aiwiki.page
English
Science / Biology / amp-activated-protein-kinase

AMP-activated Protein Kinase

AMP-activated protein kinase is a conserved cellular energy sensor that coordinates metabolism, growth, and maintenance through protein phosphorylation.

30 keywords4 linked from8 not yet writtenWritten by AI
EnzymeAdenosine Tripho…BiochemistryMetabolismEukaryoteGlycogenX-ray Crystallog…Cryo-electron Mi…AMP-activa…

AMP-activated protein kinase (AMPK) is a conserved enzyme complex that helps cells match their energy expenditure to their energy supply. It responds to changes in adenine nucleotides and to signals associated with nutrients, calcium, and cellular stress. Once activated, AMPK phosphorylates target proteins to promote processes that restore adenosine triphosphate (ATP) and restrain processes that consume it. Its effects range from rapid changes in glucose transport and fatty-acid metabolism to longer-term changes in gene expression, organelle maintenance, and cell growth.

AMPK is a heterotrimer, consisting of a catalytic α subunit and regulatory β and γ subunits. The γ subunit binds adenine nucleotides, while phosphorylation of a conserved threonine in the α subunit is central to activation. This arrangement allows AMPK to combine information about cellular energy status with signals from upstream protein kinases and from particular subcellular locations. Although often described as a cellular “fuel gauge,” it is not a simple switch controlled by a single metabolite.

AMPK has become a major subject in biochemistry and biomedical research because it connects nutrient availability with metabolism, growth control, and stress adaptation. Its functions are studied in exercise physiology, diabetes, inherited cardiac disorders, and cancer biology. The consequences of AMPK activation depend on the tissue, the molecular composition of the complex, the duration of activation, and the physiological or experimental setting.

Discovery and evolutionary conservation

AMPK emerged from research on the regulation of lipid synthesis rather than from an initial search for a general energy sensor. In 1973, studies identified protein kinase activities capable of inactivating acetyl-CoA carboxylase and HMG-CoA reductase, enzymes involved in fatty-acid and cholesterol synthesis, respectively. The activities were initially investigated separately. Work during the 1980s established that a common AMP-responsive kinase could regulate both enzymes, and the name AMP-activated protein kinase came into use.

This history was important mechanistically. Acetyl-CoA carboxylase consumes ATP to produce malonyl-CoA, whereas cholesterol synthesis requires substantial energy and reducing power. A kinase that inhibits both pathways during energy shortage provides a coordinated means of reducing biosynthetic expenditure. Subsequent identification of many other substrates showed that these two enzymes were examples of a much broader regulatory system rather than isolated targets. Studies associated with D. Grahame Hardie and other investigators helped establish AMPK as a central regulator of energy balance.

The identification and cloning of its subunits in the 1990s clarified that mammalian AMPK was related to the yeast SNF1 protein kinase system. SNF1 had already been recognized as essential for the response of budding yeast to glucose limitation. When readily fermentable glucose becomes scarce, SNF1 helps alter transcription and metabolism so that alternative carbon sources can be used. Plant SnRK1 kinases belong to the same broad evolutionary lineage and coordinate growth and metabolic adaptation when energy or carbon availability is restricted.

The conservation is functional as well as structural, but the regulatory details differ. Mammalian AMPK is directly regulated by AMP and ADP binding to its γ subunit. Yeast SNF1 does not reproduce the complete mammalian pattern of AMP-dependent allosteric activation, although adenine nucleotides influence its regulation. Plant SnRK1 also responds to a distinct network of metabolic signals. These systems therefore illustrate conservation of an energy-stress response architecture without implying identical sensing mechanisms in every organism.

Comparative studies place the AMPK–SNF1–SnRK1 family among ancient regulators of eukaryotic metabolism. Its central problem is universal: biosynthetic demand and ATP generation must remain compatible with survival. In multicellular animals, this ancestral cellular function became integrated with hormones, specialized tissues, and whole-body nutrient allocation. Consequently, findings from yeast or plants can illuminate basic principles of AMPK biology, but the endocrine and tissue-specific roles of mammalian AMPK require separate investigation.

