How general anaesthesia works
How exactly do general anaesthetics switch off consciousness?
Summary
Why it is open
General anaesthesia combines several effects (unconsciousness, amnesia, analgesia and immobility), and memory is blocked at concentrations considerably lower than those needed for unconsciousness and immobility [10]. The drugs act on different molecules, yet their large-scale effects are highly conserved across species [15]; volatile anaesthetics disrupt function in organisms from bacteria and plants to people within a roughly tenfold range of concentrations [11]. Hugh Hemmings and colleagues call anaesthetic-induced unconsciousness "a major unsolved problem, perhaps because of fundamental limitations in establishing the mechanistic basis for consciousness per se" [10].
An explanation has to cross scales. George Mashour separates a "root cause", which starts the causal chain, such as propofol binding to a subunit of the GABA-A receptor, from a "proximate cause" that more directly produces unconsciousness, such as impaired communication between cortical areas [8]. At the molecular end, "rather than closing in on a unitary target, multiple approaches consistently reveal multiple targets" [10].
The end point is also hard to measure. Unresponsiveness is not the same as unconsciousness: with the isolated forearm technique, patients under general anaesthesia can sometimes communicate with hand signals and later deny having been awake, and "retrospective oblivion is no proof of unconsciousness" [6]. Animals cannot report at all, so animal studies rely on behavioural markers [15].
No body sets a bar for a resolution. Hemmings and colleagues write that a clear understanding of anaesthetic-induced unconsciousness requires "an a priori understanding of the mechanisms underlying consciousness itself, the proverbial 'hard problem'" [10]; see the neural basis of consciousness and the hard problem of consciousness.
What is known
Lipids or proteins (1984–1994). For most of the twentieth century lipids were the favoured target, because of the Meyer–Overton relationship: an anaesthetic's potency correlates with its solubility in fat over six orders of magnitude [11]. In 1984 Nicholas Franks and William Lieb showed that a pure soluble protein, firefly luciferase, is half-inhibited at anaesthetic concentrations "essentially identical to those which anaesthetize animals", over a 100,000-fold range of potencies [1]. In 1991 they found that the two mirror-image forms of isoflurane differed about twofold in their effects on sensitive ion channels in mollusc neurons, while disrupting lipid bilayers equally [2]. Their 1994 review concluded that anaesthetics act "directly on proteins rather than on lipids", chiefly by potentiating inhibitory ligand-gated channels [3]. Also in 1994, Donald Koblin and colleagues reported that in rats nine of 14 halogenated compounds were less potent than their lipid solubility predicts, and five had no anaesthetic effect on their own although they dissolved in lipids and reached the nervous system [4].
GABA-A receptors (2003–2020). GABA-A receptors are the most abundant inhibitory ligand-gated channels in the mammalian brain, and low doses of all volatile anaesthetics potentiate them [11]. Rachel Jurd and colleagues made mice with a single amino-acid change (N265M) in the receptor's β3 subunit. In these mice, etomidate and propofol no longer suppressed movement in response to painful stimuli, while the volatile agents enflurane and halothane were only slightly less effective; the authors concluded that volatile anaesthetics act through a broader spectrum of targets [5]. In 2020 Jeong Joo Kim and colleagues used cryo-electron microscopy to image α1β2γ2 GABA-A receptors, reconstituted in a lipid environment, bound to propofol and etomidate; both occupied the two β–α interfaces in the part of the receptor that spans the membrane, sites shared in part by the benzodiazepine diazepam [14].
Many targets (2019–2020). Photolabelling experiments suggest that 10–15% of detectable proteins, over 300, bind anaesthetics [10]. Ketamine works differently: it is an antagonist of NMDA receptors and depresses excitatory transmission [10]. Other proposed targets include potassium channels, mitochondria and the cytoskeleton; in tadpoles, drugs that stabilise microtubules increased resistance to anaesthesia [11]. Mahmud Arif Pavel and colleagues reported in 2020 that chloroform and isoflurane activate the TREK-1 potassium channel by disrupting the localisation of the enzyme phospholipase D2 to lipid rafts in the cell membrane, and that flies lacking the enzyme resist anaesthesia [13].
