Cellular Physiology & Membrane Potentials

Overview & definition

The cellular physiology — the membrane potential (the electrical difference across the cell membrane). The resting potential (minus 70 mV). The action potential. The Na/K ATPase. The Nernst and the Goldman. The clinical correlations.[1]

Cellular physiology is the biophysical and biochemical basis of every phenomenon intensivists manipulate: the electrical excitability that antiarrhythmics and neuromuscular blockers target, the excitation-contraction coupling that inotropes and calcium-channel blockers alter, and the cellular energetics (glycolysis, the Krebs cycle, oxidative phosphorylation) whose failure produces the lactataemia, cytopathic hypoxia and multiorgan dysfunction of sepsis. This topic draws together membrane structure, bioelectricity, muscle coupling, metabolism, oxygen sensing, and the modes of cell death — the cellular substrate of critical illness. [1]

Cell membrane structure — the phospholipid bilayer

Every cell is bounded by the plasma membrane, a 5–8 nm phospholipid bilayer that is the substrate for the membrane potential, for signalling and for transport. The membrane is NOT a passive barrier: it is a dynamic, selectively permeable structure whose protein complement determines the electrophysiology of the cell. [1]

The lipid bilayer

• Phospholipid molecules are amphipathic — a hydrophilic phosphate head and two hydrophobic fatty-acid tails. In water they self-assemble tail-to-tail into a bilayer, with the polar heads facing the aqueous extracellular and intracellular compartments.
• The hydrophobic core is essentially impermeable to ions (Na+, K+, Ca2+, Cl−), to large polar molecules (glucose, amino acids) and to proteins — all of which require specific transport proteins. It is permeable to lipid-soluble substances (O2, CO2, steroid hormones, volatile anaesthetics) which diffuse down their concentration gradient directly through the lipid.
• Cholesterol is intercalated between phospholipids (roughly 1 cholesterol per phospholipid) — it modulates membrane fluidity (rigidifies at high temperature, prevents crystallisation at low temperature) and decreases water permeability.
• Asymmetric distribution: the outer (extracellular) leaflet is enriched in phosphatidylcholine and sphingomyelin; the inner (cytoplasmic) leaflet in phosphatidylethanolamine and phosphatidylserine. Phosphatidylserine externalisation to the outer leaflet is an "eat me" signal for apoptosis — the basis of the Annexin V assay. [1]

Membrane proteins — channels, carriers, receptors, pumps

The functional specificity of the membrane resides in its proteins, classified by function: [1]

[1]

Gating mechanisms — how channels open

• Voltage-gated channels open in response to a change in membrane potential (a charged S4 transmembrane segment — the "voltage sensor" — moves outward as the membrane depolarises). Examples: voltage-gated Na+, K+, Ca2+ channels — the channels of the action potential.
• Ligand-gated channels open when a chemical ligand binds an extracellular or intracellular site. Examples: nicotinic ACh receptor (ACh), GABA-A receptor (GABA, benzodiazepines), NMDA receptor (glutamate).
• Mechanically-gated channels open with physical deformation (stretch, pressure). Examples: cochlear hair cells, vascular baroreceptors.
• Second-messenger-gated channels open in response to an intracellular messenger (cAMP, cGMP, IP3, Ca2+). Example: cGMP-gated channels of photoreceptors; IP3 receptors on the sarcoplasmic reticulum. [1]

Resting membrane potential

• Minus 70 mV (inside negative).[1]
• Na/K ATPase — 3 Na out, 2 K in (electrogenic).[1]
• K leak — the dominant determinant (K diffuses out; the membrane most permeable to K).[1]
• Nernst — E(K) = minus 90 mV (close to resting = K dominant).[1]
• Goldman — the permeability-weighted average of all ions.[1]

Why the resting potential exists — three determinants

The resting membrane potential (RMP) — typically minus 70 to minus 90 mV in excitable cells (inside negative relative to outside) — is generated by three factors operating together: [1]

1. Concentration gradients established and maintained by the Na+/K+ ATPase. Intracellular K+ is high (approx 140 mmol/L) while extracellular K+ is low (approx 4 mmol/L); intracellular Na+ is low (approx 10–15 mmol/L) while extracellular Na+ is high (approx 140 mmol/L). The pump moves 3 Na+ out and 2 K+ in per ATP hydrolysed — it is electrogenic (net one positive charge out per cycle, contributing minus 4 to minus 10 mV to the RMP directly).
2. Selective membrane permeability. At rest the membrane is 25–100 times more permeable to K+ than to Na+ (because of constitutively open K+ "leak" / two-pore-domain K2P channels). K+ therefore diffuses OUT down its concentration gradient, carrying positive charge out and leaving the inside negative. Na+ cannot follow to any significant extent (the membrane resists it), so the negative charge accumulates.
3. The Donnan effect / fixed intracellular anions. Large impermeant intracellular anions (proteins, organic phosphates) cannot leave the cell, contributing a small sustained negative intracellular charge. [1]

