Visible Diode Laser-Induced Fluorescence Detection of Phenylacetic

Detection of Phenylacetic Acid in Plasma. Derivatized with Nile Blue and Using Precolumn. Phase Transfer Catalysis. Sadayappan V. Rahavendran and H...
0 downloads 0 Views 267KB Size
Download PDF
Anal. Chem. 1997, 69, 3022-3027

Visible Diode Laser-Induced Fluorescence Detection of Phenylacetic Acid in Plasma Derivatized with Nile Blue and Using Precolumn Phase Transfer Catalysis Sadayappan V. Rahavendran and H. Thomas Karnes*

Virginia Commonwealth University/Medical College of Virginia, Department of Pharmacy and Pharmaceutics, Box 533, Richmond, Virginia 23298-0533

This study reports the application of Nile blue (NB), a farred oxazine label, as a precolumn derivatization reagent for the measurement of free levels of phenylacetic acid (PAA) in plasma. The measurement of PAA in psychiatric populations is important because it provides a marker for 2-phenylethylamine (PEA), which has been implicated in the pathogenesis of schizophrenia and major depression. PAA was derivatized with NB through an amide linkage in the presence of 2-chloro-1-methylpyridinium iodide (carboxylic acid activator, CMP) and triethylamine (base catalyst, TEA), respectively. The formation of the NBPAA derivative was confirmed using normal phase and reversed phase thin-layer chromatography, reversed phase liquid chromatography, and electrospray mass spectrometry. The formation of the NB-PAA derivative was optimized using a sequential single factor approach. The optimal conditions for the formation and chromatographic separation of the derivative were determined to be 8.0 nmol/mL NB, 390 nmol/mL CMP, 2 µmol/mL TEA, a reaction time of 45 min, and a reaction temperature of 25 °C. This derivatization scheme was performed in a phase transfer catalysis mode that enabled the simultaneous extraction, preconcentration, and derivatization of the analyte in a single step. The limit of derivatization of PAA was determined to be 1.0 × 10-9 M in phosphatebuffered saline, a PAA-free matrix. This derivatization was limited not by the kinetics of the reaction but by the chromatographic separation of the derivative from a side reaction product. The method was used to estimate endogenous free levels of PAA in human plasma samples. The levels of PAA in four sources of plasma were determined to be within 30-70 ng/mL using the method of standard addition and reflected levels that have been reported in the literature. The limit of detection of the derivative was determined to be 7.33 × 10-11 M using a laboratory-constructed HPLC-VDLIF detector. The ability to quantify ultratrace levels of drugs/metabolites in biological matrixes presents a unique challenge to bioanalytical chemists. One of the methods by which this challenge has been met is through the use of high-performance liquid chromatography (HPLC) with laser-induced fluorescence detection (LIF).1 Traditionally, LIF work has been performed using argon ion and (1) Rahavendran, S. V.; Karnes, H. T. Pharm. Res. 1993, 10(3), 328-334.

