Deep thiopental anesthesia alters glucose homeostasis but not the neurochemical profile in rat cortex

Introduction Barbiturates are widely used as anesthetics and may suppress brain energy metabolism and inhibit both glucose transport at blood-brain barrier (BBB) and cerebral glucose utilization (CMRglc) (1). However, commercial pentobarbital is normally supplied in alcoholic solutions containing, for example, ethanol and propylene glycol (2), which affect glucose transport and CMRglc in rat cortex (3), and possibly cause osmotic opening the BBB (4). Consequently, results from previous studies using those commercially available pentobarbital might reflect mixing effects of this particular cocktail and thus mask the influence of pentobarbital itself, for example, on the glucose transport kinetics (1). Recent in vivo studies using H NMR spectroscopy suggested that both ethanol and propylene glycol exhibit visible peaks (2,5) and thus could negatively affect absolute quantification of metabolites, in particular glucose (Glc). Therefore, instead of pentobarbital for deep anesthesia, we investigated the effect of another barbiturate, thiopental on glucose transport kinetics and neurochemical profile in the rat cortex using H NMR spectroscopy. Materials and Methods Under authorizations of the local veterinary authorities, male Sprague Dawley rats (n=14, 284±45g) were prepared under 2% isoflurane anesthesia in O2 gas and mechanically ventilated (1). Anesthesia was switched to either α-chloralose (40mg/kg initial bolus and followed by 27mg/kg/hr continuous rate infusion) or thiopental (50mg/kg initial bolus and followed by 80-90mg/kg/hr continuous rate infusion, which induced iso-electricity confirmed by electroencephalogram (EEG, data not shown)). Temperature, PaCO2 and pH were maintained in the physiological range (temperature~38 °C, pH~7.4 and PCO2~45mmHg). All MR experiments were performed in an actively shielded horizontal 9.4T. After the automatical adjustment of field inhomogeneities with the resulting water linewidth in the range of 13-17Hz, the neurochemical profile could be obtained from rat cortex in a 6.5×1.5×3.5μl volume (TE=2.8 ms, TR=4 sec, nt = 320) using SPECIAL (6). Spectra were quantified using LCModel (7). To evaluate cortex glucose transport kinetics, steady-state glycemia was maintained at least 20 min before NMR measurement by adjusting continuous rate of 20% (w/v) D-glucose solution in response to the concomitant measured plasma glucose. The glucose transport kinetic parameters were estimated using the reversible MM model (ReMM) (8). The neurochemical profile was compared with the two-way ANOVA followed by Bonferroni’s post-test. Results and Discussion No apparent difference was observed in the spectra (Figure 1A and 1B) or either in the difference of two spectra (Figure 1C). Consequently, nearly identical neurochemical profiles (Figure 1D) were obtained under light α-chloralose and deep thiopental anesthesia. When plotting the steady-state cortex glucose concentrations as function of plasma glucose, elevated glucose levels in cortex were systematically observed under deep thiopental anesthesia (Figure 2). The resulting apparent transport kinetics, Tmax/CMRglc, was increased by 47% in deep thiopental anesthesia (Table 1) when compared to α-chloralose, which is consistent with the previously observed an increase in deep pentobarbital anesthesia (1). In conclusion, with the successful elimination of alcohols or other potential influents such as N2O, deep thiopental anesthesia could affect cortex glucose transport kinetics with nearly no effect on other metabolites.