Interaction of Isoflurane with the Dopamine Transporter

The Emory University Institutional Animal Care and Use Committee approved all animal studies. Rhesus experiments used five female monkeys ranging in weight from 4 to 7 kg. These animals are housed in the department of animal resources in a cohort of approximately 10 animals. When studied, they were initially hypnotized with an intramuscular injection of ketamine, 8–10 mg/kg, and atropine, 0.01 mg/kg. The ketamine was administered at least 2 h before beginning the first PET scan. Shortly after, the animal was brought into the PET scanner suite, an endotracheal tube was placed, and the animal was mechanically ventilated with oxygen and 1% isoflurane (end-tidal concentration). For half of the experiments, the anesthetic concentration was increased to 2% after the first scan and before the second scan. These two levels were chosen because 1% is approximately the minimum concentration that can be used to keep the animal motionless during the 90-min scanning session, and 2% is near the maximum that can safely be used without administering other, confounding drugs. A peripheral intravenous catheter was inserted for injection of the radiotracer. We also measured core body temperature, electrocardiogram, and end-tidal respiratory gas concentrations (Datex-Ohmeda, Madison, WI). Rodent studies used male Sprague-Dawley rats (weight, 250–300 g). Handling of the rats was specific to the different experiments and is described below.

[¹⁸F]FECNT was prepared by a two-step reaction sequence as previously described. 14Briefly, alkylation of 1-[¹⁸F]fluoro-2-tosyloxyethane with 2β-carbomethoxy-3β-(4-chlorophenyl)nortropane in dimethyl formamide at 135°C for 45 min affords [¹⁸F]FECNT, which is purified by semi-prep, reverse-phase high-performance liquid chromatography (HPLC) to produce a product free from the precursor, 2β-carbomethoxy-3β-(4-chlorophenyl)nortropane with specific activity 56 MBq/nmole (1.5 Ci/μmole). [¹⁸F]FECNT is at or near quasiequilibrium 30 min after injection and is rapidly displaced by β-CIT. 15 

The pet studies used a within-animal experimental design that allows each animal to act as his or her own control. The animals were initially imaged, and then the anesthetic conditions were changed (or not), and the animal was rescanned. Efforts were taken to ensure that the animal was stable before the second scan, but otherwise the time between scans was minimized. [¹⁸F]FECNT binding potential was determined from the pet data. Studies where the conditions were not changed acted as the null control. The parameter of interest was the fractional change in the binding potential caused by increasing the anesthetic concentration.

All studies were collected on a Siemens 951 (Hoffman Estates, IL) scanner. After positioning the monkey in the scanner, the head was immobilized using a thermoplastic (Tru Scan, Annapolis, MD) face mask. Transmission scanning was performed with ⁶⁸Ge ring sources. [¹⁸F]FECNT was injected as a slow bolus for 5 min with an infusion pump. Scanning started coincident with the start of injection. The acquisition sequence was 28 frames long, starting with 30-s scans and ending with 20-min scans for a total duration of 90 min. (The exact sequence of number of frames, time per frame in seconds was 10, 30; 5, 60; 5, 180; 5, 600; 1, 900.) At the end of the first scanning sequence, the isoflurane concentration was either left the same (to determine the test–retest variability) or increased to 2%. After a 1-h delay, PET images were again acquired with the same sequence except it included a 10-min scan immediately before the [¹⁸F]FECNT injection to measure background activity remaining from the first injection.

Images from the 951 scanner were reconstructed with measured attenuation correction, zoom factor 8, and Shepp-Logan reconstruction filler cutoff at 1 cycle/cm. This produced images with in-plane pixel size of 1.17 mm, 8 mm isotropic resolution, and axial slice thickness of 3.375 mm. At the early and late time points, the cerebellum and striatum are easily discerned in the PET images. Regions of interest were manually drawn over appropriate images and then overlaid on all images to obtain the time–activity curves (TACs) used in kinetic modeling.