The RMP therefore sits CLOSE to the potassium equilibrium potential EK, because K+ is the dominant permeant ion at rest. [1]

The Nernst equation — equilibrium potential for a SINGLE ion

The Nernst equation gives the membrane potential at which a single ion is at electrochemical equilibrium (no net flux — the electrical force exactly balances the chemical/concentration force): [1]

E(ion) = (RT / zF) × ln( [ion]outside / [ion]inside ) [1]

Where R is the gas constant, T the absolute temperature (310 K at 37 °C), z the ion valence, and F the Faraday constant. At 37 °C the term RT/F × ln(10) evaluates to 61.5 mV, giving the practical logarithmic form: [1]

E(ion) = (61.5 / z) × log10( [ion]outside / [ion]inside ) [1]

Worked examples (typical mammalian myocyte concentrations): [1]

The RMP (minus 70 to minus 90 mV) sits near EK (minus 95 mV) — confirming that K+ is the dominant determinant of the resting potential. Note Ca2+ has the largest driving force (its equilibrium potential is very positive, so it always wants to ENTER the cell whenever a Ca2+ channel opens — the basis of excitation-contraction coupling and neurotransmitter release). [1]

The Goldman-Hodgkin-Katz (Goldman) equation — all permeant ions together

Real membranes are simultaneously permeable to several ions. The Goldman equation (Goldman 1943; Hodgkin and Katz 1949) generalises the Nernst equation into a permeability-weighted average under the constant-field assumption (the electric field across the ~5 nm bilayer is assumed uniform). For monovalent ions: [1]

Vm = (RT/F) × ln( (PK·[K+]o + PNa·[Na+]o + PCl·[Cl−]i) / (PK·[K+]i + PNa·[Na+]i + PCl·[Cl−]o) ) [1]

Note the Cl− terms are inverted (intracellular in the numerator, extracellular in the denominator) because Cl− is an anion — the opposite sign convention. The permeability ratio PK : PNa : PCl at rest is approximately 1 : 0.04 : 0.45 — K+ dominates. Worked example at rest (Vm = 61.5 × log10(11.6 / 190.1) = minus 74.5 mV). During the action-potential upstroke PNa rises ~500-fold and Vm swings towards ENa (+60 mV) — exactly as observed. The equation reduces to the Nernst equation if only one ion is permeant. [1]

Key teaching point the equation encodes: the membrane potential is pulled towards whichever ion currently has the highest relative permeability. At rest, K+ dominates (Vm near EK); during Phase 0, Na+ dominates (Vm swings towards ENa). [1]

Action potential generation

1. Threshold (minus 55 mV — voltage-gated Na channels open).[1]
2. Depolarisation — Na influx (positive feedback — more depolarisation opens more channels).[1]
3. Repolarisation — Na channels inactivate; K channels open; K efflux.[1]
4. Hyperpolarisation — K channels remain open briefly (overshoot the resting).[1]
5. Refractory — absolute (Na channels inactivated); relative (membrane hyperpolarised).[1]

The molecular events underlying each phase

The action potential is generated by sequential opening and closing of voltage-gated ion channels, each with distinct gating states: [1]

• Voltage-gated Na+ channels have three functional states: RESTING (closed, available — the gate is the activation gate; requires a negative RMP to be available), OPEN (activated — both gates open, Na+ rushes in), and INACTIVATED (the ball-and-chain inactivation gate plugs the pore; the channel cannot reopen until the membrane repolarises and the inactivation gate is removed). This inactivation is the molecular basis of the refractory period.
• Depolarisation (the upstroke). At threshold (about minus 55 mV), voltage-gated Na+ channels open → Na+ rushes in down its electrochemical gradient (ENa = +60 mV, so a huge inward driving force) → the inside becomes more positive → more Na+ channels open → a regenerative positive feedback → the upstroke overshoots towards +30 to +40 mV. The dV/dt is set by the number of available Na+ channels — which is why hyperkalaemia (which inactivates Na+ channels by depolarising the RMP) slows the upstroke and widens the QRS.
• Repolarisation. Na+ channels inactivate (inactivation gate closes); delayed-rectifier voltage-gated K+ channels open (slowly, because their activation gate is voltage-sensitive but slow) → K+ efflux carries positive charge out → the membrane repolarises back towards EK.
• Afterhyperpolarisation. K+ channels close slowly, so they remain open briefly past the resting potential → the membrane transiently hyperpolarises towards EK (minus 95 mV) before the K+ leak equilibrium re-establishes minus 70 to minus 90 mV. [1]

