{
  "id": "second-messenger",
  "title": "Second Messenger",
  "lang": "en",
  "source_lang": "en",
  "summary": "A second messenger is an intracellular signaling molecule or ion that relays and regulates information from receptors to cellular targets.",
  "category": "science",
  "subcategory": "biology",
  "type": "concept",
  "aliases": [
    "Intracellular second messenger",
    "second-messenger molecule"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/second-messenger/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/second-messenger/",
      "json": "https://aiwiki.page/zh/second-messenger.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "A second messenger is a small intracellular molecule or ion whose concentration, availability, or distribution changes in response to a signal and thereby alters the activity of cellular targets. Second messengers connect signal detection to processes such as secretion, contraction, metabolism, and gene expression. Familiar examples include cyclic adenosine monophosphate, cyclic guanosine monophosphate, calcium ions, inositol trisphosphate, and diacylglycerol.\n\nThe term distinguishes these intracellular intermediates from first messengers, such as hormones and neurotransmitters, which usually convey information between cells. In a typical pathway, a first messenger binds a receptor, the receptor regulates an enzyme or channel, and the resulting second messenger acts on downstream proteins. However, second-messenger signaling is not restricted to responses initiated at the cell surface: intracellular receptors, metabolic changes, mechanical forces, and microbial components can also initiate it.\n\nSecond messengers make signaling amplifiable, reversible, and spatially organized. The same messenger can produce different responses in different cells because its effects depend on where it is generated, how long it persists, and which targets are available. Their study has therefore been central to understanding signal transduction, rather than merely identifying linear chains of biochemical reactions."
    },
    {
      "heading": "Definition, organizing principles, and discovery",
      "text": "A second messenger is defined primarily by its signaling role, not by a single chemical structure. Some are water-soluble nucleotides, some are membrane-associated lipids, and calcium is an inorganic ion. They generally act without requiring new protein synthesis, allowing responses on timescales from milliseconds to minutes. Their downstream consequences can nevertheless last much longer, particularly when they alter transcription, protein abundance, or synaptic organization.\n\nA useful distinction separates second messengers from the proteins that generate or detect them. A receptor, a signaling enzyme, and a protein kinase may all transmit information, but they are not ordinarily called second messengers. Rather, they produce, remove, bind, or respond to the messenger. Likewise, a metabolic intermediate does not become a second messenger merely because its concentration changes. The signaling designation requires evidence that its regulated availability communicates information to a target.\n\nThree functions recur across otherwise different systems: a source generates or releases the messenger, an effector detects it, and a removal mechanism terminates or restricts its action. Sources include nucleotide cyclases, lipid-cleaving enzymes, and calcium channels. Effectors include kinases, ion channels, exchange factors, and transcriptional regulators. Removal occurs through chemical degradation, transport, sequestration, or conversion into another compound. Reversibility is essential: a pathway must distinguish a new stimulus from residual activity caused by an earlier one.\n\nThe second-messenger concept emerged from work on hormone-regulated carbohydrate metabolism. During the 1950s, Earl W. Sutherland Jr. and colleagues investigated how epinephrine and glucagon stimulated glycogen breakdown in liver preparations. Their experiments identified a heat-stable intracellular factor that could connect hormone action at membranes with activation of enzymes in the cytoplasm. This factor was cyclic adenosine monophosphate, or cAMP. Sutherland received the 1971 Nobel Prize in Physiology or Medicine for discoveries concerning the mechanisms of hormone action.\n\nSubsequent research showed that cAMP was one member of a much broader class of intermediates. Calcium signaling, cyclic GMP signaling, and receptor-stimulated phosphoinositide turnover established that cells use chemically distinct messengers in parallel and in combination. The terminology remains functional rather than absolute. Nitric oxide, for example, can be produced inside a cell and signal locally, but it can also diffuse to a neighboring cell and serve as an intercellular messenger. Its classification depends partly on which portion of the signaling pathway is being described."
