Volatile anaesthetic toxicity in the genetic mitochondrial disease Leigh syndrome

Abstract

Background: Volatile anaesthetics are widely used in human medicine. Although generally safe, hypersensitivity and toxicity can occur in rare cases, such as in certain genetic disorders. Anaesthesia hypersensitivity is well-documented in a subset of mitochondrial diseases, but whether volatile anaesthetics are toxic in this setting has not been explored.

Methods: We exposed Ndufs4(-/-) mice, a model of Leigh syndrome, to isoflurane (0.2-0.6%), oxygen 100%, or air. Cardiorespiratory function, weight, blood metabolites, and survival were assessed. We exposed post-symptom onset and pre-symptom onset animals and animals treated with the macrophage depleting drug PLX3397/pexidartinib to define the role of overt neuroinflammation in volatile anaesthetic toxicities.

Results: Isoflurane induced hyperlactataemia, weight loss, and mortality in a concentration- and duration-dependent manner from 0.2% to 0.6% compared with carrier gas (O₂ 100%) or mock (air) exposures (lifespan after 30-min exposures ∗P<0.05 for isoflurane 0.4% vs air or vs O₂, ∗∗P<0.005 for isoflurane 0.6% vs air or O₂; 60-min exposures ∗∗P<0.005 for isoflurane 0.2% vs air, ∗P<0.05 for isoflurane 0.2% vs O₂). Isoflurane toxicity was significantly reduced in Ndufs4(-/-) exposed before CNS disease onset, and the macrophage depleting drug pexidartinib attenuated sequelae of isoflurane toxicity (survival ∗∗∗P=0.0008 isoflurane 0.4% vs pexidartinib plus isoflurane 0.4%). Finally, the laboratory animal standard of care of 100% O₂ as a carrier gas contributed significantly to weight loss and reduced survival, but not to metabolic changes, and increased acute mortality.

Conclusions: Isoflurane is toxic in the Ndufs4(-/-) model of Leigh syndrome. Toxic effects are dependent on the status of underlying neurologic disease, largely prevented by the CSF1R inhibitor pexidartinib, and influenced by oxygen concentration in the carrier gas.

Keywords: electron transport chain complex I; mitochondrial disease; neurodegenerative disease; paediatric disease; toxicity.

Fig 1

Brief isoflurane exposures cause respiratory depression, metabolic changes, and weight loss in Ndufs4(−/−) mice. (a) Schematic of mouse anaesthesia exposure chamber with major components indicated, see also Methods. (b) Overview of the course of disease onset in untreated Ndufs4(−/−) animals and the paradigm for testing VA toxicity. Animals are born healthy, beginning to show neurological symptoms around postnatal day 37 (P37). Median and maximum survival in untreated Ndufs4(−/−) mice are ∼P60 and ∼P80, respectively. Animals were exposed to isoflurane (Iso), carrier gas (O₂ 100%), or air in matched conditions, for 30 min at P50, a post-symptom onset age. (c) Righting reflex at 15 and 30 min of exposure. 1=righting reflex is present (animals are unanaesthetised), 0=absent (animals are anaesthetised by this measure). ∗∗P<0.01, ∗∗∗P<0.0005, ∗∗∗∗P<0.00005 compared with oxygen 100% at the matched time, ^††P<0.005 15 vs 30 min of exposure, by Mann–Whitney test. (d) Ventilatory frequency at 15 and 30 min of exposure. ∗∗P<0.005, ∗∗∗P<0.0005, ∗∗∗∗P<0.00005 compared with the O₂ 100% group at the same time by Welch's t-test. ^†P<0.05, ^†††P<0.0005 by pairwise t-test against 15 min timepoint in the same treatment group. (c–d) n≥5 per group, see Supplementary Figure S2 for control animal data. (e) Change in blood lactate concentration in control mice during a 30-min exposure at P50 to Iso 1.25% or 0.4% (equipotent/anaesthetising and equimolar compared with Ndufs4(−/−), respectively), O₂ 100%, or air. One-way analysis of variance (anova) ∗P<0.05. (f) Change in blood lactate concentration in Ndufs4(−/−) mice during a 30-min exposure at P50 to Iso 0.2%, 0.4%, or 0.6%, O₂ 100%, or air. One-way anova ∗∗∗∗P<0.0001. (g) Change in blood glucose concentration in control mice during a 30-min exposure at P50 to Iso 1.25% or 0.4% (equipotent/anaesthetising and equimolar, respectively), O₂ 100%, or air. One-way anova=not significant. (h) Change in blood glucose concentration in Ndufs4(−/−) mice during a 30-min exposure at P50 to Iso 0.2%, 0.4%, or 0.6%, O₂ 100%, or air. One-way anova ∗∗∗∗P<0.0001. (e–h) n≥7 per group. (i) Change in control animal weight over the 24 h after a single 30-min exposure at P50 to Iso, O₂ 100%, or air. One-way anova=not significant. (j) Change in Ndufs4(−/−) weight over the 24 h after a single 30-min exposure at P50 to Iso, O₂ 100%, or air. One-way anova ∗P<0.05. (i–j) n≥5 per group. (c–j) Error bars=standard error of the mean (sem) centred on the mean. All datapoints represent biological replicates (individual animals). Any pairwise comparisons not shown are non-significant (P>0.05). (e–j) P-values shown are Tukey's multiple testing corrected P-values: ∗∗∗∗P<0.0001, ∗∗∗P<0.0005, ∗∗P<0.005, ∗P<0.05. All datapoints shown. VA, volatile anaesthetic.