Refractory periods — why impulses travel one way

• Absolute refractory period — no stimulus, however large, can generate a second action potential. It corresponds to the period when most Na+ channels are in the INACTIVATED state (the inactivation gate is shut). This is ESSENTIAL — it prevents the action potential from propagating backwards, and it limits the maximum firing frequency of a neuron or muscle cell.
• Relative refractory period — a supranormal (larger-than-usual) stimulus is required. It corresponds to the afterhyperpolarisation (the membrane is closer to EK, further from threshold) PLUS partially recovered Na+ channels. The relative refractory period is the basis of the phenomenon that repeated stimuli are harder to transmit at high frequency (relevant to the decrement in myasthenia gravis). [1]

Propagation of the action potential

Local current flows from the depolarised (active) region into the adjacent resting region → bring it to threshold → it fires → the wave propagates. In myelinated axons, conduction is saltatory — the impulse "jumps" between nodes of Ranvier (where Na+ channels are concentrated), giving high conduction velocity (up to 120 m/s) with low energy cost. Local anaesthetics block propagation by physically plugging the inner pore of voltage-gated Na+ channels from the intracellular side (they must cross the membrane in uncharged form first — which is why inflamed, acidic tissue [pH 6 — the protonated form cannot cross] is relatively resistant to local anaesthesia, "the infected tooth that won't freeze"). [1]

Excitation-contraction coupling — skeletal, cardiac and smooth muscle

Excitation-contraction (E-C) coupling is the process by which an electrical action potential is converted into a mechanical contraction. All three muscle types share the SAME contractile machinery (actin and myosin, regulated by Ca2+) but use RADICALLY different strategies to deliver Ca2+ to the contractile apparatus. [1]

The common contractile mechanism

In ALL muscle types, contraction occurs by the sliding-filament / cross-bridge cycle:

1. Ca2+ binds the regulatory protein (troponin C in striated muscle; calmodulin in smooth muscle).
2. This exposes actin binding sites for the myosin head.
3. The myosin head (carrying ADP + Pi) binds actin → power stroke (Pi release → myosin pivots → pulls the actin filament towards the centre of the sarcomere) → ADP release.
4. A new ATP binds → myosin detaches from actin → ATP hydrolysis → myosin head re-cocks → the cycle repeats as long as Ca2+ is bound and ATP is available.
5. Relaxation requires both Ca2+ removal (back into the SR / out of the cell) AND ATP (to detach the myosin head — in ATP depletion the muscle stiffens in the "rigor" state, the basis of rigor mortis). [1]

[1]

Skeletal vs cardiac vs smooth muscle — the three coupling strategies

[1]

Why skeletal and cardiac muscle differ — the molecular basis

• Skeletal muscle (RyR1 + DHPR mechanical coupling). The DHPR in the T-tubule physically "bumps" RyR1 through a direct foot-process interaction → RyR1 opens → Ca2+ release. Because coupling is mechanical and direct, every action potential releases a near-fixed quantum of Ca2+ (all-or-none within a sarcomere). Skeletal muscle contracts for HOURS in zero-Ca2+ solution — the SR store is self-sufficient.[4]
• Cardiac muscle (RyR2 + L-type Ca2+ channel CICR). Ca2+ ENTRY through the L-type channel is the trigger; without extracellular Ca2+ the heart CANNOT contract. This is why hypocalcaemia and calcium-channel-blocker overdose cause cardiac (but not skeletal-muscle) failure, and why calcium gluconate is the specific antidote.[1]
• Smooth muscle (pharmacomechanical coupling). There are no T-tubules. Ca2+ enters via L-type and receptor-operated channels AND is released from the SR via IP3 receptors (IP3 generated when an agonist binds a Gq-coupled GPCR, e.g. noradrenaline on vascular alpha-1 receptors). Ca2+ binds calmodulin → activates MLCK → phosphorylates the myosin regulatory light chain → cross-bridge cycling. Relaxation requires myosin light-chain PHOSPHATASE (dephosphorylation). This is the target of nitric oxide (→ cGMP → MLCK inhibition → vasodilation) and of beta-2 agonists in bronchial smooth muscle.

