|Home | About | Journals | Submit | Contact Us | Français|
Sensitive analytical methods are needed for the separation and quantification of neurotransmitters obtained in microdialysate studies. This unit describes methods that permit quantification of nanomolar concentrations of monoamines and their metabolites (high-pressure liquid chromatography electrochemical detection), acetylcholine (HPLC-coupled to an enzyme reactor), and amino acids (HPLC-fluorescence detection; capillary electrophoresis with laser-induced fluorescence detection).
This unit describes methods for the separation and quantification of neurotransmitters in microdialysis samples by high-pressure liquid chromatography (HPLC) coupled to either electrochemical (EC) or fluorescence detection and a more recently developed technique, capillary electrophoresis laser-induced fluorescence detection, for detection of monoamines and amino acids. Other methods including gas chromatography (GC; Griffin et al., 2007), radioimmunoassay (RIA; Maidment et al., 1989), and mass spectrometry (Lanckmans et al., 2006) are used for quantification of drugs and other analytes. Methods of sufficient sensitivity for the routine quantification of neuropeptides are limited (Baseski et al., 2005; Babu et al., 2006) and are not discussed here.
When selecting a method for separation and quantification, several issues must be considered. First, the concentration of most neurochemicals in the extracellular space is quite low. Because temporal resolution of an analyte in a dialysis sample is inversely related to volume, the analytical method employed should be that which provides detection limits below the lowest concentration expected in the dialysate and that which requires the smallest sample volume. Because of the low volume of microdialysate samples, pipetting or sample-cleanup techniques are often impossible. Finally, the perfusion medium itself contains neurochemicals and inorganic ions that may interfere with the quantification method employed (unit 7.1). The protocols describe commonly used procedures for the detection of catecholamines and indoleamines (HPLC-EC detection; see Basic Protocol 1), detection of acetylcholine (HPLC-EC detection with enzyme reactor; see Basic Protocol 2 and Alternate Protocol 1), and detection of amino acids (HPLC-fluorescence detection; see Basic Protocol 3; and capillary electrophoresis laser-induced fluorescence, see Basic Protocol 4). Each of these protocols affords detection limits in the low nanomolar range.
Ungerstedt and co-workers (Zetterstrom et al., 1983) were the first to combine HPLC with electrochemical detection (EC) to analyze dopamine and its metabolites in brain dialysate samples. This technique was refined by several laboratories (Church et al., 1987; Abercrombie et al., 1988; Church and Justice, 1989; Pettit and Justice, 1991) and provides a sensitive method with which to quantify catecholamines and indoleamines. In HPLC/EC, oxidation/reduction occurs at a fixed point along a flowing eluent (mobile phase). The mobile phase passes in a thin layer through the cell over the electrode. As the single layer of mobile phase passes over an electrode held at a fixed potential, it is oxidized or reduced. If the potential is greater than that required for oxidation/reduction of the analyte, a charge passes between the electrode and solute. The resultant current is directly proportional to the concentration of analyte passing through the cell. This current is amplified and sent to a recorder to yield a chromatogram.
There are two approaches to EC detection for HPLC: amperometric and coulometric. In the amperometric technique, some fraction (usually 5% to 15%) of the analyte is oxidized or reduced. This is in contrast to the coulometric detector, which enables essentially 100% conversion of the analyte. By placing an additional electrochemical cell before the analytical cell, it is possible to oxidize or reduce compounds that may co-elute with the analyte of interest, thereby preventing them from interfering with the analysis. Both the amperometric and coulometric EC methods depend on electron transfer between the mobile phase (solute) and the electrode surface. Therefore, mobile-phase composition is critical. The mobile phase must have a sufficiently high dielectric constant to permit ionization of the electrolyte and must be electrochemically inert at the electrode surface (i.e., background current should be low). Column size and composition are also important factors in determining sensitivity of the method. This protocol may be used for the amperometric detection of either catecholamines/indoleamines together with their metabolites (using mobile phase buffer I, running time: ~15 min) or dopamine alone (using mobile phase buffer II, running time: ~5 min).
A phosphate buffer allows fast elution and optimal separation selectivity for choline and acetylcholine on a cation exchange column. Acetylcholine is neither electroactive nor detectable by optical methods. The most frequently used method for acetylcholine detection is based on enzymatic conversion of acetylcholine into choline and acetate by acetylcholinesterase, and then subsequent oxidation of choline by choline oxidase to betaine and H2O2. The latter can be oxidized on a platinum electrode. Enzymes can be covalently attached to packing material to form an immobilized enzyme reactor (IMER). Bioanalytical Systems and ESA sell a kit that includes the analytical column and the IMER, both in microbore and normal bore versions, which greatly facilitates method development. This method allows detection of acetylcholine concentrations as low as 50 fmol on the column, permitting quantification of dialysate acetylcholine concentrations of 5 to 10 nM. However, these concentrations are higher than those in typical brain dialysates. Therefore, an acetylcholine esterase inhibitor (1 to 10 μM neostigmine) is frequently added to the perfusion buffer to increase dialysate acetylcholine to detectable levels. It is important to perform pilot experiments to determine the minimum concentration of neostigmine needed to reliably quantify dialysate acetylcholine under the experimental and analytical conditions employed. Artificially increasing acetylcholine levels by using acetylcholinesterase inhibitors alters the mechanisms responsible for the physiological regulation of extracellular acetylcholine levels and may interfere with the goals of the experiment. Recently, improvements in sensitivity have been reported by replacing the platinum electrode with a carbon electrode coated with peroxidase enzyme. The peroxidase oxidizes H2O2 and the resulting electrons are transferred to the electrode surface by a redox polymer and detected on the carbon surface operating in reduction mode. A kit containing the peroxidase and the redox polymer necessary to coat the carbon electrode is available from Bioanalytical Systems (see Alternate Protocol 1).