From molecules to networks (2008–2015). Michael Alkire, Anthony Hudetz and Giulio Tononi noted that the most consistent regional effect at loss of consciousness is reduced thalamic metabolism and blood flow, suggesting a thalamic "switch"; but ketamine increases thalamic metabolism, and thalamic changes may be "a readout of global cortical activity rather than a consciousness switch" [6]. They proposed instead that anaesthetics cause unconsciousness when they block "the brain's ability to integrate information" [6]. In people with intracranial electrodes, Laura Lewis and colleagues found that propofol-induced unconsciousness began within seconds of a slow (below 1 Hz) oscillation; local networks stayed intact, but activity was fragmented in time and space, and the oscillations were out of step across cortex [7]. Mashour's "cognitive unbinding" proposal, which grew out of work on cognitive binding, including binding by synchrony, holds that processing continues under anaesthesia while its synthesis into a conscious representation fails [8]; see the binding problem. He reports that the role of gamma synchrony "has not been robustly or consistently supported" [8]. In 18 volunteers, Simone Sarasso and colleagues found the cortical response to magnetic stimulation low in complexity under propofol and xenon but complex and wake-like under ketamine; after ketamine participants reported long, vivid dreams, after propofol and xenon no experience [9].
Thalamus and integration (2019–2026). Hemmings and colleagues describe functional disconnection of the thalamus from the cortex, and of prefrontal from posterior cortex, under anaesthesia; whether such loss of connectivity is necessary and sufficient for unconsciousness "is an open question" [10]. In macaques anaesthetised with isoflurane or propofol, Michelle Redinbaugh and colleagues found neurons in the thalamus and deep cortical layers most sensitive to the level of consciousness, and 50 Hz stimulation of the central lateral thalamus restored arousal and wake-like neural processing [12]. In 2026 Andrea Luppi and colleagues compared fMRI data from humans, macaques, marmosets and mice under six anaesthetic regimes. A measure of integrated information between pairs of brain regions fell in every species and regime except one (medetomidine–isoflurane in mice), recovered when people woke, and rose again in two propofol-anaesthetised macaques woken by deep-brain stimulation of the central thalamus; regional losses coincided with the distribution of PVALB, a marker of inhibitory interneurons [15].
Main approaches
Specific protein targets. Anaesthetics bind particular sites on a limited set of proteins, above all GABA-A receptors for propofol and etomidate [3, 5, 14]. Strongest evidence: the luciferase and mirror-image results [1, 2], the loss of etomidate and propofol effects in knock-in mice [5], and binding sites resolved by cryo-electron microscopy [14]. Main difficulty: volatile anaesthetics act through a broader spectrum of targets [5], hundreds of proteins bind anaesthetics [10], no single gene mutation fully abolishes volatile anaesthetic responsiveness in worms, flies or mice [11], and ketamine acts on a different receptor [10].
Membrane-mediated action. A revised lipid view holds that inhaled anaesthetics reach some ion channels through changes in the membrane [13]. Strongest evidence: the Meyer–Overton correlation [11] and the phospholipase D2 mechanism [13]. Main difficulty: the protein results [1, 2] and compounds that dissolve in lipids but do not anaesthetise [4]; the evidence for the new mechanism comes from cells and flies [13].
Thalamus and arousal circuits. Anaesthetics cause unconsciousness by acting on the thalamus and on the hypothalamic circuits that regulate sleep and arousal [6, 10]; see also the function of sleep. Strongest evidence: reduced thalamic metabolism at loss of consciousness [6] and reversal of anaesthesia by central thalamic stimulation in macaques [12, 15]. Main difficulty: ketamine increases thalamic metabolism, and thalamic changes may follow cortical ones [6].
Disrupted cortical integration. Whatever the molecular route, unconsciousness follows when cortical areas can no longer integrate information [6, 8]. Strongest evidence: fragmented cortical activity in people under propofol [7], loss of integrated information across four species and its recovery on waking [15], and complexity that tracks reported experience, including under ketamine [9]. Main difficulty: whether disconnection is necessary and sufficient is open [10], gamma-synchrony predictions have not held consistently [8], and Hemmings and colleagues cite evidence that different anaesthetics produce states that are not neurophysiologically identical [10].
Open sub-questions
- One proximate cause? At the moment responsiveness is lost, do propofol, sevoflurane and ketamine converge on the same loss of integration [15], although ketamine leaves cortical complexity high [9]?
- From receptor to network. In β3(N265M) mice, which resist the effects of propofol on movement [5], does propofol still fragment cortical activity [7] or reduce integration [15]?
- Switch or readout. Does the thalamic change at loss of consciousness precede or follow the cortical one [6], and does thalamic stimulation [12] restore more than arousal?
- Experience without response. Can a complexity measure [9] detect disconnected experience during anaesthesia as it happens, given that later reports can miss it [6]?
- Membranes in mammals. Does the phospholipase D2 pathway shown in cells and flies [13] contribute to anaesthesia in mammals?
- Binding under anaesthesia. Is local feature processing preserved while its synthesis across areas fails, as cognitive unbinding predicts [8]? See the binding problem.
Work on this in Papex
If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.
