- Email: [email protected]
Accurate and efficient method for quantification of urinary histamine by gas chromatography negative ion chemical ionization mass spectrometry
ANALYTICAL
BIOCHEMISTRY
Accurate
136,258-263
(1984)
and Efficient Method for Quantification of Urinary Histamine by Gas Chromatography Negative ion Chemical ’ Ionization Mass Spectrometry II AND JOHN A. OATES
L. JACKSON ROBERTS, Departments
of Pharmacology
and Medicine,
Vanderbilt
University,
Nashville,
Tennessee
37232
Received July 5, 1983 Because of the recognized inaccuracy and unreliability of currently available methods for the quantification of histamine in biological fluids, a method for quantification of urinary histamine by stable isotope dilution assay with negative ion chemical ionization mass spectrometry has been developed. Following the addition of [2H4]histamine to 1 ml of urine, histamine is extracted into butanol, back-extracted into HCl, derivatized to the pentafluorobenzyl derivative (CH,C&,),histamine, extracted into methylene chloride, and then quantified with negative ion chemical ionization mass spectrometry by selected ion monitoring of the ratio of ions m/z 430/434. Twenty samples can be assayed in 2 days. Precision of the assay is +2.7% and the accuracy is 97.6%. Lower limits of sensitivity are approximately 100-500 fg injected on-column. This assayprovides a very sensitive, accurate, and efficient method for the quantification of histamine in human urine. KEY WORDS: histamine; mass spectrometry; negative ion chemical ionization; gas chromatography; urinary.
The biological importance of histamine is well recognized. Histamine is a physiological mediator of gastric acid secretion (1) and is a granule-associated mediator in mast cells which is released during immediate hypersensitivity reactions (2-4). We have also recently provided evidence that histamine plays an important role as a mediator of the foregut carcinoid syndrome associated with a gastric carcinoid tumor (5,6). In patients with mastocytosis overproduction of histamine also occurs and is released in substantial quantities during attacks of flushing and hypotension (7-l 1). Because of these important physiological and pathophysiological roles of histamine, a means to accurately quantify histamine is obviously of value in attempts to assessthe participation of histamine in various biological events. In addition, the demonstration of overproduction of histamine is considered important biochemical evidence to support a diagnosis of mastocytosis. 0003-2697184 $3.00 Copyright 0 1984 by Academic Fxss. Inc. All rights of reproduction in any form mewed.
A variety of methods have been described for the determination of histamine in biological fluids. Currently the most widely used methods for quantification of histamine are the radioenzymatic assays(4,12- 14) and either manual or automated fluorometry (15-17). With increasing awareness, however, it has been recognized that these methods for analysis of histamine are associated at times with inaccuracy and unreliability. A recent study convincingly elucidated the substantial inaccuracies associated with both single and double isotope radioenzymatic assays as well as manual and automatic fluorometric methods for the determination of histamine (18). In our own attempts to quantify histamine in urine by single and double isotope radioenzymatic assays, we also encountered severe problems and inaccuracy which could not be re@y rectified. We have therefore undertaken the development of an accurate method for the quantification of histamine in biological fluids. Be258
CHROMATOGRAPHIC/SPECTROPHOTOMETRIC
cause of the recognized accuracy associated with the physical method of analysis of biological compounds by stable isotope dilution assay with quantification by gas chromatography-mass spectrometry (GC/MS),’ we have adapted this methodology for the determination of histamine. Recent reports have described a method for the analysis of histamine using electron impact ionization EI-MS. ( 192 1). Negative ion chemical ionization (NICI) mass spectrometry in most instances, however, has been shown to be associated with substantially enhanced sensitivity and specificity compared to EI-MS. We have, therefore, directed our efforts to adapt NICI-MS methodology for the determination of histamine. The enhanced sensitivity and specificity of NICI-MS compared to EI-MS potentially would allow more simplified and efficient purification procedures of histamine from biological fluids prior to analysis as well as permit the more routine determination of histamine in small volumes of certain biological fluids in which the levels of histamine are very low. We report here a very sensitive, accurate, and efficient method developed for the quantification of histamine in urine using GC/NICIMS. As an initial basis for the development of this method, we examined potentially suitable derivatives of histamine for analysis by NICI-MS and reported the formation and characteristics of a pentafluorobenzyl derivative of histamine formed by treatment of histamine with pentafluorobenzyl bromide (22). With selected ion monitoring GC/NICI-MS of this derivative of histamine, (CH&F&histamine, lower limits of detection were found to be in the range of 100-500 fg injected on-column. MATERIALS
AND
GENERAL
METHODS
Pentafluorobenzyl bromide and diisopropylethylamine were obtained from Pierce Chemical Company (Rockford, Ill.). Hista’ Abbreviations used: GC, gas chromatography; MS, mass spectrometry; MCI, negative ion chemical ionization; EI, electron impact ionization.