Subunit architecture and molecular organization

Mammalian AMPK contains one α, one β, and one γ subunit. Humans have two genes encoding α subunits, PRKAA1 and PRKAA2; two encoding β subunits, PRKAB1 and PRKAB2; and three encoding γ subunits, PRKAG1, PRKAG2, and PRKAG3. These provide 12 theoretically possible αβγ combinations, although their abundance and assembly are not uniform across tissues. Differences among complexes influence intracellular localization, nucleotide sensitivity, substrate access, and responses to pharmacological activators.

The α subunit contains an amino-terminal protein kinase domain that transfers phosphate from ATP to serine or threonine residues in target proteins. Its activation loop contains a conserved threonine conventionally called Thr172. Phosphorylation at this position greatly increases catalytic activity. The remainder of the α subunit includes an autoinhibitory region and regulatory elements that connect the kinase domain to the nucleotide-binding machinery. Thus, the catalytic site is physically coupled to the energy-sensing portion of the complex rather than operating independently.

The β subunit forms a structural bridge between α and γ. It also contains a carbohydrate-binding module capable of interacting with glycogen and related carbohydrates. This creates a potential connection between AMPK regulation and stored carbohydrate, although the physiological significance depends on the tissue and experimental conditions. Myristoylation near the β-subunit amino terminus contributes to membrane association and to some activation mechanisms. The interface between its carbohydrate-binding module and the α kinase domain forms an important regulatory pocket known as the allosteric drug and metabolite, or ADaM, site.

The γ subunit contains four cystathionine β-synthase, or CBS, repeats arranged into two paired structural units. Together they form nucleotide-binding pockets. Structural studies of mammalian complexes commonly distinguish four potential sites: sites 1 and 3 exchange regulatory nucleotides, site 4 contains tightly bound AMP, and site 2 is generally unoccupied. These sites are not equivalent, and their roles cannot be reduced to a single average nucleotide-binding affinity. Occupancy at regulatory sites changes the arrangement of the α-subunit linker and helps determine whether the activation loop remains phosphorylated.

X-ray crystallography and cryo-electron microscopy have revealed how these domains communicate. In particular, they show how nucleotide occupancy and ADaM-site ligands stabilize conformations that favor kinase activity or protect the activation loop. AMPK is consequently best understood as a dynamic molecular assembly. Its regulation depends not only on whether a ligand is bound, but also on how that binding changes communication among domains and accessibility to other enzymes.

Adenine nucleotides and energy sensing

ATP hydrolysis supports muscle contraction, active transport, biosynthesis, and many other cellular processes. As ATP is consumed, ADP accumulates. The enzyme adenylate kinase interconverts adenine nucleotides through the reaction

2 ADP⇌ATP+AMP.2\,\mathrm{ADP}\rightleftharpoons\mathrm{ATP}+\mathrm{AMP}.

This reaction couples AMP abundance to the balance between ATP and ADP. When adenylate kinase is near equilibrium, relatively modest changes in ATP availability can produce proportionally larger changes in AMP. AMPK exploits this relationship to detect energetic disturbance before a cell has exhausted its ATP supply.

The approximate equilibrium relationship can be written as

K=[ATP][AMP][ADP]2.K=\frac{[\mathrm{ATP}][\mathrm{AMP}]}{[\mathrm{ADP}]^2}.

Rearrangement gives

[AMP][ATP]=K([ADP][ATP])2.\frac{[\mathrm{AMP}]}{[\mathrm{ATP}]} = K\left(\frac{[\mathrm{ADP}]}{[\mathrm{ATP}]}\right)^2.

This expression illustrates the amplified relationship between the AMP:ATP and ADP:ATP ratios under the stated equilibrium assumption. It is not a complete quantitative model of a living cell: nucleotide compartmentation, magnesium binding, metabolic flux, and departures from equilibrium all matter. Nonetheless, it explains why AMP is an informative signal even though its concentration is usually much lower than that of ATP.

AMP activates AMPK through several cooperating mechanisms. It directly stimulates the activity of suitably phosphorylated complexes, promotes activation-loop phosphorylation in appropriate upstream-kinase settings, and protects phosphorylated Thr172 against dephosphorylation. ADP can also protect Thr172 from dephosphorylation, but does not reproduce the characteristic direct allosteric activation caused by AMP. ATP antagonizes these regulatory effects by competing at nucleotide-binding sites. The net response therefore reflects competition among nucleotides, not simply the absolute concentration of AMP.