3022 Analytical Chemistry, Vol. 69, No. 15, August 1, 1997

helium cadmium lasers which provide outputs in the 300-500 nm spectral range. Most of the commercially available derivatization reagents have therefore been developed for this wavelength region.2 Although, very low mass detection limits (femtomole and attomole levels) have been achieved in this region, the sensitivity of analysis is limited by Raman scatter.3 Additionally, endogenous porphyrin-type molecules provide high background fluorescence which results in reduced selectivity when analysis is conducted in biological matrixes.4 HPLC in combination with visible diodelaser induced fluorescence detection (VDLIF) has been shown to provide superior sensitivity and selectivity for the measurement of analytes in biological matrixes.5-7 Nile blue (NB), a far-red oxazine dye (λex ) 633 nm, λem ) 660 nm,  ) 75 000, Φf ) 27%) has been shown, in previous work, to be a useful precolumn derivatization reagent suitable for VDLIF detection of carboxylic acid- containing analytes in neat solution.8 A detection limit of 3.98 × 10-11 M was achieved for benzoic acid derivatized with NB using VDLIF detection with excitation at 635 nm.8 A limit of detection (S/N ) 3) of 5 × 10-15 M was obtained for underivatized NB using a reversed phase column. The linear dynamic range extended over 6 orders of magnitude.9 The applicability of NB for the quantitation of a carboxylic acid analyte in a biological matrix has not been demonstrated. In this work, phenylacetic acid (PAA), endogenously present in plasma, was derivatized with NB and measured by HPLC-VDLIF detection. PAA was derivatized with NB through an amide linkage in the presence of 2-chloro-1-methylpyridinium iodide, a coupling reagent first developed by Saigo et al.10 and later investigated by Lingeman and co-workers.11,12 (2) Van Den Beld, C. M. B.; Lingeman, H. In Luminescence Techniques in Chemical and Biochemical Analysis; (Baeyens, W. R. G., Kenkeleive, D. D., Korkidis, K., Eds; Marcel Dekker: New York, 1990; pp 237-316. (3) Imasaka, T.; Ishibashi, N. Anal. Chem. 1990, 62, 363A-371A. (4) Karnes, H. T.; Rahavendran, S. V.; Gui, M. Proc. SPIE 1995, 2388, 4-7. (5) Mank, A. J. G.; Molenaar, E. J.; Lingeman, H.; Gooijer, C.; Brinkman, U. A. Th.; Velhorst, N. H. J. Pharm. Biomed. Anal. 1995, 13(3), 255-263. (6) Min Gui, Nagaraj, S; Rahavendran, S. V.; Karnes, H. T. Anal. Chim. Acta, in press. (7) Mank, A. J. G.; Molenaar, E. J.; Lingeman, H.; Gooijer, C.; Brinkman, U. A. Th.; Velhorst, N. H. Anal. Chem. 1993, 65, 2197-2203. (8) Rahavendran, S. V.; Karnes, H. T. J. Pharm. Biomed. Anal. 1996, 15, 8398. (9) Rahavendran, S. V.; Karnes, H. T. Instrum. Sci. Technol. 1997, 25(2), 121131. (10) Saigo, K.; Usui, M.; Kikuchi, K.; Shimada, E.; Mukaiyama, T. Bull. Chem. Soc. Jpn. 1977, 50(7), 1863-1866. (11) Lingeman, H.; Haan, H. B. P.; Hulshoff, A. J. Chromatogr. 1984, 336, 241-248. S0003-2700(97)00033-4 CCC: $14.00