[1]

Calcium handling in cardiac muscle — the targets of inotropes

Force of cardiac contraction is GRADED by cytoplasmic Ca2+, which is regulated at four nodes — every one a drug target in ICU: [1]

• Beta-1 adrenergic stimulation (adrenaline, noradrenaline, dobutamine, dopamine): binds beta-1 GPCR → Gs → adenylyl cyclase → cAMP → PKA → phosphorylates (a) L-type Ca2+ channel (more Ca2+ entry, more trigger, more SR release → positive inotropy), (b) phospholamban (relieves inhibition of SERCA2a → faster reuptake → positive lusitropy AND more SR Ca2+ for next beat → positive inotropy), (c) troponin I (faster Ca2+ dissociation → faster relaxation), (d) the funny current (positive chronotropy).[1][2]
• Phosphodiesterase-3 inhibitors (milrinone, enoximone): prevent cAMP breakdown → raise cAMP WITHOUT beta-receptor stimulation → positive inotropy + vasodilation ("inodilator"). Used when beta-receptors are downregulated (chronic heart failure).
• Digoxin: inhibits the Na+/K+ ATPase → intracellular Na+ rises → Na+/Ca2+ exchanger activity falls → intracellular Ca2+ rises → positive inotropy (but with a narrow therapeutic index and toxicity in hypokalaemia).
• Calcium-channel blockers (verapamil, diltiazem): block L-type Ca2+ channels → negative inotropy, negative chronotropy, negative dromotropy.

The neuromuscular junction — quantal release and safety factor

The neuromuscular junction (NMJ) is the archetypal chemical synapse and the target of every neuromuscular blocking agent and anticholinesterase. [1]

[1]

The endplate potential is typically 3–4 times LARGER than needed to reach threshold — the safety factor. This redundancy ensures that virtually every nerve action potential produces a muscle action potential. Clinical consequence: in myasthenia gravis the safety factor falls to about 1× (antibodies reduce receptor number) → at high firing rates the endplate potential no longer reaches threshold → fatigable weakness (the decremental response on repetitive stimulation). In Lambert-Eaton (antibodies against presynaptic Ca2+ channels → fewer quanta released) the weakness IMPROVES with use (incremental response).[5]

Cellular metabolism — glycolysis, Krebs cycle, oxidative phosphorylation

The cell extracts energy from nutrients to regenerate ATP — the universal energy currency. The complete oxidation of one molecule of glucose yields about 30–32 ATP (modern, lower-than-textbook figures account for the proton leak and transport costs) through three linked pathways. [1]

Glycolysis (cytoplasm, anaerobic-capable)

Glycolysis splits one glucose (6 carbons) into two pyruvate (3 carbons), with a net gain of 2 ATP (substrate-level phosphorylation) and 2 NADH. It does NOT require oxygen — which is why it is the SOLE energy source for the red blood cell (which has no mitochondria) and why it can sustain anaerobic metabolism in ischaemic or exercising muscle (pyruvate → lactate, regenerating NAD+ so glycolysis can continue). [1]

• Rate-limiting step: phosphofructokinase-1 (PFK-1), inhibited by ATP and citrate (signals of energy sufficiency), activated by AMP and fructose-2,6-bisphosphate (insulin signal).
• Fates of pyruvate: AEROBIC → enters mitochondria → acetyl-CoA (via pyruvate dehydrogenase) → Krebs cycle; ANAEROBIC → lactate (via lactate dehydrogenase, regenerating NAD+). [1]

The Krebs cycle (mitochondrial matrix, aerobic)

The Krebs (citric acid / tricarboxylic acid, TCA) cycle oxidises acetyl-CoA (2 carbons, from pyruvate, fatty-acid beta-oxidation or ketone bodies) to CO2, yielding per acetyl-CoA: 3 NADH, 1 FADH2, 1 GTP (= ATP), 2 CO2. The cycle does NOT directly use O2, but it REQUIRES O2 indirectly (to re-oxidise NADH and FADH2 via the electron transport chain — without O2 as the final electron acceptor, NADH accumulates and the cycle stops). Rate-limiting enzyme: isocitrate dehydrogenase (activated by ADP, inhibited by ATP and NADH). [1]

Oxidative phosphorylation (inner mitochondrial membrane — the electron transport chain)

Oxidative phosphorylation (OXPHOS) is the major ATP producer: about 26–28 of the 30–32 ATP per glucose come from here. The electron transport chain (ETC) consists of four multisubunit complexes (I, II, III, IV) embedded in the inner mitochondrial membrane, plus two mobile carriers (coenzyme Q / ubiquinone, and cytochrome c). [1]

[1]

The ATP yield — accounting the 30–32 ATP per glucose

[1]

Why this matters in ICU: the brain, heart and kidney are OBLIGATE aerobic organs — they cannot sustain glycolysis alone and die within minutes of losing O2 (hence the cardiac arrest timeline). The red blood cell (no mitochondria) and the renal medulla rely entirely on glycolysis. Sepsis shifts cells towards aerobic glycolysis (the "Warburg-like" phenotype) — pyruvate is preferentially converted to lactate even when O2 is adequate, contributing to septic lactataemia that does NOT simply reflect tissue hypoxia. [1]