HISTAMINE
QUANTIFICATION
259
mine dihydrochloride was purchased from Sigma Chemical Company (St. Louis, MO.). Histamine [cr,cr,B,@-2H&lihydrochloride was obtained from Merck Isotope, Montreal, Canada. Poly I- 110 gas chromatography column packing was obtained from Applied Science, (State College, Pa.). NICI mass spectral analysis was performed using a Hewlett-Packard 5982A gas chromatograph/mass spectrometer modified to detect negative ions. Conditions: electron energy, 250 eV; interface temperature, 300°C; internal source temperature, 225°C; direct inlet line, emission current 300 PA; methane as reagent and carrier flow gas; analyzer manifold pressure, 1.2 X 1Ow5Torr; injection port temperature, 250°C; conversion diode potential, -3 kV. Analysis was performed using a 2-ft packed column of Poly I- 110 operated at 250°C. RESULTS
Histamine
Derivative
As a basis for the development of this assay, we have reported the formation of a pentafluorobenzyl derivative of histamine with very favorable characteristics for determination by NICI-MS (22). Treatment of histamine with pentafluorobenzyl bromide was found to derivatize the side chain a-nitrogen to a tertiary nitrogen and also derivatized the imidazole ring tele-nitrogen yielding (CH&F&-histamine. The NICI mass spectrum of (CH2C6F&histamine is shown in Fig. 1. This mass spectrum is characterized by a single negative ion of high intensity at m/z 430 which is formed by the loss of 181 + (2 X 20) mass units, CH2C6FS + (2 X FH), from the molecular ion at m/z 65 1. The analogous ion in the NICI mass spectrum of (CH&F5)3-histamine[oc,c&/~-~H~] retains all four deuterium atoms and thus is m/z 434. The final quantification of urinary histamine in this assay is accomplished by selected ion monitoring of the ratio of the intensity of the ions m/z 430 (endogenous [2Ho]histamine) and m/z 434 ([2H4]histamine).
260
ROBERTS
AND
OATES
‘bF5
w
40
I
410
2
611
M-181-(3x20) M-181-(4x20) 0 50
100
150
200
250
300
350
400
450
500
550
600
650
700
m/2
1. Negative
ion chemical
Assay Method for Urinary
Histamine
FIG.
ionization
To 1 ml of urine is initially added 40 ng of [*H4]histamine and 150 ~1 of 1 N NaOH in 16 X 125-mm disposable glass tubes. After addition of 2 ml of butanol the tubes are vortexed. Heptane (2 ml) is added and the tubes are again vortexed and then centrifuged to facilitate separation of organic and aqueous phases. The organic layer of butanol:heptane is transferred to 12-ml conical centrifuge tubes and 100 ~1 of 1 N HCl is added. After vortexing and centrifugation, the organic layer is aspirated and discarded. The aqueous phase is transferred to a l-ml Reacti-vial and evaporated under a stream of nitrogen, and the residue is dissolved in 30 ~1 of acetonitrile to which is added 10 ~1 of diisopropylethylamine and 15 ~1 of a 25% solution of pentafluorobenzyl bromide in acetonitrile. The derivatization reaction is allowed to proceed for 30 min at room temperature after which excess reagents are evaporated under a stream of nitrogen. The residue is then dissolved in 150 ~1 of 10% Na2COJ to which is added 250 ~1 of methylene chloride followed by vortexing and centrifugation. The upper aqueous layer is aspirated and discarded. To accomplish complete removal of all Na$O,containing water in the vial before evaporating the organic layer, the vial is filled with distilled water without mixing, aspirated, and discarded twice. The methylene chloride is then evap-
mass spectrum
of (CH&F&histamine.
orated to dryness under a stream of nitrogen and the residue is dissolved in 50 ~1 ethyl acetate, and l-2 ~1 is injected for analysis by GC/NICI-MS. Assay Results and Validations
A representative selected ion current chromatogram of m/z 430 and 434 obtained from an analysis as described above of urine collected from a normal individual is depicted in Fig. 2. As seen, the purification procedure outlined satisfactorily removes urinary impurities which could interfere with the accurate determination of the intensity of the m/z 430 and 434 ion peaks from the [*Ho]- and
h em/r430
FILAMENT ON (PEN OFFSET .5 CM)