H2O2 can also be detected on a glassy carbon electrode that has been “wired” with a redox polymer containing peroxidase instead of the platinum electrode used in Basic Protocol 2. The peroxidase oxidizes H2O2 and the resulting electrons are transferred to the carbon electrode surface by the redox polymer and detected there with the electrode in the reduction mode. Bioanalytical Systems supplies a kit that allows the coating of a regular glassy carbon electrode. The peroxidase-wired electrode has much shorter equilibration times than platinum electrodes, as well as lower background current, and allows an about five to ten times increase in assay detection limits (5 to 10 fmol in column), making it possible to detect acetylcholine in the dialysate with much lower concentrations of neostigmine (in the nanomolar range) or, depending on brain area and microdialysis conditions, no neostigmine at all.
Peroxidase/redox polymer coating kit (Bioanalytical Systems cat. no. MF-2095) Glassy carbon electrode
Amino acids in microdialysates can be readily separated using reversed-phase high-performance liquid chromatography (RP-HPLC). This protocol describes the detection of primary amino acids treated with o-phthaldialdehyde (OPA) in the presence of 2-mercaptoethanol at basic pH in an aqueous solution. The reaction yields isoindole derivatives of the primary amino acids, rendering them highly fluorescent (Fig. 7.4.1). Once injected into the HPLC system, the derivatized sample is carried at a constant flow rate through the chromatography column in an optimized buffer, which constitutes the mobile phase. Bonded to the matrix of the column are long carbon chains (C18 column), which form a hydrophobic stationary phase. Separation occurs as the derivatized amino acids in the sample are retained in the column on the basis of their relative hydrophobicity in the mobile phase. Retention times range from ~2 min for aspartate to 25 min for γ-aminobutyric acid (see Anticipated Results). The fluorescence detector can measure picogram quantities of the OPA-derivatized amino acids as they elute from the column.
The method described in Basic Protocol 3 is able to resolve a mixture of 10 amino acids present in brain dialysates. However, the running time of the chromatographic method is ~30 min. This long running time can be a problem when many samples need to be analyzed, as is often the case in microdialysis experiments. If the amino acids of interest are limited to glutamate (Glu), aspartate (Asp), or GABA, there is the possibility to optimize the assay to improve detection of those amino acids and decrease running time at the expense of losing the rest of the amino acids. The two methods described here are based on those originally described by Kehr (1998a,b) and involve the use of microbore columns and an organic phase flush step. One of the problems with the chromatographic methods for amino acids is the existence of very late-eluting peaks that can contaminate chromatograms during repeated runs. One solution to this problem is to flush the column with a high concentration of organic solvent after the amino acids of interest have eluted. This causes all the late peaks to elute from the column. However, this normally involves the use of a gradient pump and more sophisticated HPLC equipment. A viable alternative is the use of microbore columns. The internal volume of those columns is so small that they can be flushed with a small volume of organic solvent (i.e., 20 μl). This can be achieved by injecting the solvent through the manual or automatic injector. The use of microbore columns enhances the assay sensitivity since the smaller column volume results in less dilution of the injected sample.
Although high-pressure liquid chromatography (HPLC) remains the most commonly used method for quantification of neurotransmitters in dialysate samples, it provides only moderate mass sensitivity. Consequently, HPLC analysis requires large injection volumes and, hence, long microdialysis sampling times (generally 5 to 15 min). As a consequence, the limited time resolution is often insufficient to monitor rapid chemical changes linked to neuronal activity. This temporal resolution, already poor when compared with the duration of neurobiological events, is further decreased when samples are split for the simultaneous determination of different classes of neurotransmitters. Temporal resolution can be improved by coupling microdialysis with capillary electrophoresis (CE), a separation technique, which not only requires lower sample volumes, possesses a high mass sensitivity (10−18 mol) and high separation efficiency, but also allows rapid separation of a large number of compounds. Due to its plug-like flow and minimal diffusion, CE possesses enormous resolving power and large peak capacities. With this approach, sampling times as short as 5 to 30 sec have been reported (Bowser and Kennedy, 2001; Kennedy et al., 2002; Parrot et al., 2004; Powell and Ewing, 2005), allowing the monitoring of rapid changes associated with neuronal events. Although many of these studies have been performed using home-made CE instruments, which cannot easily be set up in most biology laboratories, recent studies have shown that short sampling times can also be achieved by coupling microdialysis with commercially available CE laser-induced fluorescence (LIF) detection systems.