Oxygen sensing at the cellular level — HIF-1 alpha and mitochondrial oxygen utilisation

Cells must sense and adapt to changes in oxygen availability. Two molecular systems dominate: the hypoxia-inducible factor (HIF) transcription factor system (genomic, slow, adaptive) and the mitochondrial oxygen utilisation system (acute, metabolic, the actual sink for O2). [1]

HIF-1 alpha — the master oxygen sensor

Hypoxia-inducible factor 1 (HIF-1) is a heterodimeric transcription factor (HIF-1 alpha + HIF-1 beta / ARNT subunits) that is the MASTER REGULATOR of the transcriptional response to hypoxia. Its elegance lies in an oxygen-dependent degradation switch: [1]

• In NORMOXIA: HIF-1 alpha is hydroxylated on specific proline residues by prolyl hydroxylase domain (PHD) enzymes. This hydroxylation REQUIRES molecular O2 (PHDs are 2-oxoglutarate-dependent dioxygenases — they consume O2). Hydroxylated HIF-1 alpha is recognised by the von Hippel-Lindau (VHL) E3 ubiquitin ligase → polyubiquitinated → degraded by the proteasome. HIF-1 alpha is therefore undetectable in normoxia.
• In HYPOXIA: PHD activity falls (no O2 substrate) → HIF-1 alpha is NOT hydroxylated → NOT recognised by VHL → NOT degraded → it ACCUMULATES → translocates to the nucleus → dimerises with HIF-1 beta → binds hypoxia-response elements (HREs) in target genes → upregulates hundreds of adaptive genes. [1]

HIF-1 alpha target genes orchestrate a coordinated adaptation to hypoxia: erythropoietin (increased red cell mass), vascular endothelial growth factor (VEGF) (angiogenesis), glycolytic enzymes (shift to anaerobic ATP production), glucose transporter GLUT-1 (increased glucose uptake), iNOS (vasodilation), and iron metabolism genes (transferrin, ceruloplasmin).[9][10]

Clinical correlate — HIF stabilisers: roxadustat, daprodustat and other PHD inhibitors (used for renal anaemia) deliberately block PHD → stabilise HIF-1 alpha → mimic hypoxia → stimulate erythropoietin and iron mobilisation. Conversely, VHL tumour-suppressor loss (von Hippel-Lindau disease) constitutively stabilises HIF → haemangioblastomas, renal cell carcinoma, polycythaemia. [1]

Mitochondrial oxygen utilisation — the actual O2 sink

The cell's overwhelming consumer of O2 is cytochrome c oxidase (Complex IV) of the electron transport chain, which reduces O2 to water. Tissue O2 partial pressure at the mitochondrion is therefore the relevant variable for cellular energetics. Critical points: [1]

• O2 extraction is flow-limited then diffusion-limited. As blood passes through a capillary, O2 is extracted; the venous O2 reflects what was NOT extracted. In shock, low flow maximises extraction (low SvO2); in sepsis, impaired extraction (cytopathic hypoxia) leaves venous O2 paradoxically HIGH.
• The P50 of cytochrome c oxidase for O2 is extremely low (about 0.1 mmHg) — meaning mitochondria can keep making ATP until tissue PO2 is vanishingly small. Therefore cellular hypoxia (ATP depletion) is a LATE event — it requires profound tissue O2 depletion. This is why a normal arterial PaO2 does NOT guarantee adequate cellular oxygenation in sepsis (cytopathic hypoxia — see below).[13]

Cellular hypoxia — the four classical types

Hypoxia is defined as insufficient O2 delivery to / utilisation by the cell to sustain oxidative phosphorylation. Four mechanisms are classically distinguished — and identifying which type is present changes management entirely: [1]

[1]

Clinical fingerprints that distinguish the types

• Hypoxic hypoxia: LOW PaO2 + low SpO2. Correct with supplemental O2, ventilation, PEEP.
• Stagnant hypoxia: NORMAL PaO2 but LOW mixed venous O2 saturation (SvO2, measured on a pulmonary artery catheter — reflects high extraction as the body tries to compensate). Cold peripheries, oliguria, mottled skin. Correct with FLUIDS, inotropes, vasoactive drugs — restore perfusion, NOT more O2.
• Anaemic hypoxia: NORMAL PaO2 + NORMAL SvO2 early, but LOW total O2 content. The body compensates by raising cardiac output; decompensation occurs below about 70 g/L Hb (faster if acutely lost). Correct with TRANSFUSION.
• Histotoxic hypoxia: NORMAL PaO2 but HIGH SvO2 (the hallmark — cells cannot extract O2, so venous blood remains oxygenated) + severe LACTIC ACIDOSIS + a normal-to-high cardiac output. The classical cause is cyanide (Complex IV poisoning — the antidotes are hydroxocobalamin and sodium thiosiosulphate). The sepsis variant is cytopathic hypoxia (see below). [1]