1
lNJr U&.,,, 0 0
1
1
1 I
1
I
2WIO I
012345676 TIME WN.) FIG.
gram urine
2. Representative selected ion current chromatoof m/z 430 and 434 obtained from the analysis of from a normal volunteer.
CHROMATOGRAPHIC/SPECTROPHOTOMETRIC
HISTAMINE
[*I-I&&amine present. In an occasional urine we have observed an additional impurity ml z 430 peak with a retention time of approximately 1 min. However, this peak is very sharp and returns to baseline before reaching 2 min retention time and thus does not interfere with the histamine m/z 430 peak. Except for this single noninterfering m/z 430 impurity peak seen in an occasional urine, similar selected ion current chromatograms as depicted in Fig. 2 free of interfering impurity peaks have been reproduced in 125 different urines thus far analyzed by this method. Several procedures have been performed to establish the accuracy of this method for the determination of histamine. Initially a standard curve was constructed by adding varying amounts of [‘Ho]histamine to a fixed quantity of 22.81 ng of [*H4]histamine and the ratio of m/z 4301434 was determined directly following derivatization and compared to the expected ratios based on the amounts of [*Ho] and [*H4] present. The results obtained are shown in Fig. 3. The standard deviations for each determination from replicate injections into the mass spectrometer were less than the size of the dots in the figure. As can be seen, this standard curve was linear and accurate
v
I 10-Z lo-’ IO0 IO’ IO2 %
HISTAMINE
I
ADDED
FIG. 3. Standard curve for the analysis of histamine by negative ion chemical ionization mass spectrometry. Varying quantities of unlabeled histamine were added to 22.81 ng of [2H,]histamine as noted on the abcissa. On the ordinate are plotted the actual ratios of m/z 430 (unlabeled histamine) to m/z 434 ([*H,]histamine) obtained with selected ion monitoring analysis.
QUANTIFICATION
261
over a lO,OOO-fold range of varying ratios of m/z 4301434.
We then confirmed that the m/z 430 and 434 peaks observed in the analysis of urine were only generated by the urinary histamine and [*H4]histamine present in the sample rather than in part contributed by potentially coeluting urinary impurities. When urine was analyzed without the addition of [*H4]histamine, no m/z 434 peak was observed. We then treated 1 ml of urine with 16.25 mg of diamine oxidase at 37°C for 30 min. Acetone was then added to precipitate the protein and [*H4]histamine added. Precipitated protein was centrifuged and petroleum ether was added and the organic phase of acetone:petroleum ether aspirated and discarded. The remaining 1 ml aqueous phase was then processed according to the standard assay procedure. The ratio of m/z 4301434 in the diamine oxidase-treated urine was reduced by 96% compared to the ratio obtained with the analysis of 1 ml of the same urine which had not been treated with diamine oxidase. We then processed 5 ml of urine and analyzed the sample by both NICI-MS and EI-MS. The ions monitored for quantification in the EI mass spectrum of (CH2C6F&-histamine were m/z 470 for [*HJhistamine and m/z 474 for [*H,Jhistamine (22). The selected ion monitoring ratio of m/z 4701474 obtained by EIMS differed from the ratio of m/z 4301434 obtained by NICI-MS by only 2.7%. Because the mass spectra of (CH2C6FS),-histamine analyzed by NICI and EI-MS are different and thus the ions monitored for quantification are different, obtaining essentially identical quantitative data when urine was analyzed by both NICI-MS and EI-MS provides convincing evidence that the m/z 430 and 434 peaks observed in the analysis of urine by NICI-MS are only generated by the histamine present rather than in part contributed by coeluting urinary impurities. Experiments were then performed to establish the precision and accuracy of the assay. For the determination of precision, six l-ml aliquots of a single 24-h urine collection were
262
ROBERTS
AND OATES
processed and analyzed individually. For determination of accuracy, to an additional six 1-ml aliquots of the same 24-h urine collection were added 30.39 ng of [‘Halhistamine and the urine subsequently processed and analyzed. The difference between the total amount of histamine measured in these urines to which histamine had been added and the amount measured in the urines without the addition of histamine was then compared to the expected value of 30.39 ng of histamine which had been added. The precision had a coefficient of variation of 2.7%. The accuracy of measuring the histamine added to urine was 97.6%. DISCUSSION
Because of the recognized inaccuracies associated with the widely used radioenzymatic and fluorometric methods for the analysis of histamine in biological fluids, we have developed an accurate physical method for the determination of histamine using GC/NICI-MS. Although the method described in this report pertains to the quantification of urinary histamine, certainly with minor modifications in the purification scheme the procedure could be adapted for the analysis of histamine in a variety of biological fluids. In addition to accuracy this assay is also associated with a remarkable degree of sensitivity. As previously reported, with selected ion monitoring of m/z 430, the lower limit of detection of (CH&F&-histamine injected on column is in the range of 100-500 fg. It is primarily because of the much greater sensitivity associated with NICI-MS compared to EI-MS that we attempted to adapt NICI-MS methodology for the determination of histamine. Although the sensitivity of this assay is not absolutely required for the analysis of urinary histamine in which the normal concentrations present are in the range of several nanograms per milliliter, this degree of sensitivity should easily permit the determination of histamine in small volumes of other biological fluids such as plasma in which the normal levels of histamine are very low.