The separation of Glu and Asp is readily achieved by capillary zone electrophoresis (CZE), a separation mode in which the background electrolyte is a simple electrolyte solution such as phosphate, borate, or citrate buffer. The electrophoretic migration and the electroosmotic flow allow the separation of molecules. These forces, as well as the characteristics of the capillary (dimension, coating), differentiate each separation system. In this case, only charged molecules can be separated, since neutral compounds migrate together with the electroosmotic flow. In CE, detection of neurotransmitters can be made by LIF or by electrochemistry (EC). Though no EC detector is presently commercially available, some laboratories use home-made EC detectors to identify electroactive compounds such as catecholamines or derivatized amino acids. However, use of EC is complicated by the need to decouple the electrophoretic current from the amperometric detector. In contrast, LIF detectors are commercially available and user-friendly while providing detection limits in the same range as EC detectors.
As neurotransmitters are not fluorescent at available laser wavelengths, microdialysis samples must be derivatized with a tagging agent (Prata et al., 2001). Fluorescent reagents such as naphthalene-2,3-dicarboxyaldehyde (NDA), o-phthaldialdehyde, or fluorescein isothiocyanate, react with the primary amine function of neurotransmitters, allowing their detection following laser excitation at 442, 325, or 488 nm. Naphthalene-2,3-dicarboxyaldehyde (NDA) is a reagent of choice because (1) it is not fluorescent itself (in contrast with fluorescein isothiocyanate, for instance), (2) it reacts rapidly in presence of CN− to give stable fluorescent derivatives (cyanobenzo[f]isoindol or CBI products) with minute amounts of amino acids and biogenic amines, and (3) NDA derivatives can be detected after excitation with the 442-nm He-Cd laser or the more robust 410-nm diode laser. OPA is also an efficient tagging agent for amine neurotransmitters, but the derivatives formed are relatively unstable. Moreover, OPA derivatives are only detected after excitation by a 325-nm He-Cd laser, which may be less reliable.
This protocol describes a commonly used CZE-LIF procedure (Robert et al., 1998) for the detection of excitatory amino acids. Alternate Protocol 3 describes a procedure that enables detection of inhibitory and excitatory amino acids using another mode of CE, namely micellar electrokinetic chromatography with laser-induced fluorescence (MEKC-LIF, Sauvinet et al., 2003). Each of these protocols affords detection limits in the low nanomolar range.
NOTE: Because primary amine contaminants can affect detection, special care must be taken to avoid any dust or bacterial contamination. HPLC-grade water, pipet tips and microcentrifuge (PCR) tubes should be autoclaved. Gloves must be worn during manipulation of samples and reagents.
NOTE: Filtration of solutions is important for proper functioning of CE since particulate matter can clog the capillary lumen leading to baseline disturbances and current failure. Degassing in an ultrasonic bath is not essential, but strongly recommended since it will avoid formation of micro-bubbles inside the capillary due to Joule heating.
Addition of surfactant to the CE buffer generates an enhanced degree of resolution and aids in the separation of analytes with similar charges, such as biogenic amines. Surfactants, the most commonly used being sodium dodecyl sulfate (SDS), are added at a relatively high concentration and form micelles within the background electrolyte. Micelles exhibit a hydrophobic core and are considered a mobile “pseudo-stationary” phase, by analogy with the fixed stationary phase of chromatographic columns. This mode of separation is called micellar electrokinetic chromatography (MEKC). The analytes are resolved according to their hydrophobicity, as a consequence of their interaction with micelles and their charge-to-mass ratio. Consequently, compounds exhibiting similar charge-to-mass ratios, like NDA derivatives of monocarboxylic amino acids, can be separated.
The following protocol may be used for the determination of γ-aminobutyric acid (GABA), Glu, and L-Asp within microdialysates.
The procedure for MEKC-LIF analysis is similar to Basic Protocol 4 described above for CZE analysis of excitatory amino acids, with the following important modifications:
Since amino acids are not fluorescent, microdialysis samples need to be derivatized with a reagent to give a fluorescent end product. The procedure for derivatization of samples with NDA is described here. NDA is not fluorescent itself but readily reacts with primary amines in the presence of CN to give stable fluorescent derivatives. After derivatization, samples can be analyzed by CE-LIF.