[1]

Cytopathic hypoxia in sepsis

In sepsis, lactataemia and organ dysfunction can occur DESPITE adequate (even supranormal) O2 delivery, normal PaO2, normal or high Hb, and high SvO2. Fink and colleagues proposed the term cytopathic hypoxia to describe this: the cells cannot UTILISE the oxygen being delivered, because the mitochondria themselves are dysfunctional.[7][13]

Mechanisms of cytopathic hypoxia

1. Inhibition of pyruvate dehydrogenase (PDH). Inflammatory mediators (notably TNF-alpha) and endotoxin inhibit PDH → pyruvate cannot enter the Krebs cycle → it is shunted to lactate instead. This produces lactataemia WITHOUT tissue hypoxia (the "aerobic glycolysis" / Warburg-like phenotype of sepsis). Thiamine is a PDH cofactor — thiamine deficiency (common in critically ill, alcoholics, malnourished) worsens this and is a reversible cause of septic lactataemia.[15]
2. Nitric oxide and peroxynitrite damage to the electron transport chain. Inducible NOS (iNOS) generates high NO in sepsis; NO and its product peroxynitrite (ONOO−, formed from NO + superoxide) inhibit Complexes I and II → reduced electron transport → reduced O2 utilisation. This is partly REVERSIBLE early, becoming irreversible as proteins are nitrated.
3. Poly(ADP-ribose) polymerase (PARP) activation. Oxidative and nitrosative stress damages DNA → PARP is activated to repair it → PARP depletes NAD+ (the substrate for many reactions, and a cofactor whose depletion cripples glycolysis too) → catastrophic energy failure. This is the "suicide hypothesis" of cellular bioenergetic failure.
4. Mitochondrial swelling and permeability transition. Sustained Ca2+ overload and oxidative stress open the mitochondrial permeability transition pore (mPTP) → mitochondrial depolarisation → uncoupling of OXPHOS → ATP synthase runs BACKWARDS, hydrolysing ATP rather than making it → a positive feedback bioenergetic catastrophe (see Cell Death below).

The evidence — Brealey 2002

Brealey and colleagues (2002) demonstrated, in skeletal-muscle biopsies of patients with septic shock, that mitochondrial complex I activity was REDUCED in proportion to the severity of shock, and that reduced complex activity was ASSOCIATED with mortality — even though oxygen delivery was maintained. This was the first direct human evidence that mitochondrial dysfunction (not oxygen delivery) is the rate-limiting step in septic organ failure.[6]

Clinical implications of cytopathic hypoxia

• Resuscitation to supranormal O2 delivery is futile (and harmful). The landmark goal-directed therapy trials (Shoemaker, then the EGDT era — Rivers 2001, then PROCESS, ARISE, ProMISe disproving benefit) assumed the problem was O2 DELIVERY. In cytopathic hypoxia the cells cannot USE the extra O2 — so aggressive O2 delivery targets do not help and may harm (fluid overload, hyperoxia-induced ROS).[8]
• Lactate is a marker, not always of hypoxia. In sepsis, a falling lactate with adequate perfusion suggests improving mitochondrial function; a persistently high lactate may reflect ongoing cytopathic hypoxia OR impaired hepatic clearance OR beta-2-agonist-induced glycolysis (salbutamol) — not necessarily tissue hypoxia.
• Mitochondrial rescue is a therapeutic frontier. Strategies investigated include exogenous cytochrome c (Fink), succinate (Protti — a Complex II substrate that bypasses the damaged Complex I), melatonin (antioxidant), and NAD+ precursors. None yet proven in large trials.[13][14]

Mitochondrial dysfunction in critical illness

Beyond sepsis, mitochondrial dysfunction is the common pathway of organ failure across critical illness — it explains why multiple organs fail SIMULTANEOUSLY in multiorgan failure (rather than the heart failing first, then the kidneys, etc.). [1]