© 1997 American Chemical Society

The study of PAA in psychiatric populations is motivated by the evidence that it is the main metabolite of 2-phenylethylamine (PEA), an endogenous biogenic tracer which has been found to be similar in structure and behavioral pharmacology to the psychostimulant amphetamine. PEA may therefore be implicated in the pathogenesis of schizophrenia and major depression.13 Measurement of PAA plasma levels provide a marker for PEA levels. Free PAA plasma levels in humans range between 30 (0.22 nmol/mL) and 300 ng/mL (2.2 nmol/mL), respectively.14 Reports from the literature have commonly used liquid/ liquid extraction of PAA from plasma into an organic solvent and analyzed by HPLC or GC.15 A disadvantage of this method has been loss of PAA due to volatilization as the organic solvent is dried. Presented in this work is application of a precolumn derivatization procedure based on the principle of phase transfer catalysis16 that allows the extraction, preconcentration, and derivatization of the analyte to be accomplished in a single step which prevents the loss of PAA. The synthesized derivative is then separated by reversed phase high-performance liquid chromatography (RP-HPLC) and measured using a laboratory-constructed VDLIF detector. EXPRIMENTAL SECTION Nile blue perchlorate and phenylacetyl chloride were purchased from Aldrich Chemical Co. (Milwaukee, WI). 2-chloro1-methylpyridinium iodide, phenylacetic acid, sodium chloride, potassium chloride, and charcoal-stripped human serum were obtained from Sigma Chemical Co.(St. Louis, MO). Triethylamine and trifluoroacetic acid (HPLC grade) were purchased from Pierce (Rockford, IL). Acetonitrile and methylene chloride were of HPLC grade and obtained from Baxter, Burdick, and Jackson brand (Muskegon, MI). Hydrochloric acid, sodium hydroxide (10 N solution), and bovine serum albumin were obtained from Fisher Scientific (Fair Lawn, NJ). Potassium phosphate monobasic and sodium phosphate dibasic were purchased from Mallinckrodt Inc. (Paris, KY). A Perkin-Elmer Model lambda 2S scanning spectrometer was used for acquisition of absorbance spectral measurements, and a Perkin-Elmer LS 50 scanning luminescence spectrometer equipped with a red-sensitive photomultiplier tube (Hamamatsu R 928) was used to obtain fluorescence spectral measurements (Perkin-Elmer Corp., Rockville, MD). Electrospray ionization mass spectral analysis (ESI-MS) was performed using an Extrel 4000 amu quadropole mass spectrometer. Thin-Layer Chromatographic System and Conditions. Normal phase silica gel GHLF and reversed phase C18 thin-layer plates (Analtech Inc., Newark, DE) with solvent systems consisting of 50% acetonitrile/50% methylene chloride and 50% water/ 50% tetrahydrofuran, respectively, were used for the separation of NB-PAA from unreacted NB and for determination of the respective retardation factors (Rf). Visual detection was performed using a Spectroline Model ENF-280C light source (Spectronics Corp., Westbury, NY). Liquid Chromatographic System and Conditions. System 1. A mobile phase containing 34% acetonitrile/66% water/0.08% (12) Lingeman, H.; Tjaden, U. R., Van Den Beld, C. M. B.; Van Der Greef, J. J. Pharm. Biomed. Anal. 1988, 6(6-8), 687-695. (13) Sharma, R. P.; Faull, K.; Javaid, J. I.; Davis, J. M. Acta Psychiatr. Scand. 1995, 91, 293-298. (14) Fellow, L. E.; Kings, G. S.; Pttit, B. R.; Goodwin, B. L.; Ruthven, C. R.; Sandler, M. Biomed. Mass. Spectrom. 1978, 5, 508-510. (15) Karege, F.; Rudolph, W. J. Chromatogr. 1991, 570, 376-381. (16) De Ruiter, C.; Otten, R. R.; Brinkman, U. A. Th.; Frei, R. W. J. Chromatogr. 1988, 436, 429-436.