Manual derivatization of sample volumes <2 μl cannot be accomplished with good reproducibility. Since NDA reacts rapidly with the primary amine moiety of neurotransmitters, continuous flow derivatization (online derivatization) can be performed on nanovolumes of microdialysates in a microreactor placed at the outlet of a microdialysis probe. This procedure has the advantage of continuously adding very small volumes of reagents to the perfusate sample with good reproducibility, while preventing evaporation and sample loss. Systems for online derivatization can easily be constructed but must have a minimal dead volume to limit diffusion of solutes in the microdialysis tubing (Parrot et al., 2004). In this respect, the use of tubing with very small inner diameters and very slow perfusion rates allows submicroliter volumes to be collected with good reproducibility, and results in a high microdialysis sampling rate. The online derivatization of submicroliter volumes of sample and femtomole amounts of compounds does not decrease sensitivity as compared to manual derivatization (see Support Protocol 1). The following protocol describes the reagents and procedures for online derivatization. The procedures can be used with the CZE or MEKC protocols described above.
Use HPLC-grade water in all recipes and protocol steps. For common stock solutions, see APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.
Stock solutions: Prepare diluent by adding 3.0 ml glacial acetic acid to 1 liter HPLC-grade water. Adjust pH to 5.25 with 1 M NaOH. Add 5 ml of commercial 1% Proclin solution (bacteriostatic, Bionalytical Systems), and store up to 1 month at 4°C. Prepare 2.0 mM acetylcholine stock solution by dissolving 36.3 mg acetylcholine chloride in 100 ml diluent and 2 mM choline stock solution by dissolving 27.9 g choline chloride in 100 ml diluent. Store up to 1 month at 4°C.
Working solutions: On day of assay, prepare fresh 20 μM working standards and other standard concentrations as needed by diluting the stock solutions with diluent.
Acetylcholine and choline are highly hygroscopic. Store powder in a desiccator and minimize the time the bottle is open.
Working solutions: Prepare 10 μM working standards on day of assay by diluting 1 ml of stock solutions in 100 ml of the perfusate buffer used to collect the microdialysate of interest. Prepare other concentrations as needed to bracket the estimated concentrations of the samples.
The monoamines can be obtained from a variety of sources including Sigma and Research Biochemicals. Keep powder and standard solutions protected from light.
To prepare 2.925 mM NDA (0.539 mg/ml of acetonitrile-water, 50:50) stock solution, accurately weigh an amount of NDA between 2 and 5 mg. Calculate the volume of acetonitrile and water to be added. Add acetonitrile first, then water. Store up to 1 week at 4°C in a foil-wrapped, capped glass container.
To prepare 87 mM NaCN (4.264 mg/ml of water) stock solution, weigh an amount of NaCN between 4 and 8 mg. Add the required volume of water. Store up to 1 week at 4°C in a capped glass container.
To prepare buffer for derivatization, mix 500 mM (3.1 g/100 ml) H3BO3 and 125 mM (4.8 g/100 ml) borax to obtain a pH 8.7 solution. Store in a sealed glass container up to 1 month at ambient temperature.
Prepare borax buffer (0.1 M sodium tetraborate, pH 10.4 with 10 M NaOH).
Prepare OPA stock solution (10 mg OPA, bring into solution in 100 μl methanol and add 9.9 ml of borax buffer).
Prepare β-mercaptoethanol stock solution (dilute 1/10 in methanol in a fume hood).
Keep the stocks in capped vials and protected from light. The stocks are stable for up to 1 month at room temperature. To prepare fresh working solution, mix 0.9 ml of borax buffer, 0.1 ml of OPA stock solution, and add 3 μl of β-mercaptoethanol stock.
The borax buffer tends to precipitate at 4°C. Filter and keep at room temperature.
Stock solutions: Prepare diluent by adding 2.07 ml of 37% HCl to 250 ml of HPLC-grade water.
Prepare 1 mM d,l-α amino adipic acid stock solution (16.1 mg in 100 ml diluent).
Prepare 1 mM d,l- glutamic acid stock solution (14.7 mg in 100 ml diluent).
Prepare 1 mM d,l- aspartic acid stock solution (13.3 mg in 100 ml diluent).
Dispense into 1.5-ml microcentrifuge tubes. Store up to 6 months at −20°C.
Working solutions: Prepare diluent for internal standard by adding 1.00 ml of 70% HClO4 to 100 ml of HPLC-grade water. On the day of assay, prepare fresh 10 μM working internal standard by diluting the α amino adipic acid stock solution with diluent. Prepare glutamic acid and aspartic acid standard concentrations as needed by diluting the respective stock solutions with aCSF.
Stock solutions: Prepare diluent by adding 2.07 ml of 37% HCl to 250 ml of HPLC-grade water.
Prepare 1 mM L-cysteic acid stock solution (8.5 mg in 100 ml diluent).
Prepare 1 mM D,L-glutamic acid stock solution (14.7 mg in 100 ml diluent).
Prepare 1 mM L-aspartic acid stock solution (13.3 mg in 100 ml diluent).
Prepare 1 mM GABA stock solution (5.2 mg in 100 ml diluent).
Dispense into 1.5-ml microcentrifuge tubes. Store up to 6 months at −20°C.
Working solutions: Prepare diluent for internal standard by adding 1.00 ml of 70% HClO4 to 100 ml of HPLC-grade water. On the day of assay, prepare fresh 50 μM working internal standard by diluting the stock solution of L-cysteic acid with diluent. Prepare amino acid standard concentrations as needed by diluting the respective stock solutions with aCSF.
