Calcium as an Intracellular Messenger
An egg is fertilized. Within seconds, a wave of light sweeps across its surface. This is visible under a fluorescent microscope as a ring of glowing calcium. A muscle fiber twitches within milliseconds of a nerve impulse arriving. A pancreatic beta cell senses rising glucose. A few hundred milliseconds later, it releases insulin. All three events are triggered by the same tiny ion: Ca²⁺. Free calcium is one of the most versatile signals in biology. It couples electrical, mechanical, and chemical stimuli to almost every cellular process. These processes include contraction, secretion, gene transcription, metabolism, fertilization, and cell death. The puzzle calcium signaling must solve is this: cells need calcium for structural purposes. It stabilizes membranes, activates extracellular enzymes, and is a building block of bone. Yet high concentrations of free cytosolic Ca²⁺ are toxic. They cause proteins to aggregate, mitochondria to fail, and phosphate to precipitate. Evolution's solution was to keep resting cytosolic free Ca²⁺ almost 10,000-fold lower. This is compared to the extracellular fluid and the endoplasmic reticulum (ER) lumen. That steep gradient turns Ca²⁺ into a coiled spring. Opening a channel for a few milliseconds releases a large, fast, spatially restricted signal. Then pumps and buffers snap the concentration back down. The information is encoded not just in how much calcium rises. It also depends on how fast, how often, and where. This article builds directly on the The Phosphoinositide Pathway: IP3 and DAG. There, IP₃ generation is the trigger that opens ER calcium channels. Here we go one level deeper. We examine how the resting gradient is built and maintained. We look at how the cell decodes a calcium transient once it happens. We also explore how oscillations and waves let one ion carry many different messages without ambiguity. Calcium signaling is a paradigm case for Signal Amplification and the Cascade Concept. It is also a paradigm case for Signal Integration and Crosstalk. This is because nearly every major second-messenger pathway eventually reads or writes to the calcium channel. The story is also historically important. Calcium signaling was worked out through some of the most elegant single-cell imaging experiments in cell biology. These experiments used calcium-sensitive dyes, aequorin from luminescent jellyfish, and genetically encoded sensors. They give students a clean window into how modern experimental cell biology actually operates.
- Resting cytosolic free Ca²⁺ (~100 nM) is ~10,000-fold lower than extracellular fluid (~1–2 mM) and the ER lumen (~100–800 µM), a gradient maintained by pumps, exchangers, and buffering proteins.
- IP₃ generated by phospholipase C opens IP₃ receptor channels on the ER, releasing stored Ca²⁺ into the cytosol; voltage-gated and store-operated channels admit Ca²⁺ from outside.
- Calmodulin, a four-EF-hand Ca²⁺ sensor, changes shape upon binding Ca²⁺ and activates dozens of downstream enzymes, including CaM kinase II, myosin light-chain kinase, and calcineurin.
- Signals are encoded spatiotemporally: calcium sparks, puffs, waves, and oscillations of varying frequency carry distinct information decoded by frequency-sensitive effectors like CaMKII.
- Roger Tsien's synthetic fluorescent indicators (fura-2, indo-1) and Osamu Shimomura's aequorin let researchers directly visualize calcium transients, transforming signaling from inference to observation.
- Dysregulated calcium handling underlies cardiac arrhythmia, neurodegeneration, and some cancers, and is a direct drug target of calcium-channel blockers and calcineurin inhibitors.
Calcium signals work because the cell keeps cytosolic Ca²⁺ almost 10,000-fold below the extracellular and ER concentrations, so that briefly opening a channel produces a fast, steep, information-rich transient that dedicated sensor proteins like calmodulin can decode in amplitude, time, and space.