Fig 2

Repeated brief isoflurane exposures increase anaesthesia sensitivity, cause weight loss, and accelerate mortality in a concentration-dependent manner in Ndufs4(−/−) mice. (a) Overview of Ndufs4(−/−) disease course and the paradigm for assessing the impact of repeated anaesthesia exposures. Animals were exposed to isoflurane, or air or carrier gas (O₂ 100%) in matched conditions, for 30 min once per day on P50, P51, and P52, with exposures at approximately the same time each day (see Methods). (b) Loss of righting reflex by exposure day and time exposed in isoflurane 0.2%-exposed Ndufs4(−/−) animals. ∗∗P<0.005 (exposure number factor) and ∗∗∗∗P<0.00005 (time factor) by Mann–Whitney test (see Methods). (c) Ventilatory frequency by time into exposure and exposure number in isoflurane 0.2%-exposed Ndufs4(−/−) animals. Two-way analysis of variance (anova): time factor ∗∗∗P<0.0005, exposure number factor ∗∗∗∗P<0.0001. P-values shown are Tukey's multiple comparisons test ∗P<0.05, ∗∗P<0.005, ∗∗∗P<0.0005, ∗∗∗∗P<0.0001. P-values for irrelevant comparisons (different duration and day) not shown. See Supplementary Figure S2 for equivalent comparisons of other treatment groups. (d–e) Plot of control (d) and Ndufs4(−/−) (e) weights normalised (by individual animals) to P50, focused on the days spanning the exposures. Plotted are mean with standard error of the mean (sem). Controls—no differences between groups on any day. Ndufs4(−/−)—significance of differences between groups not assessed at individual days; rather, the rate of weight change was assessed in (f). (f) The rate of change in weight (grams per day, g day⁻¹) in data from (e) during the P50–P53 period, calculated using a slope equation with available datapoints (i.e. where animals did not survive to P53, the slope was calculated using available days). One-way anova: ∗∗∗∗P<0.00001. (b–f) Datapoints represent individual animals, with bars at the mean and error bars representing sem. P-values shown are Tukey's multiple testing corrected P-values, ∗P<0.05, ∗∗P<0.005. n≥6 in each group. (g) Survival of Ndufs4(−/−) animals in the P50–P52 30-min exposures. ∗P<0.05 vs O₂ 100%, ^#P<0.05 vs air (mock) treatment, ∗∗P<0.005 vs O₂ 100%, ^##P<0.005 vs air (mock) treatment, Gehan–Breslow–Wilcoxon test. (h) Cause of death in Ndufs4(−/−) mice from these experiments (those animals in survival curves in [d]). No mortality was observed in any control cohort in these experiments. FDIC, found dead in cage; ISO, isoflurane; WCO, euthanised as a result of reaching weight cut-off.