Liquid chromatography was the first method used for the measurement of catecholamines and indoleamines in biological samples. Isolation/purification of specific catecholamines was achieved by adsorption onto either acid-washed alumina (catecholamines) or cation-exchange resins (catecholamines and indoleamines; Anton and Sayre, 1962). Fluorescence analysis with or without chemical derivatization was then employed to identify specific amines (Welch and Welch, 1969). These methods, however, afforded limited sensitivity and selectivity. Three techniques that offered greater sensitivity and selectivity were subsequently developed. Gas chromatography with mass-spectrometric detection (Koslow et al., 1974) resulted in improved selectivity and detection limits. However, its use was limited by expense and the need to derivatize compounds. The second approach employed the enzyme catechol O-methyl-transferase and radioactive labeling to increase the selectivity of primary amine determination. This technique increased detectability relative to that attainable via fluorescence methods. However, safety concerns and expense precluded its routine use. The third approach, liquid chromatography combined with electrochemical detection (HPLC/EC; Kissinger et al., 1977), was developed in the early 1970s and is now the most commonly used technique for the quantification of amines and their metabolites. However, the use of HPLC with EC is often more problematic than with ultraviolet or fluorescence detection. Although detection limits are typically lower, noise can profoundly affect performance. Electrode connections, pump pulsations, bubbles on the electrode, and temperature must be rigorously controlled. The applied potential is a major determinant of the signal-to-noise ratio, and voltammograms can be used to determine the optimal setting to be applied. Mobile phase composition is critical for the separation of the various amines. Knowledge of pKa values and the effects of ion-pairing agents is essential for the development of a mobile phase that will provide optimal separation and detection of electroactive species.
Dialysate levels of acetylcholine can be readily quantified using reversed-phase high performance liquid chromatography coupled with electrochemical detection (HPLC/EC). The HPLC/EC method for acetylcholine analysis was originally developed by Potter et al. (1983). This technique was modified and further refined by Damsma et al. (1987), who converted the reversed-phase analytical column into a cation-exchange column by loading sodium lauryl sulfate (or sodium dodecyl sulfate; SDS) onto the reversed-phase column.
The typical assay involves an analytical column packed with a material based on silica, with covalently bonded cationexchange/reversed-phase groups. A phosphate buffer allows fast elution and optimal separation selectivity for choline and acetylcholine. The covalent binding of acetylcholinesterase and choline oxidase onto an immobilized enzyme reactor (IMER) allows the sequential and stoichiometric conversion of acetylcholine to acetate and choline, and of choline to betaine and hydrogen peroxide. The hydrogen peroxide is then electrochemically detected via oxidation on a platinum electrode. Recently, a new method has been developed involving coating of a glassy carbon electrode with a peroxidase/redox polymer, which allows for faster stabilization and lower background currents, resulting in lower detection limits (Kehr et al., 1998).
First described by Roth (1971), the reaction of primary amino acids with OPA in the presence of mercaptans yields strongly fluorescent isoindole derivatives (Fig. 7.4.1). Basic Protocol 3 describes a precolumn derivatization in which the reaction with OPA is performed before injection of the sample into the HPLC system. The reaction with OPA is rapid and amenable to automation. Retention of OPA-derivatized amino acids is governed by their relative hydrophobic and hydrophilic interactions with the hydrocarbon stationary phase and the aqueous-methanolic mobile phase (Lindroth and Mopper, 1979). Although amino acids derivatized with OPA can also be detected by coulometry (e.g., using a dual-electrode coulometric-amperometric analytical cell with potentials of −0.03V and +0.65V), fluorimetric detection offers a higher level of sensitivity, which is often necessary when quantifying low nanogram amounts of amino acids present in microdialysate samples.
Microdialysates can be analyzed online, i.e., just after the outlet of the probe, through an analytical interface allowing the direct coupling of the microdialysis probe to the CE instrument, or offline, i.e., after sample collection in vials, which may be stored frozen before analysis.
Since dialysates are protein-free, online coupling between microdialysis and laboratory-made CE-LIF instruments have been described to prevent sample loss and evaporation. The online instrument removes the requirement of collecting, storing, derivatizing, and analyzing large numbers of nanoliter-volume dialysate fractions. Using various interfaces connected to a continuous flow derivatization device, 20-sec to 3-min sampling rates have been reported for the in vivo monitoring of excitatory amino acids (Robert et al., 1998; Bowser and Kennedy, 2001). Microdialysis sampling could be coupled via a flow-gated interface online to CE for in vivo monitoring of neuroactive amino acids and amines. In the instrument, analytes are derivatized precolumn with OPA and β-mercaptoethanol to form fluorescent isoindole products (Bowser and Kennedy, 2001). With online methods, however, temporal resolution becomes limited by the speed of the analytical method used. In addition, for an accurate monitoring of fast events, one must take into account the total delay in the response of the online system to a biochemical event.