The core idea
Calcium signaling rests on an electrochemical gradient rather than on synthesis and degradation of a molecule, which is what distinguishes it from most other second messengers. Cyclic AMP or IP₃ are made and destroyed by enzymes; Ca²⁺ is an element, so cells cannot synthesize or destroy it. Instead, cells control its free concentration by pumping it across membranes into compartments — mainly the extracellular space and the ER lumen (and, transiently, mitochondria) — and by binding much of it to buffer proteins. Resting free cytosolic Ca²⁺ is about 100 nanomolar; extracellular Ca²⁺ is about 1–2 millimolar; ER luminal free Ca²⁺ is roughly 100–800 micromolar. Opening a channel therefore lets Ca²⁺ rush down its electrochemical gradient into the cytosol within milliseconds, producing a fast, large-fold-change signal — exactly the property a good signaling system needs. The second core idea is decoding. A rise in free Ca²⁺ has no intrinsic meaning until it binds a sensor protein that changes conformation and, in turn, changes the activity of an enzyme, channel, or transcription factor. The master universal sensor is calmodulin (CaM), a small, highly conserved 148-residue protein with four helix-loop-helix "EF-hand" Ca²⁺-binding motifs arranged in two lobes. Calmodulin itself has no catalytic activity; it is a pure allosteric switch. When Ca²⁺ occupies its EF hands, calmodulin wraps around a short target peptide in a Ca²⁺/CaM-dependent enzyme, exposing hydrophobic surfaces that dock onto and activate targets like CaM kinase II (CaMKII), myosin light-chain kinase, calcineurin (a phosphatase), and many others. Other sensors, such as troponin C in muscle, synaptotagmin at synapses, and protein kinase C's C2 domain, perform analogous roles in specific tissues. The third core idea, and the one that most distinguishes calcium from a simple on/off switch, is that information is encoded in the pattern of the signal, not just its peak amplitude — a property called spatiotemporal encoding. The same total calcium exposure delivered as one large pulse versus many small oscillatory pulses can activate different genes and different physiological outcomes, because downstream decoders like CaMKII are frequency-sensitive.
Mechanism
1. Building and maintaining the gradient. At rest, the plasma membrane Ca²⁺-ATPase (PMCA) and the Na⁺/Ca²⁺ exchanger (NCX) continuously extrude Ca²⁺ from the cytosol to the extracellular space, while the sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) pumps Ca²⁺ from the cytosol into the ER lumen. Both pumps work against a large concentration gradient: SERCA hydrolyzes one ATP to move roughly two Ca²⁺ ions into the ER, and PMCA moves one Ca²⁺ out per ATP; NCX instead uses the inward Na⁺ gradient (three Na⁺ in for one Ca²⁺ out) rather than ATP directly. Mitochondria can also take up Ca²⁺ transiently through the mitochondrial calcium uniporter, acting as a fast local buffer and a way to couple calcium signaling to ATP production, since Ca²⁺ activates several matrix dehydrogenases. Cytosolic Ca²⁺-binding proteins (calbindin, parvalbumin, calretinin) further buffer free Ca²⁺, shaping how far and how fast a signal spreads from its point of entry.
2. Triggering release — the ER as an internal store. Following activation of a Gq-coupled receptor or a receptor tyrosine kinase, phospholipase C cleaves PIP₂ into IP₃ and DAG, exactly as described for The Phosphoinositide Pathway: IP3 and DAG. IP₃ is water-soluble and diffuses through the cytosol to bind IP₃ receptors (IP₃Rs), which are tetrameric ligand-gated Ca²⁺ channels embedded in the ER membrane. Binding of IP₃ together with a modest rise in Ca²⁺ itself opens the channel — Ca²⁺ is a co-agonist, which is the molecular basis of calcium-induced calcium release (CICR), a positive feedback loop that converts a local opening into a propagating signal. Ryanodine receptors (RyRs), a structurally related channel family, perform an analogous CICR role in muscle, where they are triggered directly by voltage sensors (skeletal muscle) or by Ca²⁺ entering through L-type Ca²⁺ channels (cardiac muscle).
3. Entry from outside. Plasma-membrane channels admit extracellular Ca²⁺ down its steep inward gradient. Voltage-gated Ca²⁺ channels open upon membrane depolarization (crucial at nerve terminals and in cardiac/smooth muscle). Store-operated Ca²⁺ channels (SOCs), the best-characterized being the Orai1 channel, open when the ER store empties: the ER-membrane protein STIM1 senses low luminal Ca²⁺, oligomerizes, and physically relocates to ER–plasma membrane junctions to gate Orai1 — a mechanism called store-operated calcium entry (SOCE) that refills the ER and sustains long signals.
