ATP — The Energy Currency

Why ATP matters

• Universal across life. Every known organism — bacteria, archaea, eukaryotes — uses ATP as the primary energy currency. The conservation across 3.5 billion years of evolution implies ATP arose before the last universal common ancestor; alternative chemistries (acetyl phosphate, polyphosphates) exist as supplements but never as the main currency.
• Coupling drives unfavorable biosynthesis. Glucose + Pi → glucose-6-phosphate is +13.8 kJ/mol (unfavorable); coupling to ATP → ADP + Pi (−30.5 kJ/mol) gives a net −16.7 kJ/mol. Hexokinase performs this directly via a phosphoryl-transfer mechanism, the template for hundreds of biosynthetic kinases.
• Daily flux is enormous relative to pool size. ~250 g of ATP at any instant; ~50 kg of turnover per day means the pool flips ~1000 times. During vigorous exercise muscle ATP turnover exceeds 1 kg/minute. The system runs on flow, not storage.
• Phosphorylation potential ΔG_p is the real energy unit. Standard ΔG°' is −30.5 kJ/mol but cells maintain [ATP]/[ADP][Pi] far from equilibrium; Q ≈ 10⁻⁷ in vivo, so RT ln Q adds ~−39 kJ/mol pushing the total to ~−50 kJ/mol — the actual driving force for cellular work.
• ATP synthase is the most efficient molecular machine known. The F₁F₀ rotor produces 3 ATP per 360° rotation, driven by ~10–14 protons crossing the membrane down their electrochemical gradient. Energetic efficiency is ~70%, far above any human-built turbine. Boyer and Walker shared the 1997 Nobel for the rotational mechanism.
• Phosphocreatine extends the burst window. Muscle stores ~25 mM phosphocreatine (vs. ~5 mM ATP) with ΔG°' = −43 kJ/mol (~12 kJ/mol more exergonic than ATP). Creatine kinase regenerates ATP from ADP using phosphocreatine in <10 ms — sufficient for ~10 s of all-out sprint before glycolysis takes over.
• ATP signaling beyond energy. Extracellular ATP binds purinergic receptors (P2X, P2Y) regulating immune response, pain, and neurotransmission. Cyclic AMP (cAMP), made from ATP by adenylyl cyclase, is the canonical second messenger described by Sutherland (Nobel 1971).

Common misconceptions

• "High-energy bonds" are not unusually strong covalent bonds. The phosphoanhydride bond enthalpy is ordinary (~30 kJ/mol C–O bond strength); what matters is the free energy difference between reactants and hydrolysis products, driven by charge repulsion relief and resonance stabilization of phosphate.
• ATP is not stored energy. It is a transient currency — at typical concentrations of 1–10 mM and turnover of seconds, a cell would be dead in minutes if synthesis stopped. Long-term energy storage is glycogen, fat, or starch; ATP is the spending money.
• The standard ΔG°' is not what cells use. −30.5 kJ/mol is defined for 1 M ATP/ADP/Pi at pH 7. Real cells operate far from equilibrium with [ATP]/[ADP] ~10, [Pi] ~5 mM, putting actual ΔG closer to −50 kJ/mol. Quoting the standard value as "the" energy of ATP is a textbook simplification.
• γ-phosphate is not the only point of cleavage. ATP can hydrolyze to ADP + Pi (γ cleavage, most common) or to AMP + PPi (α-β cleavage, used when needing extra energy because PPi → 2 Pi adds another ~−19 kJ/mol). Aminoacyl-tRNA synthetases use the AMP + PPi route to drive amino acid activation effectively irreversibly.
• ATP is not made directly by glycolysis at large scale. Substrate-level phosphorylation in glycolysis nets only 2 ATP per glucose. The other ~30 come from oxidative phosphorylation in mitochondria via ATP synthase coupled to the electron transport chain.
• Mg²⁺ is not optional. Active ATP in cells is the MgATP²⁻ complex, with one or two Mg²⁺ ions chelated to the phosphate oxygens. Free ATP⁴⁻ is rare and chemically different — most enzymes bind MgATP, not ATP alone.