    },
    {
      "heading": "Cyclic AMP signaling",
      "text": "cAMP is formed from adenosine triphosphate by adenylyl cyclase. Its production is often controlled by a G protein-coupled receptor at the cell membrane. Binding of an extracellular ligand changes the receptor’s conformation and promotes nucleotide exchange on a heterotrimeric G protein. The activated G-protein alpha subunit then regulates the cyclase. The stimulatory alpha subunit, \\(G_{\\mathrm{s}}\\), activates several mammalian adenylyl cyclases, whereas \\(G_{\\mathrm{i}}\\) inhibits particular isoforms. Calcium, other G-protein components, and additional regulators also influence cyclase activity.\n\nA principal cAMP effector is protein kinase A (PKA). In the classical model, inactive PKA contains two regulatory and two catalytic subunits. Binding of cAMP to the regulatory subunits relieves inhibition of the catalytic subunits, allowing phosphorylation of selected proteins. The extent to which these components separate in intact cells depends on their molecular organization and local conditions. PKA can regulate metabolic enzymes, channels, contractile proteins, and nuclear factors. In liver, the glucagon–cAMP pathway promotes glycogen mobilization and supports glucose production during fasting.\n\ncAMP also acts through effectors other than PKA. Exchange proteins directly activated by cAMP, known as EPAC proteins, regulate small GTPases of the Rap family and influence secretion, adhesion, and other processes. Cyclic nucleotide-gated channels bind cyclic nucleotides directly, while hyperpolarization-activated cyclic nucleotide-gated channels combine voltage sensitivity with cyclic-nucleotide regulation. These alternative effectors explain why a cellular response to cAMP cannot automatically be attributed to PKA.\n\nSignal termination depends substantially on cyclic nucleotide phosphodiesterases, which hydrolyze cAMP to 5′-AMP. Mammals possess multiple phosphodiesterase families with different substrate preferences, regulatory domains, and cellular distributions. Consequently, the measured cAMP content of an entire cell may conceal sharply different concentrations near individual receptors or effectors. Production and destruction occur simultaneously, and their local balance determines which targets become activated.\n\nThe biological meaning of cAMP is especially clear when different tissues are compared. In cardiac muscle, beta-adrenergic stimulation increases cAMP signaling and modifies calcium handling and contraction. In many smooth-muscle settings, cAMP favors relaxation through a different complement of targets. In olfactory sensory neurons, odorant receptors can stimulate cAMP production, opening cyclic nucleotide-gated channels and initiating an electrical response. The shared messenger does not specify the final outcome by itself; the surrounding molecular machinery supplies that specificity."
    },
    {
      "heading": "Cyclic GMP, nitric oxide, and sensory transduction",
      "text": "Cyclic guanosine monophosphate (cGMP) is synthesized from guanosine triphosphate by guanylyl cyclases. Two major receptor arrangements generate it. Membrane-associated receptor guanylyl cyclases combine ligand recognition with an intracellular catalytic domain; important examples respond to natriuretic peptides. Soluble guanylyl cyclase is an intracellular receptor for nitric oxide. These arrangements allow cGMP to convey information from both peptide signals and a small diffusible gas.\n\nIn vascular signaling, nitric oxide is synthesized by nitric oxide synthases from arginine. Nitric oxide released by endothelial cells can diffuse into adjacent vascular smooth-muscle cells, where it binds the heme group of soluble guanylyl cyclase and stimulates cGMP production. cGMP activates protein kinase G, which acts on targets involved in calcium regulation and the contractile apparatus. Together, these effects favor relaxation. Robert F. Furchgott, Louis J. Ignarro, and Ferid Murad received the 1998 Nobel Prize in Physiology or Medicine for discoveries concerning nitric oxide as a signaling molecule in the cardiovascular system.\n\nNatriuretic peptides use a different input mechanism. Atrial and B-type natriuretic peptides activate the membrane receptor guanylyl cyclase GC-A, while C-type natriuretic peptide principally activates GC-B. These pathways participate in cardiovascular, renal, and skeletal physiology. Although both membrane and soluble cyclases generate cGMP, their signals need not reach identical targets. Receptor location, phosphodiesterase activity, and effector distribution can maintain functionally distinct cGMP pools within the same cell.\n\nPhototransduction in vertebrate rods illustrates that a second-messenger signal can be encoded by a decrease rather than an increase. In darkness, cGMP holds cyclic nucleotide-gated channels open in the outer-segment membrane. Light-activated rhodopsin activates the G protein transducin, which stimulates phosphodiesterase 6. The resulting fall in cGMP closes the channels and hyperpolarizes the photoreceptor. Calcium-dependent regulation of guanylyl cyclase subsequently contributes to recovery and adaptation. Thus, messenger removal is itself the central information-carrying event.\n\ncGMP is degraded by phosphodiesterases to 5′-GMP. Some phosphodiesterases preferentially hydrolyze cGMP, while others act on both cyclic nucleotides or are regulated by one nucleotide while degrading another. This creates direct biochemical connections between cAMP and cGMP pathways. For example, cGMP can stimulate cAMP degradation through PDE2 or inhibit it through PDE3, depending on the enzymes present. Such interactions make the effect of changing one cyclic nucleotide dependent on the broader signaling environment."