• Post-cardiac-arrest syndrome: the ischaemia-reperfusion injury generates a burst of mitochondrial ROS at reperfusion → opening of the mPTP → mitochondrial uncoupling → secondary energy failure. Targeted temperature management and avoidance of hyperoxia aim to limit this mitochondrial injury.
• Acute kidney injury: the renal tubular cells are mitochondria-rich and exquisitely O2-sensitive. Ischaemia → mitochondrial fragmentation → tubular cell injury. Recovery requires mitochondrial biogenesis (PGC-1alpha driven).
• ICU-acquired weakness: critical illness myopathy and polyneuropathy are associated with mitochondrial dysfunction in skeletal muscle and nerve — contributing to the prolonged weakness and delayed weaning seen in long-stay ICU patients.
• Singer's hypothesis (2004): multiorgan failure is an EVOLUTIONARILY CONSERVED, ADAPTIVE, metabolically-mediated response to overwhelming inflammation — cells DELIBERATELY enter a hibernation-like state (reduced mitochondrial activity, reduced O2 consumption) to weather the storm, and most patients RECOVER organ function if they survive. This re-frames multiorgan failure not as catastrophic damage but as a protective "shut-down" that we should support, not over-stimulate. It explains the otherwise-puzzling near-complete recovery of organ function in survivors.[8]

[1]

Cell death — apoptosis versus necrosis versus necroptosis

Cell death is not a single event — there are distinct, regulated programmes with different morphologies, immunological consequences, and clinical implications. The Nomenclature Committee on Cell Death (2018) codified these.[12]

[1]

The intrinsic (mitochondrial) apoptotic pathway

Cell stress (DNA damage, oxidative stress, growth-factor withdrawal, ER stress) → BH3-only proteins (PUMA, NOXA, BID) activated → they neutralise the anti-apoptotic proteins (Bcl-2, Bcl-xL) → Bax and Bak oligomerise on the outer mitochondrial membrane → mitochondrial outer membrane permeabilisation (MOMP) → cytochrome c (normally resident in the intermembrane space) is released into the cytoplasm → binds APAF-1 → forms the apoptosome → activates caspase-9 → activates executioner caspases (3, 6, 7) → ordered cell dismantling. This is why cytochrome c release is BOTH the apoptotic signal AND (in cytopathic hypoxia) a marker of catastrophic electron-transport disruption. [1]

The mitochondrial permeability transition pore (mPTP) — the catastrophic switch

Sustained mitochondrial Ca2+ overload, oxidative stress, or phosphate accumulation open the mitochondrial permeability transition pore (mPTP) — a non-specific high-conductance channel in the inner membrane. Opening collapses the proton gradient (uncoupling) → mitochondrial SWELLING (water enters osmotically) → outer membrane RUPTURE → cytochrome c release → if ATP is still available, APOPTOSIS; if ATP is depleted (e.g. ischaemia), NECROSIS. This single switch explains why ischaemia-reperfusion produces both apoptotic (penumbra) and necrotic (core) death in the same tissue. Cyclosporin A inhibits mPTP opening (by binding cyclophilin D) — investigated as cardioprotection in reperfusion (the CIRCUS trial was negative, but the concept stands). [1]

Why lymphocyte apoptosis matters in sepsis

Hotchkiss and colleagues showed that in fatal sepsis there is a marked INCREASE in lymphocyte apoptosis (the spleen and lymph nodes are depleted of lymphocytes at autopsy), contributing to the immunoparalysis / immunosuppression that characterises the later phase of sepsis and predisposes to nosocomial infection. This is why sepsis is biphasic — an early hyperinflammatory phase ("cytokine storm") followed by a late immunosuppressive phase — and why strategies to reverse immunosuppression (interleukin-7, PD-1 blockade) have been investigated.[8]

Reactive oxygen species and antioxidant defenses

Reactive oxygen species (ROS) are partially reduced oxygen intermediates that are more reactive than ground-state O2. They are produced NATURALLY (predominantly by mitochondria) as by-products of oxidative phosphorylation, and are also generated in large quantities by inflammatory cells (neutrophil NADPH oxidase — the "respiratory burst") as a microbial killing mechanism. At low levels ROS serve as SIGNALLING molecules; at high levels they damage lipids, proteins and DNA (oxidative stress). [1]

The ROS species — a hierarchy of reactivity

The PRIMARY site of physiological ROS production is Complex III of the electron transport chain, where electrons can "leak" directly onto O2 to form superoxide — Murphy (2009) estimates about 0.1–2% of consumed O2 becomes superoxide even in health.[11]

Antioxidant defenses — the protective systems

Cells maintain a layered antioxidant defense to keep ROS in check: [1]

[1]

Oxidative stress in critical illness

An imbalance between ROS production and antioxidant defense — oxidative stress — is implicated in the pathogenesis of ARDS, sepsis, reperfusion injury, traumatic brain injury, and neurodegeneration. Mechanisms of injury: [1]