The analysis of low-volume micro-dialysates can also be performed offline using a commercially available CE-LIF system, provided that online derivatization of sample is carried out (Parrot et al., 2004). Consequently, this technique can be set up in neuroscience laboratories that have no access to a specialized workshop for making custom-made analysis instruments. Furthermore, one advantage of the offline approach is to uncouple micro-dialysis sampling from the CE analysis. Off-line analysis can be useful since, if any breakdown of the CE-LIF system occurs, microdialysates can be stored up to 3 days prior to analysis. In contrast, if online analysis is used, samples cannot be saved and data are lost. Thus, offline analysis offers more flexibility for planning experiments.
In CE, the small inner diameter of the capillary tube is well adapted to the analysis of extremely small sample volumes: samples ranging from a few microliters down to nanoliters can routinely be injected.
When microdialysates are manually collected and analyzed offline, the minimal sample volume required for the CE analysis must be determined. Sample injections are currently made at one side of the capillary by applying pressure. The volume of the sample must be sufficient (1) to allow the capillary to plunge into it, preventing the injection of air microbubbles with the sample, and (2) to avoid any significant loss by evaporation when a series of dialysates (i.e., at least 30 samples) is placed in the CE sample rack before being injected. The authors have found that 940 nl is the minimal volume required for the analysis of a large series of sample with a good reproducibility. With a shorter series of samples, determination of amino acid neurotransmitter content can be performed using 500 nl of derivatized microdialysates.
The reduction of microdialysis sample volumes needed for neurotransmitter analysis using CE-LIF allows slower perfusion rates, thereby increasing temporal resolution. With slow perfusion rates, the concentration gradients developed by the probe are less marked, a characteristic that improves the spatial resolution and decreases disturbance of neural tissue. Moreover, the increased concentration of sample that is removed at slow perfusion rates improves the detection limit for concentration-limited measurements.
Finally, once it has become possible to handle even lower sample volumes, the use of slower perfusion rates (<0.1 μl/min) will allow direct measurement of the extracellular concentration of neurotransmitters in dialysates. As a consequence, in vivo calibration methods such as quantitative no-net-flux or extrapolation to zero flow rate, which are time consuming, could be avoided. However, if the “no-net-flux” method is used for determining the true extracellular concentration, short collection times allow the duration of these “no-net-flux” experiments to be shortened.
Since a derivatization reaction with a fluorescent tag is needed to detect the molecules of interest and since most of the fluorescent reagents used react with primary amines, a large number of the neurochemical constituents of brain microdialysis samples can potentially be detected. In this respect, the use of CE-LIFD makes possible the high-sensitivity determination of various neurotransmitters in a single microdialysis sample with improved temporal resolution. This can be performed via several analyses. The low volume requirement of CE-LIFD allows several subsequent offline analyses to be performed on the same sample of microdialysate. The advantage of such multiple assays is the ability to study functional or drug-induced interactions between neurotransmitters.
Measurement of dopamine and its metabolites in dialysates using an HPLC/EC detection method (see Basic Protocol 1) presents a number of challenges. Some of these are due to characteristics of the dialysate in general and some to characteristics of brain dopamine systems. These issues are as follows.
During routine analysis, it is a constant struggle to prepare a mobile phase that provides good separation of dopamine from its metabolites. In the authors' experience, minimizing organic solvent, manipulating pH, and prolonging the run time improve chromatographic separation of the dopamine peak from other electrochemical signals. Mobile phase and electrochemical detector conditions need to be modified to suit the characteristics of the transmitter of interest. Similarly, protocols for amino acid analysis in rat samples (unit 7.2) may not be suitable for monkey samples (unit 7.3). It is critical that mobile phase and detector conditions be modified for optimum detection of the transmitter of interest.
Mobile phase composition is critical for HPLC/EC detection of dopamine and its metabolites. Retention times are affected by alterations in room temperature as well as by pH, the concentration of ion-pairing reagent, and the inorganic solvent employed. These can be varied to modify analyte resolution. Increasing pH decreases the retention time of acidic metabolites (e.g., DOPAC and HVA). Increasing the concentration of ion-pairing reagents (e.g., octane-sulfonate, octylsulfate, pentanesulfonate, or hexanesulfonate) delays retention of amines, enabling better separation from charged molecules. Increasing solvent concentration results in a decreased retention time of amines and their metabolites. Each of these parameters can be altered to increase analyte resolution.
Additional troubleshooting guidelines can be found in Table 7.4.1.
The microbore column is easily clogged, making an inline precolumn filter or guard column useful. Guard columns should be replaced on a regular basis, especially when increased pressure or a loss of performance is observed.
Retention times of acetylcholine and choline primarily depend on the mobile phase ionic strength. A higher ionic strength will result in shorter retention times.
Platinum is a soft material and can scratch easily. If a loss of sensitivity can be traced to the performance of the electrode, avoid unnecessary polishing and try first to activate the electrode electrochemically.
