Urethane provides an unparalleled anaesthetic model for natural sleep: Commentary on Mondino et al., 2022

The mechanistic overlaps across pharmacological sedation and anaesthesia are a topic of keen interest in the study of unconsciousness (as reviewed by us and others: Franks & Wisden, 2021; Moody et al., 2021; Ward-Flanagan & Dickson, 2019). Amongst common veterinary anaesthetics, urethane (ethyl carbamate) is unique (cf. Ward-Flanagan et al., 2022) when used at a surgical plane, in that it allows for spontaneous changes in forebrain state that mimic the cycle between non-rapid eye movement sleep (non-REM) (i.e., 'deactivated' electroencephalographic [EEG] patterns) and rapid eye movement sleep (REM) (i.e., 'activated' EEG patterns) during natural sleep and similar associated changes in peripheral physiological signals (Clement et al., 2008; Pagliardini et al., 2012; Whitten et al., 2009). Not only is this of interest to neuroscientists interested in the mechanisms of pharmacologically-induced unconsciousness but also to those interested in the mechanisms and function of the alterations of the forebrain state during natural sleep. In a recent paper, Mondino et al. (2022) assert that urethane anaesthesia is not an accurate model of natural sleep based on their comparative analysis of electrocortical recordings performed in sleeping and then urethane-treated rats. We believe that their conclusion is erroneous because it derives from some modest and questionable differences between sleep and urethane. Moreover, they report differences that have been previously characterised and acknowledged and ignore other evidence showing very strong neurobiological and physiological parallels between urethane anaesthesia and natural sleep. We maintain that urethane anaesthesia is currently the best documented and most complete pharmacological model of the full electrographic and physiological spectrum of natural sleep—other than sleep itself. Although Mondino et al. (2022) confirm that administration of intraperiotoneal (i.p.) urethane (1.2–1.5 g/kg) allows for spontaneous brain state alternations that resemble those previously recorded during natural sleep, their specific approach seems to evaluate whether urethane produces a condition identical to natural sleep. Yet, to our knowledge, no one has ever made this claim. In fact, our original study characterising the sleep-like neurobiological properties of urethane anaesthesia summarised several ways in which it differed from natural sleep, namely: (1) dependence on pharmacological, as opposed to physiological, induction and regulation, (2) inability to rouse, (3) lack of rapid-eye movements during activated forebrain patterns and finally (4) a general slowing of forebrain activity patterns (Clement et al., 2008). As such, we consider the original premise of the Mondino et al. (2022) study to be predicated on an obviously disingenuous straw-man argument. Furthermore, the electrographic differences reported by Mondino et al. (2022) are minor and biassed by their methodological approach. In every case, their spectral analyses that demonstrated differences between urethane and sleep states used a fixed-bandwidth analysis, ignoring the left-ward spectral shift (frequency slowing) apparent under urethane. This is most apparent at the lower end of the frequency spectrum due to the natural relationship of power to the inverse of frequency values (or the 1/f properties) of the EEG, and the authors' use of large-area cortical surface recordings. Furthermore, the authors report as a novel finding a previously acknowledged difference (cf. Clement et al., 2008) that the theta rhythm (in their case, presumably volume-conducted from the underlying hippocampus) had a lower peak frequency during activated REM-like activity under urethane as compared with natural sleep. Lastly, the differences that they report in terms of the proportion of time spent in either state are confounded by a lack of determination of the anaesthetic plane. Nowhere is there a description of how (or even whether) the plane of anaesthesia was evaluated. As we previously showed (Clement et al., 2008; Ward-Flanagan et al., 2022) depth of anaesthesia directly influences the proportion of time spent in either state and the depth of each. Thus, the differences reported, including a lack of change of electromyographic (EMG) tone across brain state alternations, are of limited utility. More pointedly, the limited focus on electrographic measures in Mondino et al. (2022) expressly ignores many other compelling consistencies between urethane anaesthesia and natural sleep. For instance, in urethane anaesthetised rats, central, but not peripheral, muscarinic receptor antagonism abolishes the REM-like state, whereas central muscarinic agonism blocks the NREM-like state. This demonstrates that the REM-like state in urethane is entirely dependent upon cholinergic mechanisms and is equivalent to the pharmacological sensitivity observed for REM in natural sleep (Clement et al., 2008). Furthermore, and also homologous to sleep, alternations in the state under urethane are dependent upon brainstem (but not diencephalic) manipulations that target forebrain-projecting cholinergic nuclei (Clement et al., 2008). Lastly, although slower, the laminar and current source density profile of hippocampal theta under urethane is consistent with that exhibited during the tonic phase of REM during natural sleep (Bland, 1986; Wolansky et al., 2006). Together with the similar periodic time frame of state changes and consistently coupled changes in peripheral physiological measures (including changes in skeletal EMG tone, as well as respiratory and cardiac rate changes), we concluded that this was powerful evidence for the similarity of brain state changes across both urethane and natural sleep (Clement et al., 2008). This interpretation of our data has been supported by several studies that have since replicated these results, demonstrating the reliability of urethane as a model for both the spontaneous alternations of brain state and accompanying changes in physiological parameters that are also observed in natural sleep (Blasiak et al., 2013; Crook & Lovick, 2016; Pagliardini et al., 2012). Whereas we do not dispute that urethane anaesthesia is not identical to sleep, we do disagree with the claim that it "is not an accurate model of sleep" (Mondino et al., 2022). Indeed, the purpose of a biomedical model is to have the greatest possible fidelity to the actual physiological process (in this case, sleep), whereas also alleviating some of the technical or ethical issues that arise in the study of that process. To this point, although the physiological expression of sleep across humans and rats is also not identical (i.e., nocturnal versus diurnal), it does not preclude the rat from being a useful model for the human condition. With respect to urethane, although the minutiae of electrocortical patterns may show differences to sleep, we fear that Mondino et al. (2022) have effectively missed the forest for the trees. Urethane is still the best-documented anaesthetic that, when used at a clinical plane, allows for sleep-like spontaneous forebrain state changes that co-vary with alternations of peripheral physiological measures. In fact, we argue that it is the cardinal differences between urethane and sleep—the inability to rouse animals (and to perform technically difficult or ethically problematic surgical manipulations), and the consistent, stereotyped and predictable sequencing of state alternations—that make it a superlative model for understanding sleep-like changes in the forebrain state. Whereas we welcome further critical studies examining the similarities and differences between urethane and sleep, rejecting urethane as a model for sleep based on the results of Mondino et al. (2022) is highly premature. Rachel Ward-Flanagan: Conceptualization; writing–review and editing. Silvia Pagliardini: Conceptualization; writing–review and editing. Clayton T. Dickson: Conceptualization; writing–original draft; writing–review and editing. This work was funded by NSERC grant 2021-02926 to CTD. The authors have no conflicts of interest to declare. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/ejn.15985.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.