• Lipid peroxidation — ROS attack polyunsaturated fatty acids in membranes → loss of membrane integrity → cell lysis. Measured as malondialdehyde (MDA).
• Protein oxidation and nitration — ROS/RNS damage enzyme active sites → loss of function; nitrated proteins accumulate (e.g. nitrated alpha-synuclein in Parkinson's).
• DNA damage — ROS cause base modifications (8-oxo-guanine) and strand breaks → PARP activation → NAD+ depletion → bioenergetic failure (the suicide hypothesis, above).
• Mitochondrial ROS amplification — ROS damage the ETC → MORE electron leak → MORE ROS → a vicious cycle. Damaged mitochondria are the SOURCE as well as the TARGET of ROS. [1]

[1]

Clinical correlations

• Hyperkalaemia — E(K) shifts positive; resting depolarises; the ECG changes (peaked T, wide QRS).[1]
• Hypocalcaemia — the threshold shifts; the neuronal excitability increases.[1]
• Local anaesthetics — Na channel block (prevent depolarisation).[1]
• Channelopathies — the genetic ion channel disorders.[1]

These classical correlations extend to the whole of critical-care pharmacology. Every inotrope (adrenaline, milrinone, digoxin, levosimendan), every antiarrhythmic (Class I Na-channel block, Class III K-channel block, Class IV Ca-channel block), every neuromuscular blocker and reversal agent (suxamethonium, rocuronium, sugammadex, neostigmine), every antiepileptic and anaesthetic acts at one of the channels, pumps, receptors or enzymes described above. Mastering cellular physiology is therefore not optional for the intensivist — it is the single framework that unifies the pharmacology. [1]

[1]

Exam practice — SAQs

[1] [1]

Clinical pearls

[1]

Red flags

[1] [1] [1] [1]

Key trials and evidence

[1] [1] [1] [1] [1] [1] [1] [1] [1] [1] [1] [1]

Synthesis — cross-cutting cellular mechanisms

[1] [1]

References

1. [1]Bers DM Cardiac excitation-contraction coupling Nature, 2002.PMID 11805843
2. [2]MacLennan DH, Kranias EG Phospholamban: a crucial regulator of cardiac contractility Nat Rev Mol Cell Biol, 2003.PMID 12838339
3. [3]DiFrancesco D The role of the funny current in pacemaker activity Circ Res, 2010.PMID 20167941
4. [4]Rios E, Pizarro G Voltage sensor of excitation-contraction coupling in skeletal muscle Physiol Rev, 1991.PMID 2057528
5. [5]Wood SJ, Slater CR Safety factor at the neuromuscular junction Prog Neurobiol, 2001.PMID 11275359
6. [6]Brealey D, Brand M, Hargreaves I, et al Association between mitochondrial dysfunction and severity and outcome of septic shock Lancet, 2002.PMID 12133657
7. [7]Fink MP Cytopathic hypoxia and sepsis: is mitochondrial dysfunction pathophysiologically important or just an epiphenomenon Pediatr Crit Care Med, 2015.PMID 25560289
8. [8]Singer M, De Santis V, Vitale D, Jeffcoate W Multiorgan failure is an adaptive, endocrine-mediated, metabolic response to overwhelming systemic inflammation Lancet, 2004.PMID 15302200
9. [9]Semenza GL HIF-1: mediator of physiological and pathophysiological responses to hypoxia J Appl Physiol (1985), 2000.PMID 10749844
10. [10]Semenza GL O2-regulated gene expression: transcriptional control of cardiorespiratory physiology by HIF-1 J Appl Physiol (1985), 2004.PMID 14766767
11. [11]Murphy MP How mitochondria produce reactive oxygen species Biochem J, 2009.PMID 19061483
12. [12]Galluzzi L, Bravo-San Pedro JM, Vitale I, et al Molecular mechanisms of cell death: recommendations of the Nomenclature Committee on Cell Death 2018 Cell Death Differ, 2018.PMID 29362479
13. [13]Fink MP Administration of exogenous cytochrome c as a novel approach for the treatment of cytopathic hypoxia Crit Care Med, 2007.PMID 17713377
14. [14]Protti A, Carr J, Moran J, et al Succinate recovers mitochondrial oxygen consumption in septic rat skeletal muscle Crit Care Med, 2007.PMID 17855829
15. [15]Ji R, Li D, Liu Q, et al The Warburg Effect Promotes Mitochondrial Injury Regulated by Uncoupling Protein-2 in Septic Acute Kidney Injury Shock, 2021.PMID 32496419
16. [16]Brooks GA, Arevalo JA, Osmond AD, et al The tortuous path of lactate shuttle discovery: From cinders and boards to the lab and ICU J Sport Health Sci, 2020.PMID 32444344