We routinely process 1 ml of urine for ease and efficiency since the sensitivity of this method would allow the absurd number of well over 200 injections into the mass spectrometer for quantification of each 1-ml sample of urine being analyzed. However, because of this sensitivity, the efficiency of the assay is enhanced due to the fact that time-consuming repeat extractions and careful solvent aspirations to accomplish a high recovery of the histamine through the assay procedure can be avoided. Routinely about 20 samples can be processed and analyzed in 2 days. The sample capacity of this assay is approximately the same as the sample capacity of the radioenzymatic methods of analyses. We are currently in the process of a study rigorously establishing the range of the urinary excretion of histamine in normal volunteers and assessing potential influences of sex, age, and other factors which may alter the levels of urinary histamine excreted. The results of this study will be the subject of a future report. In summary, we have developed a very sensitive, accurate, and efficient method for the quantification of urinary histamine. With minor modifications, it should be possible to adapt this method for the determination of histamine in a variety of biological fluids. The use of this method should contribute to investigations assessing the physiological and pathophysiological participation of histamine in various biological events and diseases. ACKNOWLEDGMENTS This work was supported by Grants GM 15431 and 507-RR05424 from the National Institutes of Health. Dr. Roberts is a Burroughs Wellcome Scholar in Clinical Pharmacology. Dr. Oates is the Joe and Morris Werthan Professor of Investigative Medicine. We would also like to acknowledge the generous contribution of Dr. David Begs of the Hewlett-Packard Corporation who supplied blueprints to allow modification of the HP-5982A mass spectrometer to detect negative ions. The invaluable technical assistance of K. A. Aulsebrook and A. M. Kelleghan was greatly appreciated. REFERENCES 1. Black, J. W., Duncan, W. A. M., Durant, C. J., Ganellin, C. R., and Parsons, E. M. (1972) Nature (London)
236, 385-390.
CHROMATOGRAPHIC/SPECTROPHOTOMETRIC 2. Ishizaka, T., Ishizaka, K., and Tomioka, H. (1972) J. Immunol. 108, 513-520. 3. (1973) in Mechanisms in Allergy: Reagin-mediated Hypersensitivity (Goodfiiend, L., Sehon, A. H., and Orange, R. P., eds.), Dekker, New York. 4. Brown, M. J., Ind, P. W., Causon, R., and Lee, T. H. (1982) J. Allergy Clin. Immunol. 69,20-24. 5. Roberts, L. J., II, Mamey, S. R., Jr., and Oates, J. A. (1979) New Engl. J. Med. 300, 236-238. 6. Roberts, L. J., II, Bloomgarden, Z. T., Mamey, S. R., Jr., Rabin, D., and Oates, J. A. (1983) Gustroenterology 84, 272-275. 7. Roberts, L. J., II, Fields, J. P., and Oates, J. A. (1982) Trans. Assoc. Amer. Phys. 45, 36-4 1. 8. Demis, D. J. (1963) Ann. Int. Med. 59, 194-206. 9. Roberts, L. J., II, Sweetman, B. J., Lewis, R. A., Austen, K. F., and Oates, J. A. (1980) New Engl. J. Med. 303, 1400- 1404. 10. Roberts. L. J., II, Turk, J. W., and Oates, J. A. (1982) Advan. Shock Res. 8, 145-152. I 1. Turk, J., Oates, J. A., and Roberts, L. J. (1983) J. Allergy C/in. Immunol. 71, 189-192.
HISTAMINE
QUANTIFICATION
263
12. Snyder, S. H., Baldessarini, R. J., and Axelrod, J. (1966) J. Pharmacol. Exp. Thu. 153, 544-549. 13. Miller, R. L., McCord, C., Sauda, M., Boume, H. R., and Melmon, K. L. (1970) J. Pharmacol. Exp. Ther. 175,228-234. 14. Beaven, M. A., Jacobsen, S., and Horakova, Z. (1972) Clin. Chim. Acta 37, 91-103. 15. Roberts, M., and Adam, H. M. (1950) Brit. J. Pharmacol. 5, 526-541. 16. Oates, J. A., Marsh, E., and Sjoerdsma, A. (1962) Clin. Chim. Acta 7, 488-497. 17. Siraganian, R. P. (1974) Anal. Biochem. 57,383-394. 18. Gleich, G. J., and Hull, W. M. (1980) J. Allergy Clin. Immunol. 66,295-298. 19. Mita, H.. Yasueda, H., and Shida, T. (1979) J. Chromatogr. 175, 339-342. 20. Mita, H., Yasueda, H., and Shida, T. (1980) J. Chromatogr. 181, 153-159. 21. Mita, H., Yasueda, H., and Shida, T. (1980) J. Chromatogr. 221, l-7. 22. Roberts, L. J., II, and Oates, J. A., submitted for publication.