trifluoroacetic acid was delivered at a flow rate of 2.0 mL/min using a Shimadzu Model LC-6A pump (Columbia, MD) to a Vydac Model 218 TP, C18 reversed phase HPLC column (250 mm × 4.6 mm i.d.; 5 µm particle size) and a Vydac Model 201GCC guard column (20 mm × 4.6 mm i.d.; 5 µm particle size) packed with Vydac Model 201 TP packing (The Separations Group, Mojave, CA). The guard and analytical columns were maintained at 36 °C using a column heater (Jones Chromatography, Hengoed, Wales). The eluent from the analytical column was alkalinized on-line to a pH of 12. This was achieved using a mixing tee (Model U 429; Upchurch Scientific, Oak Harbor, WA) and a Gilson Model 302 reciprocating pump (Middleton, WI) adjusted to a flow rate of 50 µL to deliver 1 N sodium hydroxide. The postcolumn alkalinized eluent was passed through a conventional fluorescence detector or a laboratory-constructed VDLIF detector. Data acquired from the detector (VDLIF or conventional) were recorded with a Hewlett−Packard Model 3396 A integrator (San Fernando, CA). A Rheodyne Model 7125 manual injector equipped with a 50 µL loop was used to introduce the samples onto the column (Cotati, CA). System 2. The chromatographic system equipped for heart cut column switching17 consisted of a Gilson Model 302 reciprocating pump equipped with a Model 802 manometric module that delivered 35% acetonitrile/65% water/ 0.08% trifluoroacetic acid at 2 mL/min to a precolumn (Vydac Model 201 TP, C18; 150 mm × 4.6 mm i.d., 5 µm particle size) equipped with a Vydac Model 201 GCC guard column (20 mm × 4.6 mm i.d., 5 µm particle size). A mobile phase also containing 35% acetonitrile/65% water/0.08% trifluoroacetic acid was delivered at a flow rate of 2.0 mL/min using a Shimadzu Model LC-6A pump to the analytical column (Vydac Model 218 TP, C18; 250 mm × 4.6 mm i.d.; 5 µm particle size). Heart cut column switching was achieved with an Autochrome Model 401 six-port switching valve (Milford, MA) that was operated manually. The eluent from the analytical column was alkalinized using the method reported in system 1. Conventional Fluoresencee Instrumentation. Detection was performed with a Waters Model 470 scanning fluorescence detector. The excitation and emission monochromators were set at 630 and 680 nm in the absence of postcolumn alkalinization and at 630 and 700 nm in the presence of postcolumn alkalinization. The monochromator slit widths (excitation and emission), gain, attenuation, and time constants were adjusted to 18 nm (excitation and emission), 1000×, 1.0, and 1.5 s, respectively. VDLIF Instrumentation. The laboratory-constructed HPLCVDLIF detector used in this study has been described elsewhere.9 The Melles Griot Model 03FIB016 band-pass filter used in the previous work9 was replaced with a Melles Griot Model 03FIV 024 band-pass filter (λ ) 700 nm, fwhm ) 10 ( 2 nm, peak transmittance 50%) in this work. Derivatization Procedures. Method 1. The NB-PAA derivative was synthesized by reacting 0.78 mL (2.2 nmol/mL) of PAA in the presence of 0.02 mL (100 µmol/mL) of triethylamine (TEA, a base catalyst), 0.10 mL (3.9 µmol/mL) of 2-chloro-1methylpyridinium iodide (CMP, carboxylic acid activator), and 0.10 mL (80 nmol/mL) of NB for 45 min at 25 °C in acetonitrile (Figure 1). The reaction mixture was dried under a stream of dry nitrogen. The residue was reconstituted to 1.0 mL in a solvent (17) Snyder, L. R.; Kirkland, J. J. Introduction to Modern Liquid Chromatography, 2nd ed.; John Wiley & Sons Inc.: New York, 1979.

Analytical Chemistry, Vol. 69, No. 15, August 1, 1997

3023

Figure 1. Derivatization scheme of phenylacetic acid with Nile blue.

mixture consisting of 50% acetonitrile/50% water prior to injection onto the column. The formation of the NB-PAA derivative was maximized by using a “sequential single factor approach”18 (i.e., varying one of the reaction variables while holding the other factors constant). The reaction time, temperature, and molar concentrations of NB, CMP, and TEA were all optimized using this approach. In all of the optimization studies, 0.78 mL of 2.2 nmol/mL PAA was used. Method 2. The NB-PAA derivative was also synthesized by reacting NB with phenylacetyl chloride in the presence of triethylamine for 20 min at 40 °C as described elsewhere.8 Phase Transfer Catalysis (PTC). A 0.20 mL aliquot of blank or PAA-spiked matrix [human plasma, charcoal-stripped human serum, synthetic serum, and phosphate-buffered saline (PBS)] was pipetted into a microcentrifuge vial along with 0.05 mL of 4 M HCl and the resultant mixture vortexed for 10 s. To a separate microcentrifuge vial, 0.78 mL of methylene chloride, 0.02 mL (100 µmol/mL) of TEA, 0.10 mL (11.74 µmol/mL) of CMP, and 0.10 mL (80 nmol/mL) of NB were added and the resultant mixture was vortexed (10 s). The contents of the vial containing NB were transferred to the vial containing the acidified sample; vortex mixed (30 s), and centrifuged (5 min) at 10 000 rpm. The aqueous layer was discarded and the organic layer dried under nitrogen. The residue was reconstituted with 1.0 mL of 50% acetonitrile/ 50% water and a 0.05 mL aliquot injected onto the chromatographic system. Determination of Derivatization Limit. A calibration curve for PAA at concentrations ranging between 73.33 and 2.2 nmol/ mL (10-300 ng/mL) was constructed by derivatization of NB using the PTC method. To determine the derivatization limit of PAA, the molar ratio of NB in the reaction mixture was increased. The derivatization limit and limit of detection were calculated as the concentration that provided a signal 3 times the mean peakto-peak noise (3 × Sp-p). The peak-to-peak noise was determined across the elution window of the intended peak. The limit of detection was determined by serial dilution of the NB-PAA reaction mixture prepared at a higher concentration. The NBPAA derivatives were measured using the HPLC-VDLIF detection (system 1). (18) Morgan, S. L.; Deming, S. N. Anal. Chem. 1974, 46, 1170 -1180.