4. Decoding by calmodulin and downstream effectors. Once free Ca²⁺ rises in a microdomain or throughout the cytosol, calmodulin's EF hands bind Ca²⁺ cooperatively; the two lobes close around exposed hydrophobic patches on target proteins. Key downstream effectors include:
| Effector | Ca²⁺/CaM effect | Physiological output |
|---|---|---|
| CaMKII | Autophosphorylates, becomes Ca²⁺-independent ("molecular memory") | Synaptic plasticity, LTP |
| Myosin light-chain kinase | Phosphorylates myosin light chain | Smooth muscle contraction |
| Calcineurin (PP2B) | Dephosphorylates NFAT | T-cell activation, cardiac hypertrophy |
| Adenylyl cyclase (some isoforms) | Stimulated or inhibited by Ca²⁺/CaM | Cross-talk with cAMP pathway |
| Troponin C (muscle-specific) | Shifts tropomyosin off actin | Skeletal/cardiac muscle contraction |
5. Termination. SERCA and PMCA pumps, together with mitochondrial uptake and cytosolic buffers, rapidly restore resting Ca²⁺ within tens to hundreds of milliseconds, resetting the system for the next signal. Failure of this clearance (e.g., SERCA inhibition) causes prolonged, toxic Ca²⁺ elevation.
6. Spatiotemporal patterning. Because IP₃Rs and RyRs are clustered and exhibit CICR, calcium release is inherently non-uniform. A single channel opening produces a "blip" or "spark"; clusters produce local "puffs"; and if enough puffs recruit neighboring clusters through CICR, a regenerative calcium wave sweeps across or even between cells (through gap junctions). Sustained receptor stimulation frequently produces regular calcium oscillations — repeated spikes rather than a sustained plateau — because IP₃Rs desensitize and store refilling is cyclic. The frequency of these oscillations (from ~1 per few seconds to ~1 per minute, depending on cell type and receptor occupancy) itself carries information: CaMKII activation, gene expression via NFAT, and mitochondrial metabolism can each be tuned by oscillation frequency rather than average concentration, letting one pathway drive qualitatively different outputs depending on the temporal code delivered.
The evidence
Ridgway, Ashley, and the discovery of the calcium transient (1963–1967). The first direct proof that a transient rise in free calcium accompanies excitation came from work on giant single muscle fibers of the barnacle Balanus nubilus, whose cells are large enough to inject with a photoprotein. E.B. Ridgway and C.C. Ashley injected aequorin, a Ca²⁺-activated bioluminescent protein purified from the jellyfish Aequorea victoria by Osamu Shimomura and Frank Johnson, directly into the muscle fiber. When the fiber was electrically stimulated to contract, it emitted a brief flash of light exactly coincident with the twitch, proving that free Ca²⁺ rises transiently at the moment of contraction and falls again as the muscle relaxes. This was the first direct, real-time visualization of an intracellular second messenger changing concentration inside a living cell — a landmark that established calcium as a genuine signaling ion rather than merely a structural component.
Fabiato's demonstration of calcium-induced calcium release (1975–1983). Alexandre Fabiato used skinned (sarcolemma-removed) cardiac muscle cells, allowing him to control the concentration of Ca²⁺ bathing the internal organelles directly. He showed that a small, controlled puff of Ca²⁺ applied to the cytosolic face of the sarcoplasmic reticulum triggered a much larger release of Ca²⁺ from the SR itself — proving that Ca²⁺ can trigger its own further release, the CICR mechanism now known to operate through ryanodine receptors and, in non-muscle cells, IP₃ receptors. This explained how a small trigger (voltage-gated Ca²⁺ entry) could be amplified into the large release needed for a full-strength heartbeat.
Roger Tsien's fluorescent indicators and single-cell imaging (1980s). Aequorin required protein injection and gave relatively slow, dim signals. Roger Tsien designed small synthetic fluorescent Ca²⁺ chelators — quin-2, then the ratiometric dyes fura-2 and indo-1 — built around the Ca²⁺-selective chelator BAPTA fused to a fluorophore, whose excitation or emission spectrum shifts on Ca²⁺ binding. Because these dyes could be loaded into any live cell as membrane-permeant acetoxymethyl (AM) esters that are trapped intracellularly after esterase cleavage, they made calcium imaging accessible to any laboratory with a fluorescence microscope, not just specialists able to microinject cells. Using fura-2 with ratiometric imaging (measuring the fluorescence ratio at two excitation wavelengths to correct for uneven dye loading), researchers directly visualized calcium waves sweeping across fertilized eggs, oscillations in hormone-stimulated liver cells, and quantized "puffs" of Ca²⁺ release at individual IP₃R clusters. Tsien shared the 2008 Nobel Prize in Chemistry (with Shimomura and Martin Chalfie) for this and related work on GFP-based sensors, which later gave rise to genetically encoded calcium indicators (GCaMP), now standard tools in neuroscience for imaging neural activity in living, behaving animals.