BIOCHEMISTRY
Accurate
136,258-263
(1984)
and Efficient Method for Quantification of Urinary Histamine by Gas Chromatography Negative ion Chemical ’ Ionization Mass Spectrometry II AND JOHN A. OATES
L. JACKSON ROBERTS, Departments
of Pharmacology
and Medicine,
Vanderbilt
University,
Nashville,
Tennessee
37232
Received July 5, 1983 Because of the recognized inaccuracy and unreliability of currently available methods for the quantification of histamine in biological fluids, a method for quantification of urinary histamine by stable isotope dilution assay with negative ion chemical ionization mass spectrometry has been developed. Following the addition of [2H4]histamine to 1 ml of urine, histamine is extracted into butanol, back-extracted into HCl, derivatized to the pentafluorobenzyl derivative (CH,C&,),histamine, extracted into methylene chloride, and then quantified with negative ion chemical ionization mass spectrometry by selected ion monitoring of the ratio of ions m/z 430/434. Twenty samples can be assayed in 2 days. Precision of the assay is +2.7% and the accuracy is 97.6%. Lower limits of sensitivity are approximately 100-500 fg injected on-column. This assayprovides a very sensitive, accurate, and efficient method for the quantification of histamine in human urine. KEY WORDS: histamine; mass spectrometry; negative ion chemical ionization; gas chromatography; urinary.
The biological importance of histamine is well recognized. Histamine is a physiological mediator of gastric acid secretion (1) and is a granule-associated mediator in mast cells which is released during immediate hypersensitivity reactions (2-4). We have also recently provided evidence that histamine plays an important role as a mediator of the foregut carcinoid syndrome associated with a gastric carcinoid tumor (5,6). In patients with mastocytosis overproduction of histamine also occurs and is released in substantial quantities during attacks of flushing and hypotension (7-l 1). Because of these important physiological and pathophysiological roles of histamine, a means to accurately quantify histamine is obviously of value in attempts to assessthe participation of histamine in various biological events. In addition, the demonstration of overproduction of histamine is considered important biochemical evidence to support a diagnosis of mastocytosis. 0003-2697184 $3.00 Copyright 0 1984 by Academic Fxss. Inc. All rights of reproduction in any form mewed.
A variety of methods have been described for the determination of histamine in biological fluids. Currently the most widely used methods for quantification of histamine are the radioenzymatic assays(4,12- 14) and either manual or automated fluorometry (15-17). With increasing awareness, however, it has been recognized that these methods for analysis of histamine are associated at times with inaccuracy and unreliability. A recent study convincingly elucidated the substantial inaccuracies associated with both single and double isotope radioenzymatic assays as well as manual and automatic fluorometric methods for the determination of histamine (18). In our own attempts to quantify histamine in urine by single and double isotope radioenzymatic assays, we also encountered severe problems and inaccuracy which could not be re@y rectified. We have therefore undertaken the development of an accurate method for the quantification of histamine in biological fluids. Be258
CHROMATOGRAPHIC/SPECTROPHOTOMETRIC
cause of the recognized accuracy associated with the physical method of analysis of biological compounds by stable isotope dilution assay with quantification by gas chromatography-mass spectrometry (GC/MS),’ we have adapted this methodology for the determination of histamine. Recent reports have described a method for the analysis of histamine using electron impact ionization EI-MS. ( 192 1). Negative ion chemical ionization (NICI) mass spectrometry in most instances, however, has been shown to be associated with substantially enhanced sensitivity and specificity compared to EI-MS. We have, therefore, directed our efforts to adapt NICI-MS methodology for the determination of histamine. The enhanced sensitivity and specificity of NICI-MS compared to EI-MS potentially would allow more simplified and efficient purification procedures of histamine from biological fluids prior to analysis as well as permit the more routine determination of histamine in small volumes of certain biological fluids in which the levels of histamine are very low. We report here a very sensitive, accurate, and efficient method developed for the quantification of histamine in urine using GC/NICIMS. As an initial basis for the development of this method, we examined potentially suitable derivatives of histamine for analysis by NICI-MS and reported the formation and characteristics of a pentafluorobenzyl derivative of histamine formed by treatment of histamine with pentafluorobenzyl bromide (22). With selected ion monitoring GC/NICI-MS of this derivative of histamine, (CH&F&histamine, lower limits of detection were found to be in the range of 100-500 fg injected on-column. MATERIALS
AND
GENERAL
METHODS
Pentafluorobenzyl bromide and diisopropylethylamine were obtained from Pierce Chemical Company (Rockford, Ill.). Hista’ Abbreviations used: GC, gas chromatography; MS, mass spectrometry; MCI, negative ion chemical ionization; EI, electron impact ionization.