The reference tissue method 16allows kinetic modeling to be performed without collecting arterial blood samples. Parameters that govern the uptake of radiotracer in the putamen, such as the binding potential, can be determined using data from a region devoid of specific uptake. Usually, there is no radioactivity in the subject before injection of the radiotracer. However, in this application, the analysis of data from the second scan must account for radioactivity remaining from the first injection. A generalization of the method, 13which yields binding potential estimates that are independent of initial radioactivity, was used without modification as has been done before. 17 

The 10-min scan before the second [¹⁸F]FECNT injection was used to determine residual radioactivity. The generalized reference tissue method describes the putamen uptake curve with four parameters and the TAC from the cerebellum. The parameters are R, the ratio of the rates that [¹⁸F]FECNT crosses the blood–brain barrier in the putamen and cerebellum (flow times extraction in the putamen divided by the flow times extraction in the cerebellum); k², the rate that the tracer leaves the extracellular compartment for the plasma (equal to flow divided by the distribution volume of nonspecifically bound tracer 18); k³, k^onB^maxbecause this is a tracer experiment;19and k⁴, the dissociation rate of radiotracer from the transporter. Here we have made the same assumption as Lammertsma, 16that the distribution volume of nonspecifically bound tracer is the same in the putamen and cerebellum. The binding potential is identified as the ratio k³/k⁴, 19which is equal to B^max/k^D. The kinetic analysis routine adjusts the four parameters until the calculated putamen curve matches the measured data as closely as possible. This calculated curve and the associated parameters are termed the best-fit values for the data and are presented in the figures and tables below.

Fifteen rats were studied, with 7 animals used as controls. The animals were placed in a clear plastic box that contained an evaporative source of isoflurane, oxygen, and air inlets, and an exhaust port. The concentration of isoflurane was adjusted to 2% by changing the volume of air flowing through the box. The concentrations of the gases in the box were monitored using a gas concentration analyzer (Datex-Omeda, Madison, WI). The animals were maintained under these conditions for 30 min, after which they were rapidly decapitated, and the striatum was harvested. The control animals were placed in an identical setup except they never received isoflurane. Because we hypothesized that anesthetics affect DAT, the control animals were decapitated while awake and breathing room air.

Brain tissue sections of the striatum were prepared and processed using immunohistochemical methods previously described. 20Brain slices were prepared with a freezing microtome and then rinsed in 0.1 m phosphate buffered solution (PBS), 5 × 5 min at 4°C, and TBS, 3 × 5 min. Avidin binding sites in the tissue were preblocked by incubation with avidin (10 μg/ml), normal goat serum (NGS; 4%), and triton-X (0.1%) in TBS for 45 min at 4°C followed by three rinses in TBS. The primary antibody (1° Ab) was added to a buffer of 50 μg/ml biotin and 2% NGS in TBS and incubated 48 h at 4°C. After rinsing in TBS four times, the secondary antibody (biotinylated goat antirabbit or mouse immunoglobulin [Ig]) was incubated for 1 h at 4°C. After additional rinses, the sections were developed using diaminobenzidine as described previously. 20Sections were then slide-mounted and cover-slipped for microscopic analysis.

Eleven rats were studied, with 5 animals used as controls. The animals were placed in the same clear plastic box used for the immunohistochemistry experiments. The concentration of isoflurane in the box was maintained at 2%. The animals were maintained under these conditions for 30 min, after which they were rapidly decapitated, and the striatum was harvested. The control animals were treated the same except they never received isoflurane and were decapitated while awake and breathing room air.

After the brains were harvested, tissue was homogenized in the presence of protease inhibitors as previously described. 21Ten micrograms of protein from each sample was prepared in Laemmli sample buffer containing 2% sodium dodecyl sulfate, separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 12% acrylamide gels, and electrophoretically transferred to polyvinylidene difluoride membranes. Membranes were probed with rat anti-DAT monoclonal antibody (N-terminus epitope; Chemicon, 1:500) and rabbit anti–mitogen-activated protein kinase (MAPK; 1:5,000). The MAPK42 immunoreactivity was used as an internal loading control. Blots were then washed and incubated with a horseradish peroxidase–conjugated goat antirabbit secondary antibody (1:10,000) for 1 h at room temperature. Blots were visualized with enhanced chemiluminescence (ECL or ECL+).