Berridge's oscillation and wave studies (1980s–1990s). Michael Berridge and colleagues, imaging hormone-stimulated hepatocytes and other cell types with fura-2, showed that many agonists produce not a single sustained Ca²⁺ rise but repetitive spikes — oscillations whose frequency, rather than amplitude, scaled with the concentration of the applied hormone. This was the key evidence for frequency encoding: a cell reads "how much stimulus" from "how often the spike train fires," a robust way to transmit graded information through a highly nonlinear, all-or-none release channel. Together, these experiments moved calcium signaling from inference (indirect pharmacology and ion-flux measurements) to direct visual proof, in living single cells, of concentration changes on the correct timescale and in the correct place to explain physiological outputs — the gold standard of causal cell-biological evidence.
Worked examples
Example 1 — Estimating the gradient and the energetic cost of maintaining it. Suppose a cell has resting cytosolic free Ca²⁺ of 100 nM and extracellular Ca²⁺ of 1.8 mM (a typical mammalian plasma value). Fold gradient = 1.8 × 10⁻³ M ÷ 1.0 × 10⁻⁷ M = 18,000-fold. This is why opening even a handful of plasma-membrane channels can produce a large relative change in cytosolic concentration almost instantly — a 1% leak of the extracellular pool into a small cytosolic volume overwhelms the resting level many-fold, unlike (say) a doubling of ATP concentration, which requires far more molecules to move for the same fold change. Because PMCA must pump Ca²⁺ from ~100 nM cytosol against this gradient into ~1.8 mM extracellular fluid, and SERCA must pump against an even steeper ER gradient (resting cytosol ~100 nM vs. ER lumen up to ~800 µM, an 8,000-fold gradient), a substantial fraction of a resting cell's ATP budget — commonly estimated at several percent in electrically active cells such as cardiac myocytes — goes toward calcium pumping. This illustrates a general principle: steep, information-rich signaling gradients are metabolically expensive to maintain, and the cell "pays" continuously so that the release step can be fast and cheap.
Example 2 — Reading an oscillation frequency to infer hormone concentration (a physiological-reasoning problem). A researcher stimulates isolated hepatocytes with vasopressin at three doses and records cytosolic Ca²⁺ with fura-2 ratiometric imaging:
| Vasopressin dose | Ca²⁺ pattern observed | Oscillation period |
|---|---|---|
| Low (0.1 nM) | Occasional single spikes | ~1 spike per 3–5 min |
| Medium (1 nM) | Regular spiking | ~1 spike per 60 s |
| High (100 nM) | Fused into sustained elevated plateau | continuous |
A student is asked: does the peak height of each individual Ca²⁺ spike change much across these doses, and what does that imply about how the cell encodes the hormone concentration? Reasoning: Because Ca²⁺ release through IP₃Rs is a regenerative, largely all-or-none CICR event once threshold IP₃ and Ca²⁺ co-agonism are reached, individual spike amplitudes are similar across doses — what changes is how often a spike fires. The cell is therefore encoding hormone concentration in oscillation frequency, not spike amplitude — consistent with Berridge's original hepatocyte experiments. The biological logic: a frequency code lets a receptor system communicate a wide dynamic range of hormone concentration using a signal (Ca²⁺ release) that is otherwise digital/threshold-like at the single-channel level, and it allows a frequency-decoding kinase like CaMKII (whose autophosphorylation accumulates faster with more frequent pulses) to translate frequency back into a graded, analog output — for example, a graded rate of glycogen breakdown proportional to hormone dose, even though each underlying Ca²⁺ pulse looks similar.