HISTAMINE
QUANTIFICATION
259
mine dihydrochloride was purchased from Sigma Chemical Company (St. Louis, MO.). Histamine [cr,cr,B,@-2H&lihydrochloride was obtained from Merck Isotope, Montreal, Canada. Poly I- 110 gas chromatography column packing was obtained from Applied Science, (State College, Pa.). NICI mass spectral analysis was performed using a Hewlett-Packard 5982A gas chromatograph/mass spectrometer modified to detect negative ions. Conditions: electron energy, 250 eV; interface temperature, 300°C; internal source temperature, 225°C; direct inlet line, emission current 300 PA; methane as reagent and carrier flow gas; analyzer manifold pressure, 1.2 X 1Ow5Torr; injection port temperature, 250°C; conversion diode potential, -3 kV. Analysis was performed using a 2-ft packed column of Poly I- 110 operated at 250°C. RESULTS
Histamine
Derivative
As a basis for the development of this assay, we have reported the formation of a pentafluorobenzyl derivative of histamine with very favorable characteristics for determination by NICI-MS (22). Treatment of histamine with pentafluorobenzyl bromide was found to derivatize the side chain a-nitrogen to a tertiary nitrogen and also derivatized the imidazole ring tele-nitrogen yielding (CH&F&-histamine. The NICI mass spectrum of (CH2C6F&histamine is shown in Fig. 1. This mass spectrum is characterized by a single negative ion of high intensity at m/z 430 which is formed by the loss of 181 + (2 X 20) mass units, CH2C6FS + (2 X FH), from the molecular ion at m/z 65 1. The analogous ion in the NICI mass spectrum of (CH&F5)3-histamine[oc,c&/~-~H~] retains all four deuterium atoms and thus is m/z 434. The final quantification of urinary histamine in this assay is accomplished by selected ion monitoring of the ratio of the intensity of the ions m/z 430 (endogenous [2Ho]histamine) and m/z 434 ([2H4]histamine).
260
ROBERTS
AND
OATES
‘bF5
w
40
I
410
2
611
M-181-(3x20) M-181-(4x20) 0 50
100
150
200
250
300
350
400
450
500
550
600
650
700
m/2
1. Negative
ion chemical
Assay Method for Urinary
Histamine
FIG.
ionization
To 1 ml of urine is initially added 40 ng of [*H4]histamine and 150 ~1 of 1 N NaOH in 16 X 125-mm disposable glass tubes. After addition of 2 ml of butanol the tubes are vortexed. Heptane (2 ml) is added and the tubes are again vortexed and then centrifuged to facilitate separation of organic and aqueous phases. The organic layer of butanol:heptane is transferred to 12-ml conical centrifuge tubes and 100 ~1 of 1 N HCl is added. After vortexing and centrifugation, the organic layer is aspirated and discarded. The aqueous phase is transferred to a l-ml Reacti-vial and evaporated under a stream of nitrogen, and the residue is dissolved in 30 ~1 of acetonitrile to which is added 10 ~1 of diisopropylethylamine and 15 ~1 of a 25% solution of pentafluorobenzyl bromide in acetonitrile. The derivatization reaction is allowed to proceed for 30 min at room temperature after which excess reagents are evaporated under a stream of nitrogen. The residue is then dissolved in 150 ~1 of 10% Na2COJ to which is added 250 ~1 of methylene chloride followed by vortexing and centrifugation. The upper aqueous layer is aspirated and discarded. To accomplish complete removal of all Na$O,containing water in the vial before evaporating the organic layer, the vial is filled with distilled water without mixing, aspirated, and discarded twice. The methylene chloride is then evap-
mass spectrum
of (CH&F&histamine.
orated to dryness under a stream of nitrogen and the residue is dissolved in 50 ~1 ethyl acetate, and l-2 ~1 is injected for analysis by GC/NICI-MS. Assay Results and Validations
A representative selected ion current chromatogram of m/z 430 and 434 obtained from an analysis as described above of urine collected from a normal individual is depicted in Fig. 2. As seen, the purification procedure outlined satisfactorily removes urinary impurities which could interfere with the accurate determination of the intensity of the m/z 430 and 434 ion peaks from the [*Ho]- and
h em/r430
FILAMENT ON (PEN OFFSET .5 CM)