    },
    {
      "heading": "Calcium as an intracellular messenger",
      "text": "Calcium differs from synthesized second messengers because signaling usually redistributes an existing ion rather than producing it chemically. Resting free cytosolic calcium in many animal cells is approximately \\(100\\ \\text{nM}\\), whereas extracellular free calcium is roughly \\(1\\)–\\(2\\ \\text{mM}\\). Calcium is also concentrated within intracellular stores, particularly the endoplasmic reticulum. Opening a calcium-permeable channel therefore allows rapid movement down a large electrochemical gradient. Pumps and exchangers subsequently restore the gradients, directly or indirectly using metabolic energy.\n\nCalcium enters the cytosol through several routes. Voltage-gated calcium channels respond to changes in membrane potential, while receptor-operated channels couple entry to ligand recognition. Intracellular release channels include inositol trisphosphate receptors and ryanodine receptors. In skeletal muscle, membrane voltage sensors mechanically couple to ryanodine receptors in the sarcoplasmic reticulum. In cardiac muscle, calcium entering through surface channels triggers further calcium release from the sarcoplasmic reticulum, a mechanism called calcium-induced calcium release.\n\nCells detect calcium through proteins with suitable binding sites. Calmodulin binds calcium and regulates numerous enzymes and channels, including calcium/calmodulin-dependent protein kinases and the phosphatase calcineurin. Other calcium sensors are specialized for particular tasks. Troponin C participates in the regulation of contraction in striated muscle, while synaptotagmins help couple local calcium entry to rapid vesicle fusion at many synapses. The response depends on each sensor’s affinity, binding kinetics, abundance, and proximity to the calcium source.\n\nCalcium signals have characteristic spatial and temporal forms. A brief opening of a few channels can create a highly localized microdomain that activates nearby targets without greatly changing average cytosolic calcium. Coordinated release can produce cell-wide waves or repetitive oscillations. Different oscillation frequencies can produce different downstream responses because target proteins activate and deactivate at different rates. Sustained, widespread calcium elevation is therefore not equivalent to a sequence of short, localized pulses, even when the total amount of calcium movement is similar.\n\nStore-operated calcium entry connects internal stores with the plasma membrane. When endoplasmic-reticulum calcium falls, STIM proteins undergo conformational changes and accumulate at contacts with the plasma membrane, where they activate Orai channels. Calcium entry then supports prolonged signaling and store refilling. In T cells, this pathway helps sustain calcineurin-dependent activation of NFAT transcription factors. Calcium removal involves endoplasmic-reticulum pumps, plasma-membrane pumps, and sodium–calcium exchangers; mitochondria also take up calcium and influence local signaling, although their effects depend strongly on cellular conditions."