Key references
- Franks, N. P. & Lieb, W. R. (1984). Do general anaesthetics act by competitive binding to specific receptors? Nature 310(5978), 599–601. 10.1038/310599a0 · Checked 14 Oct 2026
- Franks, N. P. & Lieb, W. R. (1991). Stereospecific effects of inhalational general anesthetic optical isomers on nerve ion channels. Science 254(5030), 427–430. 10.1126/science.1925602 · Checked 14 Oct 2026
- Franks, N. P. & Lieb, W. R. (1994). Molecular and cellular mechanisms of general anaesthesia. Nature 367(6464), 607–614. 10.1038/367607a0 · Checked 14 Oct 2026
- Koblin, D. D., Chortkoff, B. S., Laster, M. J., Eger, E. I., Halsey, M. J. & Ionescu, P. (1994). Polyhalogenated and perfluorinated compounds that disobey the Meyer-Overton hypothesis. Anesthesia & Analgesia 79(6), 1043–1048. 10.1213/00000539-199412000-00004 · Checked 14 Oct 2026
- Jurd, R., Arras, M., Lambert, S., Drexler, B., Siegwart, R., Crestani, F. et al. (2003). General anesthetic actions in vivo strongly attenuated by a point mutation in the GABA(A) receptor β3 subunit. The FASEB Journal 17(2), 250–252 (published online 2002). 10.1096/fj.02-0611fje · Checked 14 Oct 2026
- Alkire, M. T., Hudetz, A. G. & Tononi, G. (2008). Consciousness and anesthesia. Science 322(5903), 876–880. 10.1126/science.1149213 · Checked 14 Oct 2026
- Lewis, L. D., Weiner, V. S., Mukamel, E. A., Donoghue, J. A., Eskandar, E. N., Madsen, J. R. et al. (2012). Rapid fragmentation of neuronal networks at the onset of propofol-induced unconsciousness. Proceedings of the National Academy of Sciences 109(49), E3377–E3386. 10.1073/pnas.1210907109 · Checked 14 Oct 2026
- Mashour, G. A. (2013). Cognitive unbinding: a neuroscientific paradigm of general anesthesia and related states of unconsciousness. Neuroscience & Biobehavioral Reviews 37(10), 2751–2759. 10.1016/j.neubiorev.2013.09.009 · Checked 14 Oct 2026
- Sarasso, S., Boly, M., Napolitani, M., Gosseries, O., Charland-Verville, V., Casarotto, S. et al. (2015). Consciousness and complexity during unresponsiveness induced by propofol, xenon, and ketamine. Current Biology 25(23), 3099–3105. 10.1016/j.cub.2015.10.014 · Checked 14 Oct 2026
- Hemmings, H. C., Riegelhaupt, P. M., Kelz, M. B., Solt, K., Eckenhoff, R. G., Orser, B. A. & Goldstein, P. A. (2019). Towards a comprehensive understanding of anesthetic mechanisms of action: a decade of discovery. Trends in Pharmacological Sciences 40(7), 464–481. 10.1016/j.tips.2019.05.001 · Checked 14 Oct 2026
- Kelz, M. B. & Mashour, G. A. (2019). The biology of general anesthesia from paramecium to primate. Current Biology 29(22), R1199–R1210. 10.1016/j.cub.2019.09.071 · Checked 14 Oct 2026
- Redinbaugh, M. J., Phillips, J. M., Kambi, N. A., Mohanta, S., Andryk, S., Dooley, G. L. et al. (2020). Thalamus modulates consciousness via layer-specific control of cortex. Neuron 106(1), 66–75.e12. 10.1016/j.neuron.2020.01.005 · Checked 14 Oct 2026
- Pavel, M. A., Petersen, E. N., Wang, H., Lerner, R. A. & Hansen, S. B. (2020). Studies on the mechanism of general anesthesia. Proceedings of the National Academy of Sciences 117(24), 13757–13766. 10.1073/pnas.2004259117 · Checked 14 Oct 2026
- Kim, J. J., Gharpure, A., Teng, J., Zhuang, Y., Howard, R. J., Zhu, S. et al. (2020). Shared structural mechanisms of general anaesthetics and benzodiazepines. Nature 585(7824), 303–308. 10.1038/s41586-020-2654-5 · Checked 14 Oct 2026
- Luppi, A. I., Uhrig, L., Tasserie, J., Mediano, P. A. M., Rosas, F. E., Singleton, S. P. et al. (2026). Convergent transcriptomic and connectomic controllers of information integration and its anaesthetic breakdown across mammalian brains. Nature Human Behaviour 10(4), 777–802 (published online 28 January 2026). 10.1038/s41562-025-02381-5 · Checked 14 Oct 2026
Source
Open, as listed by Wikipedia · Unsolved problems in neuroscience. Checked 29 Sep 2026.
The status is the source’s. Papex does not decide whether a problem is solved.
Records on this problem
No public record addresses this problem yet.