Fig 3

Impact of 60-min isoflurane exposures from P50–P52 in the Ndufs4(−/−). (a) Overview of Ndufs4(−/−) disease course and the paradigm for 60-min exposure to isoflurane (Iso). (b) Change in blood lactate in control mice during a 60-min exposure at P50 to Iso 1.25%, O₂ 100%, or air. One-way ANOVA = not significant. (c) Change in blood lactate in Ndufs4(−/−) mice during a 60-minute exposure at P50 to Iso 0.2%, O₂ 100%, or air. One-way analysis of variance (anova) ∗P<0.05. (d) Change in blood glucose in control mice during a 60-min exposure at P50 to Iso 1.25%, O₂ 100%, or air. One-way anova=not significant. (e) Change in blood glucose in Ndufs4(−/−) mice during a 60-min exposure at P50 to Iso 0.2%, O₂ 100%, or air. One-way anova ∗∗P<0.005. (b–e) n≥7 per group; in (e), one datapoint for the air cohort is outside the plotted range (−7.1 mM; range kept consistent between plots for visual comparison). (f) Change in control animal weight over the 24 h after a single 60-min exposure to Iso 1.25%, O₂ 100%, or air. One-way anova=not significant. (g) Change in Ndufs4(−/−) weight over the 24 h after a single 60-min exposure to Iso 0.2%, O₂ 100%, or air. One-way anova ∗P<0.05. (f–g) n≥7 per group. (b–g) P-values shown are Tukey's multiple testing corrected p-values: ∗∗∗∗P<0.0001, ∗∗∗P<0.0005, ∗∗P<0.005, ∗P<0.05. All datapoints shown. (h) Overview of Ndufs4(−/−) disease course and the paradigm for once daily 60-min exposure to Iso at P50, P51, and P52. (i) Loss of righting reflex by exposure day and time exposed in Iso 0.2% in Ndufs4(−/−) animals. ∗∗P<0.005 (exposure number factor) and ∗∗∗∗P<0.00005 (time factor) by Mann–Whitney test. (j) Ventilatory frequency by time into exposure and exposure number in Iso 0.2%-exposed Ndufs4(−/−) animals. Two-way anova: time factor ∗∗∗P<0.0005, exposure number factor ∗∗∗∗P<0.0001. (i–j) n≥5 per group except 60-min ventilatory frequency on P53, where n=4. P-values shown are Tukey's multiple comparisons test ∗P<0.05, ∗∗P<0.005, ∗∗∗P<0.0005, ∗∗∗∗P<0.0001. P-values for irrelevant comparisons (different duration and day) not shown. (k–l) The rate of change in weight (grams per day, g day⁻¹) in control (k) and Ndufs4(−/−) mice during the P50–P53 period, calculated using a slope equation with available datapoints; where animals did not survive to P53 the slope was calculated using available days. (k) One-way anova—not significant. (l) One-way anova ∗P<0.05. For comparisons shown, ∗P<0.05 by Tukey's multiple comparisons test pairwise P-value. (k–l) n≥6 per group. (m) Ndufs4(−/−) weights normalised (by individual animals) to P50, focused on the days spanning the exposures. Plotted are mean with standard error of the mean (sem). No weight changes occurred in controls (see Supplementary Fig. S4). Weights were not compared on individual days, slopes were compared (k–l). (n) Survival of Ndufs4(−/−) animals exposed to 60 min of Iso 0.2%, O₂ 100%, or air once daily on P50, P51, and P52. ∗∗P<0.005 against air, ^#P<0.05 against O₂ 100%, Gehan–Breslow–Wilcoxon test. n as listed in legend. (o) Cause of death in Ndufs4(−/−) mice from these experiments (those animals in survival curves in [n]). (b–m) Error bars=standard error of the mean (sem) centred on the mean. All datapoints represent biological replicates (individual animals). Any pairwise comparisons not shown are non-significant (P>0.05). FDIC, found dead in cage; ns, nonsignificant; WCO, euthanised as a result of reaching weight cut-off.