1
lNJr U&.,,, 0 0
1
1
1 I
1
I
2WIO I
012345676 TIME WN.) FIG.
gram urine
2. Representative selected ion current chromatoof m/z 430 and 434 obtained from the analysis of from a normal volunteer.
CHROMATOGRAPHIC/SPECTROPHOTOMETRIC
HISTAMINE
[*I-I&&amine present. In an occasional urine we have observed an additional impurity ml z 430 peak with a retention time of approximately 1 min. However, this peak is very sharp and returns to baseline before reaching 2 min retention time and thus does not interfere with the histamine m/z 430 peak. Except for this single noninterfering m/z 430 impurity peak seen in an occasional urine, similar selected ion current chromatograms as depicted in Fig. 2 free of interfering impurity peaks have been reproduced in 125 different urines thus far analyzed by this method. Several procedures have been performed to establish the accuracy of this method for the determination of histamine. Initially a standard curve was constructed by adding varying amounts of [‘Ho]histamine to a fixed quantity of 22.81 ng of [*H4]histamine and the ratio of m/z 4301434 was determined directly following derivatization and compared to the expected ratios based on the amounts of [*Ho] and [*H4] present. The results obtained are shown in Fig. 3. The standard deviations for each determination from replicate injections into the mass spectrometer were less than the size of the dots in the figure. As can be seen, this standard curve was linear and accurate
v
I 10-Z lo-’ IO0 IO’ IO2 %
HISTAMINE
I
ADDED
FIG. 3. Standard curve for the analysis of histamine by negative ion chemical ionization mass spectrometry. Varying quantities of unlabeled histamine were added to 22.81 ng of [2H,]histamine as noted on the abcissa. On the ordinate are plotted the actual ratios of m/z 430 (unlabeled histamine) to m/z 434 ([*H,]histamine) obtained with selected ion monitoring analysis.
QUANTIFICATION
261
over a lO,OOO-fold range of varying ratios of m/z 4301434.
We then confirmed that the m/z 430 and 434 peaks observed in the analysis of urine were only generated by the urinary histamine and [*H4]histamine present in the sample rather than in part contributed by potentially coeluting urinary impurities. When urine was analyzed without the addition of [*H4]histamine, no m/z 434 peak was observed. We then treated 1 ml of urine with 16.25 mg of diamine oxidase at 37°C for 30 min. Acetone was then added to precipitate the protein and [*H4]histamine added. Precipitated protein was centrifuged and petroleum ether was added and the organic phase of acetone:petroleum ether aspirated and discarded. The remaining 1 ml aqueous phase was then processed according to the standard assay procedure. The ratio of m/z 4301434 in the diamine oxidase-treated urine was reduced by 96% compared to the ratio obtained with the analysis of 1 ml of the same urine which had not been treated with diamine oxidase. We then processed 5 ml of urine and analyzed the sample by both NICI-MS and EI-MS. The ions monitored for quantification in the EI mass spectrum of (CH2C6F&-histamine were m/z 470 for [*HJhistamine and m/z 474 for [*H,Jhistamine (22). The selected ion monitoring ratio of m/z 4701474 obtained by EIMS differed from the ratio of m/z 4301434 obtained by NICI-MS by only 2.7%. Because the mass spectra of (CH2C6FS),-histamine analyzed by NICI and EI-MS are different and thus the ions monitored for quantification are different, obtaining essentially identical quantitative data when urine was analyzed by both NICI-MS and EI-MS provides convincing evidence that the m/z 430 and 434 peaks observed in the analysis of urine by NICI-MS are only generated by the histamine present rather than in part contributed by coeluting urinary impurities. Experiments were then performed to establish the precision and accuracy of the assay. For the determination of precision, six l-ml aliquots of a single 24-h urine collection were
262
ROBERTS
AND OATES
processed and analyzed individually. For determination of accuracy, to an additional six 1-ml aliquots of the same 24-h urine collection were added 30.39 ng of [‘Halhistamine and the urine subsequently processed and analyzed. The difference between the total amount of histamine measured in these urines to which histamine had been added and the amount measured in the urines without the addition of histamine was then compared to the expected value of 30.39 ng of histamine which had been added. The precision had a coefficient of variation of 2.7%. The accuracy of measuring the histamine added to urine was 97.6%. DISCUSSION
Because of the recognized inaccuracies associated with the widely used radioenzymatic and fluorometric methods for the analysis of histamine in biological fluids, we have developed an accurate physical method for the determination of histamine using GC/NICI-MS. Although the method described in this report pertains to the quantification of urinary histamine, certainly with minor modifications in the purification scheme the procedure could be adapted for the analysis of histamine in a variety of biological fluids. In addition to accuracy this assay is also associated with a remarkable degree of sensitivity. As previously reported, with selected ion monitoring of m/z 430, the lower limit of detection of (CH&F&-histamine injected on column is in the range of 100-500 fg. It is primarily because of the much greater sensitivity associated with NICI-MS compared to EI-MS that we attempted to adapt NICI-MS methodology for the determination of histamine. Although the sensitivity of this assay is not absolutely required for the analysis of urinary histamine in which the normal concentrations present are in the range of several nanograms per milliliter, this degree of sensitivity should easily permit the determination of histamine in small volumes of other biological fluids such as plasma in which the normal levels of histamine are very low.