A related description of cellular energetic state is adenylate energy charge:

Energy charge=[ATP]+12[ADP][ATP]+[ADP]+[AMP].\text{Energy charge} = \frac{[\mathrm{ATP}]+\tfrac{1}{2}[\mathrm{ADP}]} {[\mathrm{ATP}]+[\mathrm{ADP}]+[\mathrm{AMP}]}.

This useful biochemical index should not be treated as a quantity that AMPK directly measures. AMPK responds through particular ligand-binding and phosphorylation mechanisms, with sensitivities that vary among subunit combinations. Two cells with similar bulk energy charge may therefore differ in AMPK activity because they have different isoforms, upstream kinases, or subcellular nucleotide environments.

Energy stress commonly arises during insufficient oxygen delivery, inhibition of mitochondrial ATP production, nutrient deprivation, or unusually high ATP demand. AMPK can respond to each, but activation is not proof that total cellular ATP has fallen dramatically. Metabolic buffering can preserve bulk ATP while changing AMP and ADP, and locally restricted signaling can activate particular AMPK pools. Conversely, AMPK activation can occur through calcium-dependent pathways without a primary change in adenine nucleotide ratios.

Upstream kinases and spatially organized activation

The major upstream kinase for energy-stress activation of AMPK in many mammalian cells is liver kinase B1, encoded by STK11 and usually called LKB1. Its role as an AMPK-activating kinase was established in 2003. LKB1 functions in a complex with the pseudokinase STRAD and the scaffold protein MO25. This complex phosphorylates the AMPK activation loop, allowing nucleotide sensing by AMPK to be translated into a large increase in catalytic output.

LKB1 is often substantially active before energy stress occurs. In such settings, the stress response is not explained simply by switching LKB1 itself on. Instead, AMP-dependent changes in AMPK make the substrate more favorable for phosphorylation and less vulnerable to dephosphorylation. This distinction separates the sensing mechanism from the upstream catalytic machinery. LKB1 also activates several AMPK-related kinases, so the consequences of losing STK11 cannot automatically be attributed solely to loss of AMPK signaling.

A second important upstream enzyme is calcium/calmodulin-dependent protein kinase kinase 2, commonly called CaMKK2 or CaMKKβ. Increases in intracellular calcium promote its activity through calmodulin, allowing AMPK activation even when cellular adenine nucleotide ratios have changed little. This route links energy regulation to calcium signals in settings that include neuronal activity and receptor-mediated signaling. Its contribution varies with CaMKK2 expression, calcium dynamics, and the availability of particular AMPK complexes.

Subcellular organization adds another level of control. AMPK can associate with the lysosomal surface, where nutrient-sensitive assemblies help coordinate kinase activation. Experiments have implicated the vacuolar ATPase, Ragulator, the scaffold AXIN, and LKB1 in this process. Glucose availability can be communicated through the glycolytic enzyme aldolase: changes in occupancy by fructose-1,6-bisphosphate participate in signaling low glucose to the lysosomal machinery. This provides a route by which glucose shortage can activate a local AMPK pool before a large change in whole-cell nucleotide ratios becomes apparent.

AMPK pools also occur near mitochondria, at other membranes, in the cytoplasm, and in the nucleus. Their positioning changes which upstream signals and downstream substrates are accessible. Thus, the same cell may contain AMPK complexes participating in different responses at the same time. This spatial organization helps explain why a whole-cell measurement of Thr172 phosphorylation does not fully describe the pathway. The functional outcome depends on where activation occurs, how long it persists, and which substrate proteins are nearby.

Control of carbohydrate and lipid metabolism

AMPK rapidly adjusts metabolism by phosphorylating existing enzymes and transport regulators. A prominent effect in skeletal muscle is increased uptake of glucose. AMPK phosphorylates proteins including TBC1D1, a regulator of Rab GTPases involved in intracellular membrane traffic. This contributes to movement of GLUT4-containing vesicles toward the cell membrane, increasing the capacity for glucose transport. The pathway provides part of the response to contraction and energy demand, although contraction-stimulated glucose uptake also involves AMPK-independent mechanisms.