Western blot analyses were quantified using ECL+ chemifluorescence and Image Quant software (Kodak Scientific Imaging Systems, Eastman Kodak Company, Hemel Hempstead, UK) as described previously. 20Band intensities were calculated, and background was subtracted using local averages. Band intensities were converted to microgram equivalents using a standard curve consisting of known amounts of protein loaded on each gel. Values were normalized for variation in protein loading based on levels of MAPK immunoreactivity. All conditions were analyzed in triplicate, and experiments were each repeated at least twice.

Human DAT-expressing HEK (DAT-HEK) cells stably expressing human DAT cDNA 22were grown on chamber slides (5 × 10⁴cells per slide). Cells were treated either with 4–6% isoflurane or mock treated for 1 h in a 37°C incubator. This concentration was chosen from preliminary studies. Greater concentrations kill the cells, whereas lesser concentrations minimize the effect. The concentration of isoflurane, O², and CO²were monitored with a gas analyzer (Datex-Omeda, Madison, WI). The cells were quickly washed with ice-cold PBS and then fixed with 3% paraformaldehyde for 10 min on ice. They were then washed with PBST (PBS + 0.1% Tween). DAT were labeled by incubation with a 1:2,000 dilution of rat monoclonal anti-DAT antibody for 10 h at 4°C. The specificity and characterization of the antibody have been described. 20The cells were then washed with PBST and incubated with a 1:100 dilution of fluorescein isothiocyanate (FITC)-labeled goat antimouse IgG for 4 h at 4°C. After a final wash with PBST, the slides were mounted with cover slips and subjected to examination with immunofluorescent confocal microscopy (Zeiss LSM510, Thornwood, NY). The experiment was performed three times.

All PET data results were expressed as the percent change between the first and second scans. Standard tests were performed to verify normalcy and equal variance of the data. Then, statistically significant differences in the percentage change in binding potential between the test–retest group and the low–high isoflurane treated group was determined by unpaired t  tests. A difference was considered significant when P < 0.05.

Three blinded reviewers were asked to rank the slides in order of DAT staining intensity. After this exercise, it was clear that none of the reviewers could distinguish between the different treatments, and no further statistical analyses were performed.

Normalized band intensities from separate blots of the same experiment were averaged. Two-factor (treatment condition and replicate) analysis of variance was used to determine if the treatment had a significant (P < 0.05) effect on the striatal samples between the isoflurane-treated and nontreated rats.

To determine if isoflurane had an effect on cell surface distribution of transporter protein, 39 0.9-μm thick images through the center of randomly chosen cells from both conditions in each of the experiments (19 control and 20 isoflurane-treated cell images) were shown to four blinded reviewers. The reviewers were asked to sort the images into two groups based on the distribution of DAT. They were specifically told not to pay attention to differences in DAT staining intensity. They were not told how many cells were in each group. To address the primary research question, we compared the proportion of control images and the proportion of isoflurane-treated images that were identified as having DAT immunoreactivity concentrated in the cytoplasm. For each reviewer, we estimated the difference between these proportions and computed a 95% confidence interval to determine whether the difference was statistically significant. The power of each test was ≥ 0.82, assuming that the true proportions of cells in the control and isoflurane populations that have DAT immunoreactivity concentrated in the cytoplasm are 0.15 and 0.65, respectively. 23 

Table 1shows hemodynamic parameters for the rhesus monkeys during the 10 scanning sessions. Figure 1shows PET images from a typical animal during 1% and 2% isoflurane anesthesia 82.5 min after [¹⁸F]FECNT injection. During light anesthesia, the uptake in the putamen is approximately sixfold greater than in the cerebellum. Uptake in the rest of the brain is similar to that in the cerebellum. During deep anesthesia, uptake in the putamen is reduced to only a factor of three greater than the cerebellum, whereas the cerebellum and remainder of the brain are relatively unchanged.