Fig 4

Isoflurane (Iso) toxicity in the Ndufs4(−/−) is limited at ages preceding onset of neurologic disease. (a) Overview of the onset of disease in the Ndufs4(−/−) and the relative timing of P30 30-min exposures. (b) Change in blood lactate in control mice during a 30-min exposure at P30 to Iso 0.4% or 1.25%, O₂ 100%, or air. One-way analysis of variance (anova) ∗∗∗P<0.0001. (c) Change in blood lactate in Ndufs4(−/−) mice during a 30-min exposure at P30 to Iso 0.4%, O₂ 100%, or air. One-way anova ∗∗P<0.005. (d) Change in blood glucose in control mice during a 30-min exposure at P30 to Iso 0.4% or 1.25%, O₂ 100%, or air. One-way anova=not significant. (e) Change in blood glucose in Ndufs4(−/−) mice during a 30-min exposure at P30 to Iso 0.4%, O₂ 100%, or air. One-way anova=not significant. (b–e) n≥5 per group. (f) Overview of the onset of disease in the Ndufs4(−/−) and the relative timing of repeat (once daily) P30, P31, and P32 30-min exposures. (g) Ventilatory frequency at 15 and 30 min of exposure by exposure number in Iso 0.4%-exposed cohort. Data compared by two-way anova: time factor and exposure number factor were both non-significant. No Tukey's multiple comparisons test pairwise comparisons (all possible combinations tested) reached significance (Tukey adjusted P<0.05). See Figure 1c for comparison with Iso 0.2% at P50. (h) Ndufs4(−/−) weights normalised (by individual animals) to P30, focused on the days spanning the exposures. Plotted are mean with standard error of the mean (sem). Weights were not compared on individual days, slopes were compared in (i). We observed no weight changes in controls. (i) The rate of change in weight (grams per day, g day⁻¹) in Ndufs4(−/−) mice during the P30–P33 period, calculated using a slope equation. One-way anova ∗P<0.05. Pairwise comparison ∗P<0.05 by Tukey's multiple comparisons test. (j) Overall weight plots for Ndufs4(−/−) mice exposed to Iso, O₂ 100%, or air on P30, P31, and P32. Weight changes were transient, with no overall impact on weight gain or subsequent loss during normal disease progression starting around P37. Plotted are mean and standard error of the mean for each day, lines are LOWESS (locally weighted scatterplot smoothing) running averages. (k) Maximum weight of Ndufs4(−/−) mice was not impacted by exposure to Iso or O₂ 100%, compared with air, on P30, P31, and P32. One-way anova—not significant. No pairwise comparisons significant by Tukey's multiple comparisons test. (l) Survival of Ndufs4(−/−) animals exposed to Iso 0.4%, O₂ 100%, or air 30 min per day from P30 to P32. No significant differences in survival were detected by Gehan–Breslow–Wilcoxon test, and survival curves overlap. (m) Cause of death in Ndufs4(−/−) mice from these experiments (those animals in survival curves in (l). (g–m) n on each Ndufs4(−/−) dataset are reflected in the survival curve in (l). FDIC, found dead in cage; WCO, euthanised as a result of reaching weight cut-off.