Inside cells, AMPK can influence how glucose is used. In some tissues, including the heart, phosphorylation of particular phosphofructokinase-2 isoforms increases fructose-2,6-bisphosphate, a strong activator of glycolysis. The effect helps accelerate ATP production from glucose under stress. Such regulation is isoform-dependent: a mechanism demonstrated in one cell type should not be assumed to operate identically in all tissues. AMPK can also phosphorylate glycogen synthase and inhibit its enzymatic activity, although net glycogen accumulation in intact muscle reflects glucose uptake, substrate availability, insulin signaling, and other regulatory inputs as well.

Lipid metabolism supplies some of the clearest direct AMPK substrates. Acetyl-CoA carboxylase catalyzes ATP-dependent production of malonyl-CoA from acetyl-CoA. Mammals express ACC1, associated principally with fatty-acid synthesis, and ACC2, which helps regulate a malonyl-CoA pool relevant to fatty-acid entry into mitochondria. AMPK-mediated phosphorylation inhibits these enzymes. Reduced ACC activity limits substrate supply for fatty-acid synthesis and can relieve malonyl-CoA inhibition of carnitine palmitoyltransferase 1, facilitating mitochondrial fatty-acid import.

These effects connect suppression of an ATP-consuming pathway with increased access to an energy-yielding fuel. However, increased fatty-acid oxidation is not an obligatory result of every AMPK activation event. Oxygen supply, mitochondrial capacity, fatty-acid availability, and competing substrates influence the actual flux. In severe oxygen limitation, for example, facilitating fatty-acid entry cannot by itself restore oxidative ATP production. Measurements of enzyme phosphorylation therefore need to be distinguished from direct measurements of metabolic rate.

AMPK also phosphorylates HMG-CoA reductase, contributing to inhibition of cholesterol synthesis. Over longer periods it affects transcriptional regulators controlling lipid metabolism, reducing lipogenic programs in many experimental settings. Its regulation of hepatic gluconeogenesis is more complicated. AMPK can influence relevant signaling and transcription, but suppression of hepatic glucose production by a drug or nutrient intervention is not necessarily AMPK-dependent. Acute metabolic changes and transcriptional responses can arise through parallel mechanisms.

The overall pattern is a redistribution of resources rather than indiscriminate stimulation of catabolism. AMPK tends to restrain expensive synthesis and support pathways suited to restoring energy balance, but the available options differ among cells. A muscle fiber, a hepatocyte, and a proliferating immune cell have different fuel sources and physiological obligations. The same kinase consequently produces related but nonidentical metabolic responses in each.

Growth regulation, autophagy, and organelle maintenance

AMPK links energy availability to cell growth through its interaction with the mechanistic target of rapamycin signaling pathway. mTOR complex 1, or mTORC1, promotes protein synthesis and other anabolic processes when nutrients and growth signals are favorable. AMPK can restrain mTORC1 by phosphorylating TSC2, a component of a complex that inhibits the small GTPase Rheb, and by phosphorylating the mTORC1-associated protein Raptor. These mechanisms allow energy stress to counter growth-promoting signals.

This inhibition reduces investment in biosynthetic processes that may be unsustainable when ATP production is limited. AMPK also influences protein synthesis through additional routes, including regulation of elongation-factor signaling in particular settings. Its actions are therefore not confined to a single growth-control node. The magnitude and duration of the response matter: a temporary reduction in translation during stress differs biologically from sustained suppression of tissue growth.

A second major connection is to autophagy, the lysosome-dependent recycling of cellular material. AMPK phosphorylates ULK1, a kinase involved in initiating autophagosome formation, while inhibition of mTORC1 removes an important opposing input. The two pathways cooperate to make recycling more favorable during nutrient limitation. Autophagic degradation can provide metabolites and remove damaged components, but the amount of material ultimately degraded depends on the entire pathway, including autophagosome maturation and lysosomal function.

AMPK also participates in mitochondrial quality control. It phosphorylates mitochondrial fission factor, promoting recruitment of the fission machinery under energy stress. Division can help segregate damaged mitochondrial regions for subsequent removal, although mitochondrial fragmentation is not by itself evidence of successful degradation. AMPK additionally connects with regulators of mitochondrial biogenesis, including PGC-1α, through direct and indirect mechanisms. These responses affect the balance between preserving existing mitochondrial function, replacing damaged components, and expanding oxidative capacity.