Until recently, detection of dialysate levels of acetylcholine could only be achieved from samples perfused with an aCSF solution containing the acetylcholinesterase inhibitor neostigmine. There is, however, increasing evidence indicating that the micromolar concentrations of neostigmine employed can markedly modify the neurochemical effects of various drugs. Therefore, attempts should be made to use as low a concentration of neostigmine as possible. Several laboratories are now able to measure basal levels of acetylcholine in various brain areas in the absence of neostigmine. Therefore, pilot studies should be conducted to determine the minimum concentration of neostigmine needed for the region examined.
Bacterial contamination of the pump results in the production of enzymes that scavenge hydrogen peroxide. Therefore, the system should be flushed periodically. The addition of a preservative (Proclin) to the mobile phase is also recommended. When setting up the method for the first time, it is critical to passivate the system to ensure a clean system.
Acetylcholine and choline are hygroscopic. Therefore, the solids should be allowed to equilibrate to room temperature before weighing. The solids should be stored in a freezer. Working standards can be refrigerated for up to 2 weeks at 4°C.
Columns should not be in contact with organic solvents such as methanol or acetonitrile. If the system is not to be used for some time, the flow rate of the mobile phase should be lowered or the columns should be washed and stored at 4°C.
A decrease in sensitivity may be caused by bacterial contamination, loss of enzymatic activity in the IMER, a dirty electrode, or insufficient oxygen in the mobile phase.
Deterioration of the OPA reagent can lead to decreased sensitivity. The pH must be basic and 2-mercaptoethanol must be present as the nucleophile for the reaction to yield fluorogenic amino acid isoindole derivatives. Therefore, stock OPA reagent should be prepared weekly. The OPA reaction is also dependent on an aqueous medium. The reaction stops when the OPA/sample mixture is injected into the HPLC system, as a result of the lower pH and presence of methanol in the mobile phase. It is important to dilute the stock OPA reagent with an aqueous solution other than the mobile phase, preferably with the perfusate buffer used in the collection of microdialysates. Un-reacted OPA will elute immediately after aspartate. The presence of this peak can indicate improper mixing of the sample with the OPA reagent, resulting in an underestimation of microdialysate amino acid concentrations. Note that the amino acids asparagine and histidine may appear as peaks directly preceding and following glutamate elution. Detection of amino acids (e.g., serine and glycine) in an OPA-derivatized perfusion buffer blank indicates the presence of bacterial contamination. Amino acids can be detected in microdialysate samples that have been stored frozen at −70°C using this protocol.
The main problems encountered with this method are exogenous contaminants, increases in column pressure, especially when using microbore columns, and high background due to air in the system.
Contrary to other neurotransmitters like catecholamines, indoleamines and acetylcholine, amino acids are very stable at room temperature. They are also ubiquitous in dust, sweat, and organic matter. For this reason, exogenous contamination of samples and solutions is common. It is good practice to wear gloves when handling solutions and to thoroughly clean glassware. It is also necessary to run blanks of the aCSF or Ringer solutions used for preparing standards and running the dialysis experiments.
Increases in HPLC system pressure are also a common problem. The microbore columns are especially sensitive to clogging by particulate matter due to the small inner diameter. Adequate filtering of all the solutions is required. It is advisable to install a precolumn filter with minimal dead volume to trap any particulate matter coming from the injector (even if samples are clean, a faulty valve seal or other moving components in the injector can be the source of particles). Sudden increases in pressure are likely caused by particulate matter trapped in the filter. Sonicating or replacing the filter normally takes care of the problem. Eventually, pressure builds up over time. If replacing the filter does not help, this means that pressure has built up in the column. The repetitive acetonitrile flushes pose a strain on the column packing and eventually it will become loose and pack at the end of the column, increasing the pressure. Sometimes the life of the column can be lengthened by reversing the column and flushing with 40% acetonitrile for several hours (disconnect the column from the detector when attempting this). Finally, slow buildup in pressure may also be due to accumulation of proteins. Since dialysates do not contain proteins, accumulation of proteins in the column likely reflects bacterial contamination in the system upstream from the column. The precolumn filter will not trap proteins in solution. In this case, a guard column may be installed instead of the precolumn filter.
Occasionally, a gradual increase of background may be observed >1 hr, followed by a complete loss of signal. This may be fixed by either preparing a fresh degassed mobile phase or sometimes just degassing the old mobile phase. This problem is related to the high concentration of acetonitrile in the mobile phase since it is normally observed with the GABA mobile phase and it seems related to oxygen in the mobile phase quenching the fluorescent signal.
Since the CE-LIF technique was developed for the analysis of brain extracellular fluid obtained by microdialysis, the derivatization procedure has to be performed on small volumes and with standards dissolved in the perfusion fluid. The derivatization rate is typically not affected by the higher ionic strength of biological samples when compared to standards in aqueous solution, as evidenced by lack of variation in peak areas when samples are prepared either in water or aCSF (A. Zapata, V.I. Chefer, T.S. Shippenberg, and L. Denoroy, unpub. observ).