Common misconceptions & traps
| Misconception | Correction |
|---|---|
| "Calcium signaling means making more calcium." | Ca²⁺ is an element; the cell cannot synthesize it. Signaling works by releasing pre-stored Ca²⁺ from compartments (ER, extracellular space) into the cytosol, then rapidly pumping it back. |
| "More calcium always means a stronger signal." | Information is often encoded in oscillation frequency or spatial pattern (sparks vs. waves), not just peak amplitude; sustained high Ca²⁺ is frequently cytotoxic, not "more signal." |
| "IP₃ receptors and ryanodine receptors are the same channel." | They are structurally related but distinct channel families with different primary triggers (IP₃ + Ca²⁺ vs. voltage/Ca²⁺) and different tissue distributions. |
| "Calmodulin is an enzyme." | Calmodulin has no catalytic activity itself; it is a Ca²⁺ sensor/adaptor that allosterically activates other enzymes. |
| "The ER calcium store is just a passive reservoir." | The ER store is actively refilled by SERCA, senses its own depletion via STIM1, and signals to the plasma membrane to open Orai1 channels (store-operated entry) — an active, regulated system. |
The scientist behind it
Origins · 1 minRoger Y. Tsien and how this was discovered
Roger Tsien trained as a chemist but became one of the most influential figures in cell biology by building tools rather than discovering pathways directly. As a graduate student and young faculty member, he recognized that the existing calcium-sensitive proteins (aequorin) were too cumbersome for routine use — they required microinjection of purified protein into each cell. Tsien instead designed small, membrane-permeant synthetic molecules: BAPTA-based Ca²⁺ chelators fused to fluorophores, engineered so that Ca²⁺ binding shifted their fluorescence spectrum in a way that could be measured ratiometrically, canceling out artifacts from uneven dye loading or cell thickness. His indicators — quin-2, fura-2, indo-1 — could be loaded non-invasively into virtually any live cell as acetoxymethyl esters. This turned calcium imaging from a specialist technique into a standard assay used across immunology, neuroscience, cardiology, and developmental biology. He later extended the same design philosophy to genetically encoded sensors built from GFP variants, work recognized with a share of the 2008 Nobel Prize in Chemistry.
Where this matters today
Cardiology. Calcium-channel blockers (verapamil, amlodipine, diltiazem) directly target L-type voltage-gated Ca²⁺ channels to relax vascular smooth muscle and slow cardiac conduction, treating hypertension and arrhythmia. Abnormal ryanodine receptor leak underlies certain inherited arrhythmias (catecholaminergic polymorphic ventricular tachycardia) and contributes to heart failure progression.
Immunology and transplantation. T-cell receptor engagement triggers PLCγ, IP₃, ER Ca²⁺ release, and store-operated entry through Orai1/STIM1, activating calcineurin, which dephosphorylates NFAT to drive interleukin-2 transcription. The immunosuppressants cyclosporine A and tacrolimus (FK506) work by directly inhibiting calcineurin, blocking this pathway — a direct clinical exploitation of the calcium/calmodulin/calcineurin axis that makes organ transplantation possible.
Neuroscience. Genetically encoded calcium indicators (GCaMP family, descended conceptually from Tsien's fluorophore-engineering approach) allow researchers to image neural activity across entire brain regions in awake, behaving animals, because action potentials trigger Ca²⁺ influx through voltage-gated channels at synaptic terminals. CaMKII autophosphorylation, discovered through calcium-signaling studies, is central to long-term potentiation and models of learning and memory.
Reproductive biology and fertility medicine. The calcium wave at fertilization (first visualized with aequorin and fura-2 in sea urchin and mammalian eggs) is required to trigger egg activation; failure of sperm-triggered Ca²⁺ oscillations is a diagnosed cause of some cases of fertilization failure in assisted reproduction, treatable by artificial calcium ionophore activation of the oocyte.
Cancer and cell death research. Store-operated calcium entry (Orai1/STIM1) supports the proliferation and migration of some cancer cell lines, making SOCE components candidate drug targets; mitochondrial calcium overload is a well-established trigger of the mitochondrial permeability transition and downstream apoptotic/necrotic cell death, relevant to ischemia-reperfusion injury and neurodegeneration.
Ask the tutor: How does calcium signaling cross-talk with the cAMP/PKA pathway to integrate two second-messenger systems into one coherent cellular response?
Why can't the cell just use a soluble second messenger like cAMP instead of an ion for everything?
Ions offer properties organic second messengers cannot easily match. Because Ca²⁺ already exists at a massive, actively maintained gradient across membranes, opening a channel produces a response within milliseconds — much faster than synthesizing a new molecule via an enzyme cascade. The same gradient can be tapped locally (a few channels near one synapse) or globally (a whole-cell wave), giving spatial resolution difficult to achieve with a freely diffusible synthesized messenger. Calcium and cAMP are not competitors, though — they are extensively cross-linked, with Ca²⁺/calmodulin regulating certain adenylyl cyclases and phosphodiesterases, illustrating Signal Integration and Crosstalk between pathways rather than redundancy.