    },
    {
      "heading": "Phosphoinositides, inositol trisphosphate, and diacylglycerol",
      "text": "A major second-messenger pathway begins with phosphatidylinositol 4,5-bisphosphate, commonly abbreviated PIP2, in the cytoplasmic leaflet of a membrane. PIP2 belongs to the phosphoinositides, a family of phosphorylated membrane lipids. Activated phospholipase C cleaves PIP2 into two chemically different products: soluble inositol 1,4,5-trisphosphate (IP3) and membrane-associated diacylglycerol (DAG). One reaction thus creates two messenger branches with distinct destinations and targets.\n\nDifferent receptor systems activate different phospholipase C isoforms. Many G protein-coupled receptors act through \\(G_{\\mathrm{q}}\\)-family proteins and phospholipase C beta. Receptor tyrosine kinases and some immune-receptor pathways activate phospholipase C gamma through protein recruitment and phosphorylation. In either case, the immediate lipid-cleavage reaction is similar, but the initiating machinery determines where, when, and for how long messenger production occurs.\n\nIP3 diffuses through the cytosol and binds IP3 receptors, which are calcium-release channels in the endoplasmic reticulum. Their opening depends on both IP3 and cytosolic calcium, creating opportunities for regenerative release and oscillation. IP3 is therefore often an upstream controller of calcium, itself another second messenger. IP3 is removed through dephosphorylation or converted into more highly phosphorylated inositol phosphates. Its lifetime, together with receptor sensitivity and the state of calcium stores, shapes the resulting calcium response.\n\nDAG remains preferentially associated with membranes because of its hydrophobic acyl chains. It recruits and activates proteins containing appropriate lipid-binding domains, most notably members of the protein kinase C family. Conventional protein kinase C isoforms require both DAG and calcium, enabling them to detect the coincidence of the two branches produced by PIP2 cleavage. Novel isoforms respond to DAG without requiring calcium, whereas atypical isoforms do not use the same DAG-dependent activation mechanism. DAG also regulates other effectors, including some proteins that control small GTPases.\n\nPhosphoinositide signaling extends beyond PIP2 cleavage. Phosphoinositide 3-kinases generate phosphatidylinositol 3,4,5-trisphosphate, or PIP3, which recruits proteins such as Akt and PDK1 to membranes through lipid-binding domains. The phosphatase PTEN removes its 3-position phosphate, helping terminate the signal. Unlike a soluble messenger, PIP3 changes the membrane’s local recruitment properties. Moreover, PIP2 itself directly regulates many channels and membrane-associated proteins, so its depletion can carry information independently of IP3 and DAG production. Lipid signaling therefore combines messenger generation with modification of the membrane platform on which signaling complexes assemble."
    },
    {
      "heading": "Redox signals and cyclic nucleotide messengers beyond cAMP and cGMP",
      "text": "Hydrogen peroxide can act as a regulated intracellular signal in addition to being a product of oxidative metabolism. It may be generated downstream of receptors through NADPH oxidases or arise from other cellular sources. Its signaling effects often involve oxidation of particular cysteine residues in proteins, altering catalytic activity or molecular interactions. This connects second-messenger signaling to oxidation–reduction chemistry, but the distinction between signaling and damage depends on concentration, localization, duration, and the chemical susceptibility of targets.\n\nHydrogen peroxide does not react rapidly or selectively with every cellular protein. Peroxiredoxins, glutathione-dependent systems, and other antioxidant mechanisms strongly constrain its availability. Some signaling pathways involve redox relays in which a peroxide-reactive protein transfers an oxidative equivalent to another target. In other cases, local inhibition of a protein tyrosine phosphatase changes the balance of phosphorylation. These mechanisms explain how limited, regulated peroxide production can influence signaling without implying indiscriminate oxidation throughout the cell.\n\nBacteria use a wide repertoire of nucleotide second messengers. Cyclic di-GMP, composed of two guanosine monophosphate units joined cyclically, commonly regulates transitions between motile and surface-associated states. Enzymes with GGDEF domains synthesize it, while EAL-domain or HD-GYP-domain enzymes commonly participate in its degradation. Its effectors include proteins and RNA riboswitches. In many bacterial species, increased cyclic di-GMP promotes extracellular matrix production or biofilm formation, although specific outputs vary among organisms and local signaling circuits.\n\nOther bacterial messengers include cyclic di-AMP and the alarmones ppGpp and pppGpp. Cyclic di-AMP participates in processes such as ion transport, osmotic regulation, and cell-envelope homeostasis. The alarmones regulate the stringent response, coordinating cellular activities with nutrient availability and other stresses. These systems show that second messengers need not originate from an extracellular hormone receptor. They can also communicate the internal physiological state of a cell to multiple targets.\n\nA prominent connection between nucleotide messengers and innate immunity is 2′3′-cyclic GMP–AMP, or cGAMP. In mammalian cells, the enzyme cGAS can be activated by accessible double-stranded DNA in inappropriate cellular locations and synthesize cGAMP from ATP and GTP. cGAMP binds the adaptor protein STING, initiating signaling that can promote type I interferon production and other responses. STING can also recognize certain bacterial cyclic dinucleotides. Although cGAMP is an intracellular second messenger, it can additionally move between cells through several routes, illustrating that intracellular and intercellular signaling categories can overlap."