Ten experiments were performed on 5 animals to determine the sensitivity and variation in the two [¹⁸F]FECNT injection protocol. The injections were separated by only 2.5 h, so residual activity from the first injection or the duration of anesthesia might confound the second acquisition. To account for these possibilities, test–retest experiments were performed. Uptake in the putamen and cerebellum is shown in figure 2as a function of time for a typical test–retest experiment. Note the similar shape of the curves except the retest curve is displaced upward as a result of activity remaining from the first injection. The lines under the putamen data are reference tissue model fits using the shown corresponding cerebellum as the input function. The binding potential is 7.7% lower for the retest experiment. This particular experiment shows the greatest decrease between test and retest. The greatest increase was 6.2%. When all test–retest experiments are combined, the retest binding potential is decreased by 0.7 ± 2.5% (SEM).

To determine if isoflurane affects [¹⁸F]FECNT binding, the dual injection experiments were repeated except the anesthetic concentration was increased to 2% before the second injection, and figure 3shows a typical set of TACs. In all experiments, the uptake was greater and the washout faster for the high anesthetic level acquisition. Note that at quasiequilibrium (65–85 min;fig. 3) the putamen-to-cerebellum ratio for the low anesthetic state (solid symbols) is much greater than the high anesthetic state (open symbols). On average, the binding potential is decreased by 63 ± 6% (SEM) during high anesthesia.

Increasing the anesthetic level dramatically decreases the binding potential compared with the control retest experiment. Across all animals, 63% of the unbound transporters in the low anesthetic state are either occupied or altered in a way that prevents [¹⁸F]FECNT binding in the high anesthetic state. The difference is significant at P < 0.001.

The reduction in [¹⁸F]FECNT binding potential seen in the rhesus studies during isoflurane exposure could be explained by the anesthetic causing acute degradation of the DAT. To determine if this happens, immunohistochemistry experiments were performed to measure the amount of DAT protein in the striatum of control and isoflurane-treated rats. Figure 4shows a typical section through the striatum of a control rat. Staining for DAT in rat striatum under standard conditions was of moderate to high intensity. There was no qualitative difference observed in the density of striatal staining between the isoflurane and control animals. The resolution of these images was not sufficient to determine if the distribution of DAT changed as a result of isoflurane treatment.

To determine quantitatively if isoflurane degrades DAT, immunoblot studies were performed on rat brain samples. The rats were prepared using the same protocol as in the immunohistochemical staining experiments. Western blot analyses using monoclonal antibody DAT revealed no differences in DAT expression between the control and isoflurane-treated rat brain samples. An immunoreactive band at approximately 85 kd was observed at high levels in the brains of control and treated animals (fig. 5). No other bands were detected in any of the lanes. After normalizing to MAPK42 protein, the mean value for total DAT expression across all experiments was control = 7.5 ± 1.6 and isoflurane = 7.8 ± 2.6 (not significantly different).

Based on the rodent studies, it does not appear that isoflurane causes a degradation of DAT that could explain the rhesus PET results. A change in DAT cellular distribution could explain the reduction in binding if the affinity of [¹⁸F]FECNT for the DAT also changes with the distribution. To determine if isoflurane causes trafficking of DAT, immunocytochemistry experiments were performed on HEK cells stably transfected with human DAT. DAT distribution was compared in control cells and in cells that were incubated in isoflurane for 60 min using fluorescent antibody labeling and confocal microscopy. Examples of cells seen with the confocal microscope are shown in figure 6.