Mechanism: phosphoanhydride hydrolysis

ATP consists of an adenine base, a ribose sugar, and three phosphate groups linked α-β and β-γ by phosphoanhydride bonds. At pH 7 the molecule carries four negative charges (one at α, one at β, two at γ in the dianionic form). Hydrolysis at the γ position cleaves the β-γ bond, transferring a phosphoryl group to water (or in coupled reactions, to a substrate). The energy release of −30.5 kJ/mol comes from three additive sources: (1) electrostatic relief — separating four like charges on a single molecule into ADP³⁻ and Pi²⁻ reduces self-repulsion; (2) resonance stabilization — orthophosphate has more equivalent resonance structures (12) than the phosphoanhydride (3), so the products have lower-energy electronic ground states; (3) entropy — releasing one molecule into two with independent translational/rotational freedom raises ΔS, contributing −T ΔS.

Inside cells, the [ATP]/[ADP][Pi] ratio is held at ~10⁷ M⁻¹ by oxidative phosphorylation, which pumps the system far from equilibrium. The actual free energy is ΔG = ΔG°' + RT ln Q where Q = [ADP][Pi]/[ATP]. Plugging in Q ≈ 10⁻⁷ M at 37 °C gives RT ln Q ≈ −41 kJ/mol, so the total ΔG is roughly −30.5 − 41 ≈ −71 kJ/mol in the most extreme cases or ~−50 kJ/mol in typical mammalian cells. This is the phosphorylation potential ΔG_p.

Lipmann's 1941 paper "Metabolic generation and utilization of phosphate bond energy" introduced the high-energy phosphate concept and the squiggle (~) notation, drawing on Engelhardt and Lyubimova's 1939 demonstration that myosin is an ATPase. Lohmann had isolated ATP from muscle in 1929, and Cori, Cori, and Houssay's 1947 Nobel-honored work showed ATP's role in glucose phosphorylation. ATP synthase was characterized by Racker, Boyer, and Walker through the 1960s–1990s, with the rotational catalysis mechanism finalized by Boyer (1979) and confirmed crystallographically by Walker (1994). The system is closed: glucose oxidation makes ATP, ATP drives biosynthesis and motion, hydrolysis returns ADP for re-phosphorylation.

Variant comparison: high-energy phosphate compounds

Compound	ΔG°' hydrolysis (kJ/mol)	Cellular concentration	Function	Made by
Phosphoenolpyruvate (PEP)	−61.9	~0.1 mM	Pyruvate kinase substrate	Enolase (glycolysis step 9)
1,3-Bisphosphoglycerate	−49.3	~0.05 mM	Glycolytic intermediate	GAPDH
Phosphocreatine	−43.0	~25 mM (muscle)	Rapid ATP buffer	Creatine kinase
ATP → ADP + Pi (γ)	−30.5	~5 mM	Universal energy currency	ATP synthase, glycolysis
GTP → GDP + Pi	−30.5	~0.5 mM	Translation, signaling	NDP kinase, succinyl-CoA synthetase
ATP → AMP + PPi (α-β)	−45.6	—	Activation reactions	tRNA synthetases, fatty acyl-CoA synthase
Glucose-6-phosphate	−13.8	~0.1 mM	Low-energy phosphate	Hexokinase
Acetyl-CoA	−31.5	~0.01 mM	2-carbon donor	Pyruvate dehydrogenase, β-oxidation

Applications and examples

• Hexokinase glycolysis step 1. Glucose + ATP → glucose-6-phosphate + ADP. ΔG°' = −16.7 kJ/mol after coupling, traps glucose inside the cell because phosphorylated glucose cannot cross GLUT transporters.
• Myosin ATPase in muscle. Each cross-bridge cycle hydrolyzes one ATP to slide an actin filament ~10 nm. A typical human at rest contracts skeletal muscle enough to consume ~10²⁰ ATP molecules per second.
• Na⁺/K⁺ ATPase. The plasma-membrane pump uses ~30% of resting basal energy, ejecting 3 Na⁺ and importing 2 K⁺ per ATP hydrolyzed. Discovery: Jens Skou, Nobel 1997.
• Creatine supplementation in athletics. 5 g/day for 4 weeks raises muscle phosphocreatine ~20–30%, extending the ATP-buffer window during repeated 10-second sprints. Documented across hundreds of trials since 1992.
• Aminoacyl-tRNA charging. Each amino acid is activated as aminoacyl-AMP (ATP → AMP + PPi, the α-β cleavage path); PPi is hydrolyzed by inorganic pyrophosphatase, making the activation effectively irreversible at the cost of two phosphate bonds per amino acid loaded.