What exactly is calcium-induced calcium release, and why doesn't it run away uncontrollably?
Calcium-induced calcium release (CICR) is the phenomenon whereby Ca²⁺ itself acts as a co-agonist at IP₃ receptors and ryanodine receptors, so a small local Ca²⁺ rise triggers further release from the same or neighboring channels, amplifying and propagating the signal as a wave. It does not run away indefinitely because IP₃Rs and RyRs show biphasic Ca²⁺ dependence: very high local Ca²⁺ actually inhibits the channel, and because SERCA pumps and cytosolic buffers rapidly clear released Ca²⁺, limiting the spatial spread and duration of each release event. This self-limiting design is what allows discrete "puffs" and "sparks" rather than a single uncontrolled explosion of Ca²⁺ throughout the cell.
How does the fertilization calcium wave actually trigger egg activation?
At fertilization, sperm delivers a phospholipase C isoform (PLCζ in mammals) into the egg cytoplasm, generating IP₃ that opens egg IP₃ receptors and triggers a wave of Ca²⁺ release across the egg, often repeating as oscillations for the following hour. This Ca²⁺ rise activates CaMKII, which relieves inhibition on the cell cycle machinery (by degrading key inhibitory proteins), allowing the egg to exit its arrested meiotic state and begin embryonic development. Failure of sperm to trigger normal oscillations is a recognized clinical cause of fertilization failure, treatable by artificially activating the egg with a calcium ionophore during assisted reproduction procedures.
Is the resting Ca2+ concentration the same in every cell type?
No; while ~100 nM is a widely cited approximate resting value, the exact figure and the size of Ca²⁺ transients vary substantially between cell types and species, and even within a single cell between organelles and microdomains near channel mouths (which can transiently reach tens of micromolar right at the channel pore). What is conserved is the general principle: cytosolic free Ca²⁺ at rest is always orders of magnitude below both extracellular fluid and ER lumen, whatever the exact numbers, because this gradient is what makes calcium signaling functionally useful across essentially all eukaryotic cells.
Why do calcium-channel blockers and calcineurin inhibitors work as drugs, given that calcium signaling is so widespread?
Because different tissues rely on distinct calcium-handling proteins, drugs can be reasonably selective even though the underlying ion is universal. Calcium-channel blockers target L-type voltage-gated channels concentrated in vascular smooth muscle and cardiac conduction tissue, producing vasodilation and slowed heart rate without directly touching, say, T-cell calcineurin signaling. Calcineurin inhibitors (cyclosporine, tacrolimus) exploit the fact that T-cell activation specifically depends on calcineurin dephosphorylating NFAT, so blocking calcineurin suppresses immune activation relatively selectively, though side effects (including some cardiovascular and renal ones) reflect calcineurin's broader roles elsewhere in the body.
How does calcium signaling relate back to the phosphoinositide pathway?
IP₃, generated by phospholipase C cleavage of PIP₂ as covered under The Phosphoinositide Pathway: IP3 and DAG, is the direct upstream trigger for ER calcium release via IP₃ receptors. DAG, the other cleavage product, activates protein kinase C in parallel, and PKC activity itself is often calcium-dependent (via its C2 domain), meaning the two arms of the phosphoinositide pathway frequently reinforce each other. Calcium signaling should be thought of as the downstream execution arm of the phosphoinositide pathway for many receptors, converting a lipid-cleavage event at the plasma membrane into a fast, quantifiable ionic signal that dozens of calcium-sensor proteins can then read.
What does it mean that calcium signals can be "frequency encoded," and why does this matter for gene expression?
Frequency encoding means that the rate of repeated Ca²⁺ spikes, rather than their individual size, carries the biological information about stimulus strength. This matters for gene expression because transcription factors like NFAT and enzymes like CaMKII respond differently depending on pulse frequency — CaMKII, for instance, can become persistently active ("locked on") after accumulating enough closely spaced pulses, a form of molecular memory. This lets a single signaling ion drive graded, dose-dependent gene expression programs even though the underlying release events are essentially all-or-none at the single-channel level, a theme that recurs in broader discussions of Signal Amplification and the Cascade Concept.