3024 Analytical Chemistry, Vol. 69, No. 15, August 1, 1997

Figure 2. Identification of the derivative peak through stepwise elimination reactions.

Evaluation of a Phenylacetic Acid-Free Matrix. Human plasma contains endogenous levels of PAA; therefore, matrixes potentially free of PAA were evaluated as a substitute for plasma in preparation of calibration standards. A matrix free of PAA was also considered necessary for determination of the derivatization limit of PAA with NB. These matrixes included charcoal-stripped human serum, synthetic serum, and phosphate-buffered saline, respectively. Synthetic serum was prepared by adding 6.0% (w/ w) bovine serum albumin to PBS. PBS was prepared as described elsewhere.19 Estimation of Phenylacetic Acid Levels in Plasma. The endogenous levels of PAA in the plasma were determined using the standard addition method. A standard addition curve for PAA in plasma was constructed by adding PAA (100-3000 ng/mL PAA prepared in water) to the plasma to provide concentrations that ranged between 10 and 300 ng/mL. RESULTS AND DISCUSSION Confirmation of Nile Blue-Phenylacetic Acid Derivative by TLC and HPLC. The NB-PAA derivative was synthesized (19) Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning, A Laboratory Manual, 2nd ed.; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 1989; Vol. 3.

Figure 4. Effect of the concentration of Nile blue on the yield of the derivative.

Figure 3. Identification of the derivative peak through stepwise elimination reactions. (I) Typical chromatogram obtained with the following reaction mixtures: (a) PAA + TEA + NB; (b) PAA + CMP + NB; (c) TEA + CNP + NB. (II) NB + PAA + CMP + TEA; The derivative peak elutes at 15 min.

by both methods 1 and 2. The Rf values for unreacted NB and NB-PAA for NP-TLC were 0.30 and 0.51, respectively. The Rf values using RP-TLC for unreacted NB and NB-PAA were 0.37 and 0.15, respectively. The spot identified as the NB-PAA derivative synthesized by both methods had similar Rf values with respect to one another using both NP and RP-TLC. A peak identified as the NB-PAA derivative was observed reproducibly at 15 min using HPLC (35% acetonitrile/65% water/0.08% TFA) with conventional fluorescence detection. The formation of the NB-PAA derivative using HPLC was confirmed by carrying out elimination reactions using method 1 (Figure 2). The HPLC-separated NB-PAA derivative was collected into a test tube by the method of peak trapping.8 The peak-trapped derivative was then introduced into the ESI-MS system for structural confirmation based on molecular weight. The mass spectrum yielded a molecular ion peak at m/z 436.5 which closely matched the calculated molecular weight of 437 for the derivative. Fluorescence Characteristics of the Derivative. The NBPAA derivative demonstrated excitation and fluorescence emission maxima at 630 and 680 nm, respectively. The molar absorptivity and relative quantum yield of the derivative were determined to be 18 000 M-1 cm-1 and 3.10%.8 NB was used as the reference fluorophore for calculation of the relative quantum yield.8 The fluorescence of the NB-PAA derivative at pH 12 was 10-fold higher when compared to the fluorescence of the derivative at pH 2 and was determined to be similar to that reported for the NB-benzoic acid derivative.8 At a pH of 12 and with excitation set at 630 nm, the fluorescence maximum and relative quantum yield of the NB-PAA derivative was observed to be 700 nm and 15%, respectively. Optimization of Derivatization Conditions. Figure 3 shows a graph of the reaction time and the temperature conditions that were employed for optimization of the NB-PAA formation. The yield of NB-PAA was highest at a temperature of 50 °C; however, the reaction temperature was chosen to be 25 °C because the NB-PAA derivative at this temperature was stable during the reaction time studied compared to 50 °C. In addition, the reaction could be performed at room temperature without the need for a