    },
    {
      "heading": "Spatial organization, amplification, and information encoding",
      "text": "The traditional drawing of a signaling pathway as a sequence of arrows conceals its spatial organization. Second messengers are produced at particular membranes, organelles, or molecular complexes, and they encounter targets distributed unevenly throughout the cell. Their range depends on diffusion, binding, degradation, and barriers to movement. A soluble messenger may therefore remain functionally localized, while a membrane-associated messenger can spread laterally but remain confined to one membrane surface.\n\nA simplified reaction–diffusion equation describes some of these relationships:\n\n\\[\\frac{\\partial c}{\\partial t} = D\\nabla^2 c + S(\\mathbf{x},t)-kc,\\]\n\nwhere \\(c\\) is messenger concentration, \\(D\\) is its diffusion coefficient, \\(S(\\mathbf{x},t)\\) represents local production, and \\(k\\) is a first-order removal rate. Under these assumptions, a characteristic spatial range scales as\n\n\\[\\lambda \\sim \\sqrt{\\frac{D}{k}}.\\]\n\nReal cells add reversible binding, nonlinear enzymatic reactions, organelle geometry, and active transport, but the model explains why rapid removal can confine even a readily diffusible messenger.\n\nScaffolding proteins provide another level of organization. A-kinase anchoring proteins bring PKA into defined complexes that can also contain receptors, phosphodiesterases, phosphatases, and substrates. Such assemblies place messenger detection close to selected targets while limiting access to others. Calcium microdomains similarly couple channels to nearby sensors. Specificity thus depends not only on the chemical identity of the messenger but also on the physical arrangement of its production and detection machinery.\n\nAmplification occurs when one activated component generates many downstream molecules or modifies many targets. A receptor may activate multiple G proteins, an enzyme may produce numerous messenger molecules, and an activated kinase may phosphorylate many substrates. Amplification is nevertheless limited by substrate availability, enzyme saturation, messenger removal, and feedback. Negative feedback can shorten a response or produce adaptation, while positive feedback can create thresholds, regenerative waves, or persistent states.\n\nSignals also encode information through timing. A transient pulse, sustained plateau, and oscillatory series can have different effects even if they reach the same maximum concentration. Targets decode these patterns according to their activation and recovery rates. Cross-talk further complicates interpretation: calcium can regulate adenylyl cyclases and phosphodiesterases, cyclic nucleotides can alter calcium channels, and lipid signaling can control both. Consequently, messenger concentration alone is rarely a complete description of the information available to the cell; location, duration, frequency, and the state of the target network also matter."
    },
    {
      "heading": "Physiological roles, disease mechanisms, and pharmacology",
      "text": "Second-messenger pathways connect extracellular conditions to coordinated physiological responses. In endocrine regulation, peptide hormones and catecholamines use these pathways to change enzyme activity, secretion, and transport without waiting for new protein synthesis. Vasopressin signaling through the V2 receptor, cAMP, and PKA promotes accumulation of aquaporin-2 water channels at the apical membrane of kidney collecting-duct cells. This example links receptor activation to membrane trafficking and, ultimately, whole-body water balance.\n\nIn the nervous system, second messengers modify both immediate electrical responses and longer-lasting cellular properties. Calcium entry triggers neurotransmitter release at many synapses, whereas cAMP- and calcium-dependent pathways regulate channel activity and synaptic proteins. These pathways can also alter transcription and contribute to forms of neuroplasticity. In immune cells, antigen-receptor signaling generates IP3, DAG, and calcium signals that cooperate with other pathways to regulate activation. Their effects depend on additional receptor inputs rather than on a single messenger acting in isolation.\n\nSome diseases illustrate the consequences of failed signal termination. Cholera toxin ADP-ribosylates the stimulatory G-protein alpha subunit, impairing its GTPase-dependent shutoff and producing prolonged activation of adenylyl cyclase. In intestinal epithelial cells, elevated cAMP promotes chloride secretion through CFTR, with accompanying electrolyte and water loss. Pertussis toxin acts differently: it modifies \\(G_{\\mathrm{i/o}}\\)-family alpha subunits and interferes with their coupling to receptors. These bacterial toxins helped establish the distinct roles of stimulatory and inhibitory G proteins.\n\nGenetic alterations can affect messenger production, sensing, or removal. Activating mutations in GNAS, which encodes the stimulatory G-protein alpha subunit, occur in several endocrine and developmental disorders. Abnormal phosphoinositide signaling is important in cancer, including tumors with activating PI3K-pathway alterations or loss of PTEN function. These examples do not make any messenger intrinsically beneficial or harmful. The consequences arise from altered pathway activity in a particular tissue and developmental or physiological context.\n\nSeveral established drug classes act on second-messenger systems. Beta-adrenergic receptor agonists and antagonists alter receptor-driven cAMP signaling. Organic nitrates act through nitric-oxide-related activation of soluble guanylyl cyclase, while PDE5 inhibitors reduce cGMP breakdown. PDE4 inhibitors alter cAMP degradation in selected clinical settings. Such drugs demonstrate that changing messenger lifetime can be as consequential as changing its production. Their effects and adverse effects also reflect the distribution of receptors, enzymes, and targets across tissues, rather than a uniform response wherever the messenger is present."