Fig 5

Pexidartinib treatment prevents or attenuates toxic effects of isoflurane (Iso) in the Ndufs4(−/−). (a) Overview of the onset of disease in the Ndufs4(−/−) and the relative timing of pexidartinib treatment and P50 exposures. (b) Changes in blood lactate in pexidartinib treated and control diet-treated Ndufs4(−/−) mice during a 30-min exposure at P50 to Iso 0.4%, O₂ 100%, or air. One-way analysis of variance (anova) ∗∗∗∗P<0.0001. Tukey's multiple testing corrected pairwise comparison P-values ∗∗P<0.005, ∗∗∗P<0.0005. (c) Changes in blood glucose in pexidartinib-treated and control diet-treated Ndufs4(−/−) mice during a 30-min exposure at P50 to Iso 0.4%, O₂ 100%, or air. One-way anova ∗∗∗∗P<0.0001. Tukey's multiple testing corrected pairwise comparison P-values ∗∗∗∗P<0.0001. (d) Change in pexidartinib-treated and control diet-treated Ndufs4(−/−) weight over the 24 h after a single 30-min exposure at P50 to Iso 0.4%, O₂ 100%, or air. One-way anova ∗∗∗P<0.0005. Tukey's multiple testing corrected pairwise comparison P-values ∗P<0.05, ∗∗P<0.005. (b–d) n≥6 per group. (e) Overview of the onset of disease in the Ndufs4(−/−) and the relative timing of pexidartinib treatment and daily Iso 0.4%, O₂ 100%, or air, exposures on P50, P51, and P52. (f) Ventilatory frequency at 15 and 30 min of exposure to Iso 0.4% in pexidartinib-treated or control diet-treated Ndufs4(−/−) mice. Two-way anova: exposure number factor ∗∗∗∗P<0.0001, treatment factor ∗∗∗∗P<0.0001. Pairwise comparisons by Tukey's multiple comparison test: ∗∗P<0.005, ∗∗∗P<0.0005, ∗∗∗∗P<0.00005. (g) Righting reflex presence in Iso 0.4% exposures in pexidartinib-treated or control diet-treated Ndufs4(−/−) mice. 1=righting reflex present (animals un-anaesthetised), 0=righting reflex absent (animals are anaesthetised to the point of loss of righting reflex). ∗P<0.05 by paired non-parametric Mann–Whitney test. (h) Weights of pexidartinib-treated and control diet-treated Ndufs4(−/−) animals exposed to Iso 0.4%, O₂ 100%, or air on P50, P51, and P52, normalised to P50. Data are mean with standard error of the mean (sem). (i) The rate of change in weight (grams per day, g day⁻¹) during the P50–P53 period. Datapoints represent individual animals, with bars at the mean and error bars representing sem. One-way anova ∗∗∗P<0.001 by. Tukey's multiple testing corrected pairwise comparison P-values ∗P<0.05, ∗∗P<0.005, ∗∗∗P<0.0001. (j) Raw weight plots of pexidartinib-treated Ndufs4(−/−) mice exposed to O₂ 100% or Iso 0.4% (in O₂ 100%) throughout the P50–P52 exposure period. Data are mean with error bars showing sem, lines are LOWESS (locally weighted scatterplot smoothing) running averages. (k) Survival of untreated and pexidartinib-treated Ndufs4(−/−) animals exposed to 30 min of Iso 0.4% or O₂ 100% at P50–P52. ∗P<0.05 and ∗∗P=0.002 vs untreated Ndufs4(−/−) mice exposed to Iso 0.4% or O₂ 100%, respectively, by Gehan–Breslow–Wilcoxon test, with pexidartinib-treated mouse cohort halted when pexidartinib reached P60–P80. No pexidartinib-treated animals showed signs of disease at study end. (a–k) Pexidartinib group n ≥ those indicated in panel (k). Control diet-treated groups as detailed in Fig 1, Fig 2. All datapoints represent individual animals, all error bars represent sem. KO, knockout; Pex., pexidartinib.