The DAT immunoreactivity was concentrated in the cytoplasm in cells treated with isoflurane. Aggregation of immunoreactivity created a distinct band within the isoflurane-treated cells, as well as a distinct demarcation of the cell nucleus. Also, there were puncta of very focal and intense immunoreactivity in the isoflurane-treated cells but not in the control cells. For each reviewer, there was a statistically significant difference between the proportion of control cells and the proportion of isoflurane-treated cells that had DAT highly concentrated in the cytoplasm. Table 2provides estimated differences between the proportions of isoflurane-treated cells and the proportions of control cells with DAT immunoreactivity concentrated in the cytoplasm. These differences ranged from 0.64 to 0.79 (a difference of 0 corresponds to no effect). All of the 95% confidence intervals excluded 0, providing strong statistical evidence that isoflurane exposure had a significant effect on the distribution of DAT in the cells (α= 0.05). When questioned, all four reviewers said the distinguishing characteristic between the groups was the presence or absence of puncta of increased immunoreactivity.

Volatile anesthetics are known to alter normal dopamine system function. Isoflurane inhibits the uptake of dopamine in COS cells heterologously expressing the DAT transporter. 24Microdialysis studies in rats have shown isoflurane increases extracellular dopamine concentration. Isoflurane enhances the inhibitory effect of cocaine on the dopamine transporter in rhesus monkeys. 25To determine if isoflurane alters DAT binding potential in nonhuman primate brain in vivo , we used PET radioligand imaging studies with the highly specific ligand [¹⁸F]FECNT in rhesus monkeys. We found that exposure to a high concentration of isoflurane causes more than a factor 2 reduction in the binding potential. Possible reasons for the decrease are isoflurane competitively inhibits [¹⁸F]FECNT binding to the DAT, degrades the DAT, changes the conformation of the DAT (and hence its affinity for [¹⁸F]FECNT), or causes internal trafficking of the DAT, which changes to a low affinity conformation in the interior environment of the cell. In the latter possibility, synaptic dopamine would increase because of decreased turnover resulting from fewer transporters available. In this case, the change in binding potential may be further reduced because of competition with endogenous dopamine. To determine which (if any) of these potential mechanisms are in action, experiments were performed on rodents and transfected cells.

Increasing the isoflurane concentration from 1 to 2% substantially reduced the binding potential of [¹⁸F]FECNT in rhesus monkey striatum. There are several other reports of acutely altering the binding potential of a DAT ligand by manipulating the anesthetic conditions. The anesthetic agent ketamine increases the binding potential of [¹¹C]β-CFT for the DAT. 26[¹¹C]β-CFT and [¹⁸F]FECNT are tropane analogs and are expected to measure similar properties of the DAT. The underlying mechanism for the change in binding potential induced by these anesthetic agents is likely to be similar and vital for understanding at least some of their physiologic effects.

Tsukada et al.  25have studied the effect of isoflurane in combination with other agents on the binding potential of [¹¹C]β-CFT. The interaction between isoflurane and DAT was not the primary purpose of their work, but from one graph in their article (fig. 4a), it appears that 1.5% isoflurane mixed with 70% nitrous oxide (N²O) increased the binding potential of [¹¹C]β-CFT by 50% compared with the awake condition. Tsukada et al.  compared awake and 1.5% isoflurane + 70% N²O on separate days, whereas we compared 1% and 2% isoflurane with 100% O²on the same day. These are the only differences noted between our PET studies.

It is not known if N²O or O²administration alters the amount of expressed DAT, but N²O increases the amount of dopamine measured by microdialysis in rats 27and causes increased metabolic breakdown of dopamine. 28Other evidence that indicates N²O may affect the dopaminergic system is a case report of N²O administration effectively relieving spastic muscular disorders. 29The other major difference between our study and Tsukada et al. ’s is that they compared awake and lightly anesthetized conditions, whereas we compared lightly and deeply anesthetized conditions. It is possible that the cellular response to isoflurane is biphasic; a small dose may increase DAT expression, whereas a larger dose inhibits it. Further studies are needed to determine if N²O plays a role when measuring the binding potential of DAT-specific radioligands.