    },
    {
      "heading": "Experimental measurement and causal analysis",
      "text": "Early second-messenger research relied heavily on biochemical extraction, tissue fractionation, and assays of enzyme activity. Competitive binding assays and immunoassays subsequently enabled sensitive measurements of cyclic nucleotides. Mass spectrometry can identify and quantify numerous nucleotide and lipid messengers, often after chromatographic separation. These methods provide chemical specificity, but measurements from pooled cells or homogenized tissue may obscure short-lived events and intracellular gradients.\n\nFluorescent calcium indicators transformed the study of living cells. Synthetic indicators such as fura-2 allow calcium-dependent changes in fluorescence to be measured, with ratiometric approaches helping control for differences in indicator amount and illumination. Genetically encoded indicators, including GCaMP-family sensors, enable calcium activity to be monitored in selected cells and subcellular compartments. Their fluorescence does not reproduce calcium concentration instantaneously: binding kinetics, affinity, dynamic range, and optical acquisition rate all shape the recorded signal.\n\nCyclic nucleotides and other messengers can be monitored with genetically encoded biosensors. Some use Förster resonance energy transfer between fluorescent proteins, while others use changes in a single fluorescent protein or redistribution of a messenger-binding domain. Sensors can be targeted to membranes, nuclei, mitochondria, or signaling complexes. This has revealed responses that differ substantially between compartments even when a whole-cell assay suggests only a modest change.\n\nEvery sensor introduces interpretive constraints. A calcium indicator can buffer the ion it measures; a cyclic-nucleotide-binding sensor can bind a meaningful fraction of a small local messenger pool. Sensor saturation can conceal further increases, and changes in pH or optical conditions can affect fluorescence independently of messenger concentration. Calibration, appropriate controls, and comparison between different measurement principles are therefore important. For membrane lipids, a fluorescent binding-domain probe may also respond to interacting molecules or perturb the lipid pool it is intended to report.\n\nEstablishing causation requires perturbation as well as observation. Researchers combine receptor stimulation with genetic disruption of messenger-producing enzymes, alteration of degradation pathways, and modification of messenger-binding sites in effectors. Light-sensitive compounds can release a messenger at a chosen time, while optogenetic cyclases can generate cyclic nucleotides in selected locations. Pharmacological inhibitors remain useful, but their selectivity and concentration-dependent off-target effects must be considered. Distinguishing messenger abundance from downstream effector activity is equally important: the same measured concentration may produce different responses if target abundance, localization, or phosphorylation state has changed."
    }
  ],
  "links": [
    "molecule",
    "ion",
    "metabolism",
    "gene-expression",
    "neurotransmitter",
    "signal-transduction",
    "enzyme",
    "cyclic-amp",
    "adenosine-triphosphate",
    "cell-membrane",
    "ligand",
    "neuron",
    "endoplasmic-reticulum",
    "electrochemical-gradient",
    "membrane-potential",
    "t-cell",
    "mitochondrion",
    "redox-reaction",
    "bacteria",
    "biofilm",
    "innate-immunity",
    "dna",
    "diffusion",
    "feedback",
    "transcription-biology",
    "neuroplasticity",
    "cancer",
    "mass-spectrometry",
    "chromatography",
    "ph",
    "optogenetics"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}