The selective removal of mitochondria, known as mitophagy, is related to but distinct from general autophagy. AMPK can contribute to this process through effects on autophagy initiation, mitochondrial dynamics, and stress-response signaling. The relevant routes vary with the insult and cell type; not every form of mitophagy requires the same upstream machinery. Similarly, increased expression of mitochondrial genes does not necessarily establish an increase in functional respiratory capacity.

AMPK-mediated maintenance can preserve cell survival during stress, but its activation is not inherently synonymous with cellular rejuvenation or increased organismal lifespan. Lifespan effects reported in model organisms involve interacting pathways, particular genetic manipulations, and environmental conditions. The molecular evidence supports a role in coordinating maintenance with energetic state; broader outcomes require direct testing at the tissue and organism levels.

Tissue specialization and whole-body physiology

In skeletal muscle, AMPK responds to increased ATP turnover during contraction. The response depends on exercise intensity, duration, training state, recruited fibers, and the subunit combinations present. Complexes containing α2 and β2 are prominent in muscle, while γ3-containing complexes are especially associated with glycolytic skeletal muscle. Their activation contributes to glucose transport and metabolic adaptation. Human exercise studies and genetic experiments nevertheless show that AMPK is one component of a broader contraction-response network rather than the sole mediator of exercise physiology.

Repeated activity produces adaptations over days and weeks, including changes in substrate handling and mitochondrial proteins. AMPK can influence these through regulators of gene expression, but calcium-dependent pathways, mechanical signals, and other kinases act alongside it. The distinction between acute activation and long-term adaptation is important. A brief rise in AMPK activity cannot be assumed to reproduce the full effects of repeated muscle contraction.

In the liver, AMPK helps limit lipid synthesis and coordinates responses to nutrient and energy availability. The liver must also maintain circulating glucose during fasting, making its role different from that of working muscle. Hepatic AMPK activation can alter lipid content and insulin-related physiology in experimental models, but glucose production is controlled by several interacting signals. Changes in glucagon action, substrate delivery, mitochondrial metabolism, and cellular adenine nucleotides can modify hepatic glucose output independently of AMPK.

The heart has exceptionally high and continuous ATP demand. During increased workload or impaired oxygen delivery, AMPK can enhance glucose uptake and support glycolytic ATP production. It also regulates lipid metabolism and longer-term growth pathways. These actions depend on the balance between fuel supply and oxygen availability, and their consequences differ during acute ischemia, restoration of blood flow, and chronic disease. Sustained changes in cardiac AMPK signaling are therefore not equivalent to a short physiological response to energy stress.

In the nervous system, AMPK integrates cellular energetic signals with calcium-dependent activity and nutrient-related information. Hypothalamic AMPK participates in experimental networks controlling feeding and energy expenditure, with different neuronal populations producing different outputs. Its effects in the brain cannot be represented by a single statement that activation either increases or decreases appetite. Neuronal survival, synaptic function, and responses to prolonged energy failure also depend on the timing and severity of activation.

AMPK further affects adipose tissue and cells of the immune system. In immune cells, it helps coordinate metabolic adaptation, growth, and persistence as activation states change. These observations connect energy sensing with inflammatory biology, but the responses differ among macrophages, lymphocytes, and other populations. Whole-body effects consequently arise from interactions among multiple tissues, sometimes with opposing changes in AMPK activity occurring simultaneously.

Human genetics and disease mechanisms

One of the clearest inherited links between AMPK and human disease involves PRKAG2, which encodes the γ2 regulatory subunit. Pathogenic variants cause PRKAG2 syndrome, an inherited disorder commonly associated with cardiac hypertrophy, ventricular pre-excitation, and progressive conduction abnormalities. Cardiac cells can accumulate excessive glycogen, producing a phenotype distinct from the sarcomeric forms of hypertrophic cardiomyopathy. The disorder demonstrates that persistent alteration of an energy-sensing complex can disrupt tissue structure as well as acute metabolism.

The molecular effects of PRKAG2 variants are not identical. Many alter nucleotide regulation and produce inappropriate basal signaling, while also changing the normal response to energetic stress. It is therefore more accurate to describe dysregulated AMPK activity than to assign every variant the same biochemical behavior. Glycogen accumulation, altered cardiac development, and disruption of conduction-system architecture contribute to the clinical phenotype. This genetic evidence also shows why generalized assumptions that more AMPK activity must always be beneficial are incorrect.