The effects of isoflurane on brain physiology have been studied. Isoflurane has been shown to cause an increase in blood flow to the brain via  vasodilatation of the cerebral blood vessels. 30The increased peak uptake and faster washout during 2% isoflurane scanning (fig. 3) is consistent with increased blood flow. Note that if this study were restricted by a flow-dependent tracer, the binding potential would be higher during the high flow state, but this is the opposite of what we see.

Isoflurane has been shown to decrease the glucose metabolic activity of neurons, presumably including dopaminergic neurons. 31However, it has not been shown that altering the metabolic rate of a neuron also alters cell membrane protein expression. Evidence exists that isoflurane interferes with other monoaminergic systems, including γ-aminobutyric acid (GABA) and glutamate. 6,24Gyulai et al.  32have found that isoflurane increases specific binding of [¹¹C]flumazenil to the γ-aminobutyric acid receptor type A (GABA^A) receptors. They suggest that a conformational change of the GABA receptor is involved in the mechanism of action of isoflurane. However, there is no previous evidence of a direct effect of isoflurane on the DAT.

Cellular binding experiments were performed to measure changes in the distribution of the DAT during exposure to isoflurane. The experiments were conducted in DAT-producing HEK cells, which may respond differently than endogenous neurons with a normal complement of genes and machinery for regulating dopamine transporters. Nonetheless, these cells have been used many times to evaluate trafficking of monoamine transporters ex vivo . 33,34Using confocal microscopic imaging, we could easily detect a redistribution of the DAT from the plasma membrane to the cytoplasmic space. This redistribution was characterized by formation of multiple puncta of staining throughout the cytoplasm, a more distinct rim, and a highlighting of the nucleus of the cells. These changes are consistent with isoflurane-causing trafficking of the DAT located in the plasma membrane to the cytoplasm.

The decrease in in vivo [¹⁸F]FECNT binding potential seen here could result from a decrease in apparent DAT density (caused by isoflurane binding to the DAT) or a change in affinity (caused by conformational changes in the DAT) or a combination of both. The trafficking mechanism can explain the results, knowing that [¹⁸F]FECNT is able to freely cross the blood–brain barrier and cell membranes because of its lipophilicity, with the following model. In the different environment inside the cell, the DAT changes conformation and releases dopamine and [¹⁸F]FECNT. [¹⁸F]FECNT only binds with high affinity to the plasma membrane-expressed DAT, which results in a decrease in measured binding potential when DAT is trafficked into the cell. The decrease may be amplified because of competition with endogenous synaptic dopamine, which increases when the functional DAT density decreases.

Taken together, the results of the three experiments support the model of isoflurane causing trafficking of DAT. The distribution of DAT was seen to change in HEK cells in the presence of isoflurane (fig. 6), but the total amount of DAT did not change in vivo  in rats (fig. 5). One drawback of this work is that three different species were used for the three experiments. However, assuming that the dopamine signaling mechanism in rats, rhesus monkeys, and transfected HEK cells is similar, we hypothesize the following model. Isoflurane causes trafficking of DAT. The details of the molecular mechanisms involved are yet to be elucidated. Once inside the cell, the DAT changes conformation and no longer will bind [¹⁸F]FECNT with high affinity. The immunohistochemistry experiments imply that the internalization does not destroy the DAT. The DAT is available for trafficking back to the membrane when the cell receives appropriate signals to cause this to happen.

The authors thank Harriet Robinson, Ph.D., Chief of the Division of Microbiology and Immunology at the Emory University Yerkes Primate Research Center, for providing laboratory space and resources to perform the immunocytochemistry experiments. Research specialists Stephanie Carter (Neurology), Eugene Malveux (Radiology), and Yan Xu (Anesthesiology) performed the in vitro  and ex vivo  experiments. Howard Rees, research specialist in neurology, oversaw collection of the in vitro  microscopic images.