Fig 6

Carrier gas oxygen concentration impacts the toxicity of isoflurane (Iso) 0.4% in Ndufs4(−/−) mice. (a) Changes in blood lactate in Ndufs4(−/−) mice during a 30-min exposure at P50 to Iso 0.4% in O₂ 100% or air, and in O₂ 100% or air alone. One-way analysis of variance (anova) ∗∗∗∗P<0.0001. Tukey's multiple testing corrected pairwise comparison P-values ∗∗∗P<0.0005, ∗∗∗∗P<0.0001. (a) Changes in blood glucose in Ndufs4(−/−) mice during a 30-min exposure at P50 to Iso 0.4% in O₂ 100% or air, and in O₂ 100% or air alone. One-way anova ∗∗∗∗P<0.0001. Tukey's multiple testing corrected pairwise comparison P-values ∗P<0.05, ∗∗P<0.005, ∗∗∗∗P<0.0001. (c) Change in weight over the 24 h after a single 30-min exposure at P50 to Iso 0.4% in O₂ 100% or air or O₂ or air alone. One-way anova ∗P<0.05; Tukey's multiple testing corrected pairwise comparison P-values ∗∗P<0.005. (d) Average weights, normalised to P50 by animal, of Ndufs4(−/−) mice exposed Iso 0.4% in O₂ 100% or air, or O₂ 100% or air alone, once daily from P50 to P52. Individual days not compared; rates of weight change compared in (e). (e) Rate of weight change from P50 to P53 in Ndufs4(−/−) animals exposed to Iso 0.4% in O₂ 100% or air or O₂ or air alone. One-way anova ∗∗P<0.005; Tukey's multiple testing corrected pairwise comparison P-values ∗P<0.05, ∗∗P<0.005. (f) Survival of Ndufs4(−/−) animals exposed to 30 min of Iso 0.4% in O₂ 100% or air, or O₂ 100% or air alone, once daily from P50 to P52. ∗P<0.05 vs air and ^#P<0.05 vs O₂ 100% by Gehan–Breslow–Wilcoxon test. Early mortality in Iso 0.4% in air was significant compared with historic data on unexposed animals during these ages (Supplementary Fig. S12). (g) Cause of death from mice in (f). For all groups and panels, n ≥7. Cohorts other than Iso 0.4% in air appear in Fig 1, Fig 2. (h) Summary of findings in this study. Isoflurane exposure results in concentration- and duration-dependent toxicities in the Ndufs4(−/−) model of Leigh syndrome including hyperlactataemia, hyperglycaemia, weight loss, and accelerated mortality. O₂ 100% resulted in acute mortality in 60-min exposures, but in most outcomes the carrier gas alone was benign. Notably, isoflurane 0.2% borders on providing anaesthesia in the first exposure (Fig. 1), but increased sensitivity to anaesthesia was observed in repeat exposures. Few toxic sequelae were present in Iso 0.4%-exposed Ndufs4(−/−) mice when exposed at pre-CNS disease onset ages, or in animals treated with the macrophage/microglia-depleting drug pexidartinib. Absent—mortality absent for the duration of the experiment, survival is extended compared with untreated Ndufs4(−/−). (i–j) Diagrammatic models of isoflurane toxicities at pre-CNS disease onset ages (i) and post-disease onset ages (j). (i) At all ages, Ndufs4(−/−) mice show hypersensitivity to anaesthesia and elevation of lactate. (j) After disease onset, additional toxic sequelae arise: hyperglycaemia, weight loss, accelerated mortality, and increased sensitivity in subsequent exposures. Pexidartinib is known to prevent CNS lesions in this model, and prevents Iso toxicity. However, hyperlactataemia and volatile anaesthetic hypersensitivity occur even in the absence of overt disease but are attenuated by pexidartinib, suggesting a role for immune cells even in the absence of overt CNS degeneration. ^‡Weight gain was prevented during the exposure period, but no weight loss occurred. ETC CI, electron transport chain complex I; FDIC, found dead in cage; KO, knockout; n.d., not detected (no statistically significant effect); n/a, not applicable (i.e. these animals were either anaesthetised even on the first exposure, or were oxygen treated so never anaesthetised); ns, nonsignificant; VA, volatile anaesthetic; WCO, euthanised as a result of reaching weight cut-off.

Figure S1

Figure S2

Figure S3

Figure S4

Figure S5

Figure S6

Figure S7

Figure S8

Figure S9

Figure S10

Figure S11

Figure S12

Figure S13