AMPK has been investigated extensively in type 2 diabetes because it can increase muscle glucose uptake and suppress hepatic lipid synthesis. These mechanisms provide a biological rationale for drug discovery, but target plausibility and clinical efficacy are separate questions. AMPK activation in cultured cells does not establish improved glucose control in humans, and changes in hepatic glucose production can occur through AMPK-independent pathways. Tissue distribution and the duration of activation are especially important for interpreting metabolic outcomes.

The relationship with cancer is similarly context-dependent. AMPK can restrain anabolic growth and oppose mTORC1 signaling, providing mechanisms compatible with suppression of tumor development in some settings. Its upstream kinase LKB1 is encoded by a tumor-suppressor gene, but LKB1 regulates multiple kinase families. The cancer consequences of STK11 loss therefore cannot be equated simply with a lack of AMPK activity.

Once a tumor exists, AMPK can also help malignant cells tolerate nutrient shortage, oxidative stress, and loss of attachment. By supporting recycling, metabolic flexibility, and maintenance of reducing capacity, it may promote survival under conditions that would otherwise be lethal. These functions are not contradictory: limiting the initiation of uncontrolled growth and supporting the survival of an established stressed cell are different biological processes. Results depend on tumor genotype, tissue origin, microenvironment, and disease stage.

Studies of AMPK in neurodegeneration, inflammatory disorders, and aging likewise show strong dependence on context. Pathway activation may be an adaptive response, a contributor to pathology, or a marker of severe energetic disturbance. Establishing causation requires interventions that distinguish AMPK from accompanying metabolic changes. Evidence from cell culture and animal models provides mechanistic information but does not by itself establish a therapeutic effect in human disease.

Pharmacological modulation and experimental interpretation

AMPK activators fall into several mechanistic classes. Indirect activators alter metabolism so that AMP or ADP rises relative to ATP. Metformin, a biguanide used in type 2 diabetes, can activate AMPK in experimental systems through effects connected to cellular energy metabolism; lysosomal mechanisms have also been described. Its actions depend on concentration, exposure, transport into cells, and tissue context. Not all effects of metformin require AMPK, and high concentrations used in some laboratory studies do not reproduce ordinary clinical exposure.

A widely used experimental compound is AICAR, more precisely the cell-permeable riboside often called AICA riboside. Within cells it is phosphorylated to ZMP, an AMP analog that can activate AMPK. ZMP also affects other AMP-sensitive enzymes and metabolic reactions. A response to AICAR is therefore not sufficient evidence that AMPK caused the response. Its value as a tool lies in combinations with genetic and biochemical tests rather than in absolute pathway specificity.

Direct activators bind AMPK itself. A-769662, reported in 2006, activates complexes through the ADaM site and shows a strong preference for β1-containing forms. Other compounds, including 991, have helped define the structural basis of direct activation. Activity can depend on the β-subunit isoform and phosphorylation of β-subunit Ser108. Thus, a compound that strongly activates one purified AMPK complex may have a different effect in a tissue dominated by another subunit combination.

MK-8722, a direct activator with broad activity across AMPK complexes, improved glucose-related metabolic measures in preclinical studies but also produced cardiac hypertrophy and glycogen accumulation. These findings illustrate a major challenge for the field: beneficial metabolic changes in one tissue can coexist with unwanted effects elsewhere. Isoform selectivity, tissue targeting, and intermittent versus sustained exposure are therefore central variables in interpreting experimental pharmacology.

Inhibitor studies have their own limitations. Compound C, also called dorsomorphin, inhibits AMPK but also inhibits other kinases and signaling pathways. It cannot serve as a selective test of AMPK dependence. Stronger causal evidence combines approaches such as deletion of both catalytic α-subunit genes, re-expression of defined complexes, substrate-site mutation, and measurements of metabolic flux. Even genetic deletion requires care because long-term adaptation may compensate for the missing kinase.

Common biochemical readouts include Thr172 phosphorylation and phosphorylation of acetyl-CoA carboxylase. Neither alone gives a complete account of pathway function: allosteric activation, localization, phosphatase activity, and substrate access can change the relationship between these measurements and actual output. Direct kinase assays, nucleotide measurements by mass spectrometry, and isotopic tracing address different parts of the mechanism. In studies of autophagy or mitochondrial function, endpoint markers must likewise be distinguished from the rates of material degradation or ATP production that they are intended to represent.