Figure 5. Optimization of CMP, a carboxylic acid activator, in the reaction mixture for maximum formation of the derivative.

Figure 6. Effect of acetonitrile/water composition (reconstitution solvent) on the peak area of the derivative.

heating block. The reaction time was established for maximum formation of the derivative and was determined to be 45 min. The concentration of NB in the reaction mixture was increased from 2 to 500 nmol/mL. Figure 4 demonstrates that the peak identified as due to the NB-PAA derivative was increased by increasing the concentration of NB. A 8.0 nmol/mL concentration of NB in the reaction mixture was established for subsequent analyses. This concentration was determined to be optimal because there was difficulty in separating the unreacted NB from the derivative using Analytical Chemistry, Vol. 69, No. 15, August 1, 1997

3025

Figure 7. Evaluation of selectivity: (I) chromatogram of derivatized blank PBS. (II) Chromatogram of a derivatized PBS sample, containing 10 ng/mL phenylacetic acid. The derivative peak elutes at 17.50 min.

system 1 at higher concentrations of NB. A 390 nmol/mL concentration of CMP in the reaction mixture was determined to be optimal for the formation of the NB-PAA derivative (Figure 5). We observed that the percentage of water in the reconstitution step of the dried reaction mixture prior to injection onto the column affected the peak area of the derivative (Figure 6). A reconstitution solvent comprised of 50% acetonitrile/50% water was therefore used because it provided solubility of the dried reaction mixture residue. The sequential single factor approach was also employed for the PTC derivatization method. The optimal reaction temperature was determined to be 25 °C and the reaction time determined to be 10 s. A CMP concentration of 1.194 µmol/mL, a NB concentration of 8 nmol/mL, and a TEA concentration of 2 µmol/ mL in the reaction mixture were established as optimal for formation of the NB-PAA derivative by the PTC method. Chromatographic System. System 2 utilizing heart cut column switching was employed to prevent overlapping of the peak from excess unreacted NB peak from the NB-PAA derivative peak. This method enabled separation of excess unreacted NB from the derivative even at NB concentrations of up to 1000 nmol/mL in the reaction mixture. The heart cut timing was established at 20 and 22 min for switching onto and off of the 3026 Analytical Chemistry, Vol. 69, No. 15, August 1, 1997

Figure 8. (I) Chromatogram of a derivatized plasma sample. (II) Chromatogram of a plasma sample to which phenylacetic acid has been added and derivatized. The derivative peak elutes at 17.50 min.

analytical column, respectively, with the derivative having a retention time of 33 min. An increase in the peak area of the derivative peak resulted upon increasing the NB concentration in the reaction mixture from 8.0 to 100.0 nmol/mL. Along with this increase in the derivative peak area was the presence of a closely eluting peak that also increased in size. The source of this peak was investigated and observed to occur when NB, TEA, and CMP were present in the reaction mixture. This peak may be due to an impurity present in the Nile blue since it increased in size only with increasing concentrations of NB and not with the other reactants. We were able to achieve baseline separation of the impurity peak from the NB-PAA peak by reducing the mobile phase composition of the analytical column to 32% acetonitrile. This, however, resulted in longer retention times with broadening of the peak and loss of sensitivity. On the basis of these results, it was decided to utilize system 1 and prepare the derivative (PTC method) using 8.0 nmol/mL NB in the reaction mixture. Resolution of greater than 1.5 was obtained between the impurity peak and the derivative by dilution of the reaction mixture prior to injection onto the column.