The rate of derivatization, stability of the labeled amino acids, and amino acid quantification varies for each amino acid. For example, the reaction of NDA/NaCN with GABA is quicker than with Glu and Asp. This is probably due to the lower steric hindrance in the vicinity of the amine group of GABA relative to other amino acids, similar to that observed after derivatization of histidine and histamine with NDA/NaCN (Zhang and Sun, 2004). In contrast, derivatives of GABA seems less stable that those of Glu and Asp; this may be due to the greater hydrophilic properties of Glu and Asp derivatives.
In vitro response to rapid step changes that occur in the external medium in which the microdialysis probe is placed have shown that rapid fluctuations (<5 sec) of the external medium produce significant responses of the system. Thus, the time spent by the solutes in crossing the membrane is not a limiting factor for monitoring short-lasting events. However, high-sampling-rate microdialysis can be performed only if dead volumes are minimized both at the outlet and at the inlet (if local application of drugs by reverse dialysis is to be carried out) of the dialysis system.
The local application of drugs through the microdialysis probe is performed by switching perfusion fluid from aCSF to an aCSF containing known concentrations of drugs. The control animals do not receive any drug, but switching between two identical perfusion fluids must be performed to take into account short-lasting flow disturbances, which become more apparent with short collection times.
On the other hand, microdialysis experiments on awake animals require long dialysis tubing to allow free movement of the animal. However, the dead volume of the tubing, i.e., between the active dialysis membrane and the outlet of the derivatization system, must be minimized when frequent collection is performed. Indeed, if the interval of time between dialysis and collection is greater than the sampling time, solutes could diffuse significantly between sample plugs. The dead volume can be calculated, but the dead time has to be determined experimentally in order to correlate neurochemical results to physiological events for in vivo studies (e.g., behavior or electroencephalographic recordings).
A lack of reproducibility in the migration time of an analyte could be due to excessive variability in the electroosmotic flow. Changing to a new capillary may take care of this problem.
Storage time is critical in maintaining the original composition of microdialysis samples. Different analytes have different preservation times.
This contamination is likely to be due to the presence of amino acids on the glassware. Purity of water and reagents should also be investigated.
A chromatogram showing retention times and peak heights of monoamines and metabolites in mouse striatal tissue samples using mobile phase I is shown in Figure 7.4.2. Note the use of the internal standard DHBA. A basal dialysate sample obtained from mouse nucleus accumbens using mobile phase II is shown in Figure 7.4.3.
A chromatogram of a mouse striatal dialysate is shown in Figure 7.4.4. The aCSF contained 30 nM neostigmine and 12 μl were injected into the HPLC system. The use of the precolumn choline oxidase catalyst IMER results in elimination of the choline peak. In combination with the wired peroxidase electrode, this system allows for a detection limit for acetylcholine of 2 to 3 nM.
Amino acids with a wide range of polarities can be detected using the HPLC fluorescence detection protocol (see Fig. 7.4.5). Later-eluting peaks such as GABA may broaden with isocratic elution (Fig. 7.4.5A). In general, increasing the methanol concentration of the mobile phase will decrease the retention times of the derivatized sample components. Full resolution of all amino acids, including hydrophobic residues (e.g., valine, isoleucine, and leucine), requires a more complex series of isocratic elution steps coupled with gradient elution displacement (Umagat et al., 1982). The regional distribution of major amino acids in rat-brain microdialysates using OPA derivatization has been published (Tossman et al., 1986). Note that changes in glutamine and taurine evoked by high potassium occur acutely (see Fig. 7.4.5B,C); however, detection of glutamate release in these microdialysates may be offset by rapid uptake via glutamate transporters. Inclusion of the competitive-uptake inhibitor l-trans-pyrrolidine-2,4-dicarboxylate (Massieu et al., 1995) will augment recovery of acutely released glutamate.
A disadvantage of the OPA reaction is its lack of suitability for the detection of thiols in microdialysates. N-(1-pyrenyl)maleimide is an appropriate derivatizing agent (Winters et al., 1995). The preparation time for mobile phase and OPA would be ~30 to 45 min for the novice. These components should be prepared the day before the run; the working OPA diluted from stock works best after overnight storage and the HPLC column requires a mobile phase equilibration period.
The detection of GABA, in particular, suffers from a long retention time, which dramatically increases the time needed to analyze a single microdialysis experiment. In addition, the longer retention time results in decreased sensitivity. If only GABA (or Glu) is of interest, use of the Alternate Protocols will decrease the chromatographic running time and, in the case of GABA, significantly increase assay sensitivity. Chromatograms of mouse nucleus accumbens dialysate are shown in Figure 7.4.6.
Electropherograms showing separation of Glu and Asp (Fig. 7.4.7A) by CZE and GABA, Glu and Asp (Fig. 7.4.7B) by MEKC are shown in Figure 7.4.7. Note the separation of the internal standards. Both electropherograms correspond to dialysis samples. See figure legend for dialysis and analysis conditions.