disables the pseudosubstrate site

Kinase activity is an inter-subunit interaction so neighbouring subunits must both be bound to CaM-Ca2+, hence it is very rare at low [Ca2+]

Slow dephosphorylation allows for the memory of signals

The longer the linker is the lower the frequency of oscillations that is required for autonomous (high levels of) activity

Phosphorylated/autonomous subunits have an increased dissociation rate, which allows them to switch into active or inactive hubs (the latter of which then activates a given hub) (Stratton, 2014, Life)

Oscillations are economical and prevent toxicity
Rapid recovery responses track Ca2+ signals whereas slower recovery responses integrate them

By draining the stores and then adding receptor agonist and Ca2+ bolus separately it can be shown that it is emptying of stores that triggers SOCE rather than receptor activation

Many processes in mast cells are dependent on SOCE (transcription, NOS, Ca2+-sensitive ACs)

Orai (PM) and STIM1 (E/SR) are essential components for SOCE
Ca2+ binding at EF hands causes STIM1 to aggregate at PM-ER contact sites, their CAD domain then interacts and activates Orai1

STIM1 KD and o/exp attenuate/increase SOCE activity in many cell types
It also redistributes towards the PM when Ca2+ stores are depleted

Most SOCE is mediated by Orai(1-3) and STIM1 interacting at membrane contact sites (siRNA KD of either attenuates SOCE, and loss-of-function mutations in either cause SCID)

STIM have SAM domains that unfurl upon Ca2+ releasing its binding
CAD/SOAR domains activate Orai1 and can activate it when expressed as a single domain - but no inhibition is seen
CC1 domain occludes CAD until Ca2+ binds
Polybasic CTD binds PIP2 upon unfurling of STIM in response to Ca2+ stopping binding to the EF hands in the ER (due to low [Ca2+])

Aggregation can be seen in fluorescence experiments where the the two proteins localises to puncta after store depletion

SCID is a rare, recessive and life-threatening immunodeficiency disease caused by mutations (Arg91Trp) in Orai1 (identified through siRNA of 74 genes)