Derivatization Limit. Charcoal-stripped human serum, synthetic serum, and PBS were evaluated as possible matrixes devoid of PAA. Charcoal-stripped human serum and synthetic serum derivatized with NB (PTC method) demonstrated peaks which appeared at the retention time of NB-PAA. Phosphate-buffered saline, however, provided a matrix free of PAA (Figure 7). The derivatization limit was therefore evaluated using PBS as the matrix. The constructed calibration curve (Y ) 1.9 × 108X + 1.6 × 107) was linear between 10 and 300 ng/mL and demonstrated a correlation coefficient of 0.981. A limit of derivatization of 1.0 × 10-9 M was obtained. A lower limit of derivatization could be accomplished provided the impurity peak could be eliminated. The application of the PTC method may reduce a problem stated by Mank et al. in which low derivatization limits were difficult to achieve due to slower kinetics at lower analyte concentrations.7 This may be true because the analyte is concentrated at the aqueous-organic layer interface which may allow for in PTC more efficient derivatization kinetics.16 Phenylacetic Acid Levels in Plasma. A correlation coefficient of 0.990 was obtained for the calibration curve (Y ) 1.94 × 105X + 1.3 × 107) constructed with the standard addition method using a plasma matrix. The free level of endogenous PAA in a single plasma sample was determined to be 68 ng/mL. Endogenous free levels of PAA in three other plasma sources were determined to be within the 30-70 ng/mL concentration range. This experiment demonstrates that the method is capable of measuring free PAA levels in human plasma.14 A typical chromatogram is shown in Figure 8. Limit of Detection. Serial dilution of the NB-PAA derivative prepared by reaction of 8.0 nmol/mL and 7.26 nmol/mL PAA in the presence of 2 µmol/mL TEA and 1.139 µmol/mL CMP (PTC method) resulted in a limit of detection (S/N ) 3) of 7.33 × 10-11 M. CONCLUSIONS This application has shown that NB is a suitable label for derivatization of PAA in plasma. The advantages of using NB as

a label include the presence of a highly reactive functional group, low cost, solubility in RP-HPLC solvents, and the ability for optimal excitation with a suitable wavelength diode laser. The impurity in NB limited the ability to measure lower levels of PAA due to the limited chromatographic separation achievable with the present system. Phase transfer catalysis as a precolumn derivatization technique has been found to be an attractive method in this application. The sample preparation time is reduced to a few minutes, and the derivatization process occurs in less than 1 min using mild conditions. Organic solvent usage is also minimized and the volatilization of the analyte has been prevented. The laboratory-constructed instrument was found to be suitable for the measurement of the NB-PAA derivative. The combination of HPLC-VDLIF detection with postcolumn addition of base is advantageous due to the enhanced Stokes shift of the ionized derivative as well as the superior fluorescence characteristics. ACKNOWLEDGMENT The authors thank Dr. William H. Soine of the Department of Medicinal Chemistry, Medical College of Virginia, for advice on the derivatization chemistries; Hari haran Nair and Dr. Vicki Wysocki, Department of Chemistry, University of Arizona for ESIMS data on the derivative; and Paul Kostel of Vydac, The Separations Group, Hesperia, CA, for the gift of a Vydac Model 218 TP, C18 reversed phase column.

Received for review January 10, 1997. Accepted April 17, 1997.X AC9700336

X

Abstract published in Advance ACS Abstracts, July, 1, 1997.

Analytical Chemistry, Vol. 69, No. 15, August 1, 1997

3027