We routinely process 1 ml of urine for ease and efficiency since the sensitivity of this method would allow the absurd number of well over 200 injections into the mass spectrometer for quantification of each 1-ml sample of urine being analyzed. However, because of this sensitivity, the efficiency of the assay is enhanced due to the fact that time-consuming repeat extractions and careful solvent aspirations to accomplish a high recovery of the histamine through the assay procedure can be avoided. Routinely about 20 samples can be processed and analyzed in 2 days. The sample capacity of this assay is approximately the same as the sample capacity of the radioenzymatic methods of analyses. We are currently in the process of a study rigorously establishing the range of the urinary excretion of histamine in normal volunteers and assessing potential influences of sex, age, and other factors which may alter the levels of urinary histamine excreted. The results of this study will be the subject of a future report. In summary, we have developed a very sensitive, accurate, and efficient method for the quantification of urinary histamine. With minor modifications, it should be possible to adapt this method for the determination of histamine in a variety of biological fluids. The use of this method should contribute to investigations assessing the physiological and pathophysiological participation of histamine in various biological events and diseases. ACKNOWLEDGMENTS This work was supported by Grants GM 15431 and 507-RR05424 from the National Institutes of Health. Dr. Roberts is a Burroughs Wellcome Scholar in Clinical Pharmacology. Dr. Oates is the Joe and Morris Werthan Professor of Investigative Medicine. We would also like to acknowledge the generous contribution of Dr. David Begs of the Hewlett-Packard Corporation who supplied blueprints to allow modification of the HP-5982A mass spectrometer to detect negative ions. The invaluable technical assistance of K. A. Aulsebrook and A. M. Kelleghan was greatly appreciated. REFERENCES 1. Black, J. W., Duncan, W. A. M., Durant, C. J., Ganellin, C. R., and Parsons, E. M. (1972) Nature (London)
236, 385-390.
CHROMATOGRAPHIC/SPECTROPHOTOMETRIC 2. Ishizaka, T., Ishizaka, K., and Tomioka, H. (1972) J. Immunol. 108, 513-520. 3. (1973) in Mechanisms in Allergy: Reagin-mediated Hypersensitivity (Goodfiiend, L., Sehon, A. H., and Orange, R. P., eds.), Dekker, New York. 4. Brown, M. J., Ind, P. W., Causon, R., and Lee, T. H. (1982) J. Allergy Clin. Immunol. 69,20-24. 5. Roberts, L. J., II, Mamey, S. R., Jr., and Oates, J. A. (1979) New Engl. J. Med. 300, 236-238. 6. Roberts, L. J., II, Bloomgarden, Z. T., Mamey, S. R., Jr., Rabin, D., and Oates, J. A. (1983) Gustroenterology 84, 272-275. 7. Roberts, L. J., II, Fields, J. P., and Oates, J. A. (1982) Trans. Assoc. Amer. Phys. 45, 36-4 1. 8. Demis, D. J. (1963) Ann. Int. Med. 59, 194-206. 9. Roberts, L. J., II, Sweetman, B. J., Lewis, R. A., Austen, K. F., and Oates, J. A. (1980) New Engl. J. Med. 303, 1400- 1404. 10. Roberts. L. J., II, Turk, J. W., and Oates, J. A. (1982) Advan. Shock Res. 8, 145-152. I 1. Turk, J., Oates, J. A., and Roberts, L. J. (1983) J. Allergy C/in. Immunol. 71, 189-192.
HISTAMINE
QUANTIFICATION
263
12. Snyder, S. H., Baldessarini, R. J., and Axelrod, J. (1966) J. Pharmacol. Exp. Thu. 153, 544-549. 13. Miller, R. L., McCord, C., Sauda, M., Boume, H. R., and Melmon, K. L. (1970) J. Pharmacol. Exp. Ther. 175,228-234. 14. Beaven, M. A., Jacobsen, S., and Horakova, Z. (1972) Clin. Chim. Acta 37, 91-103. 15. Roberts, M., and Adam, H. M. (1950) Brit. J. Pharmacol. 5, 526-541. 16. Oates, J. A., Marsh, E., and Sjoerdsma, A. (1962) Clin. Chim. Acta 7, 488-497. 17. Siraganian, R. P. (1974) Anal. Biochem. 57,383-394. 18. Gleich, G. J., and Hull, W. M. (1980) J. Allergy Clin. Immunol. 66,295-298. 19. Mita, H.. Yasueda, H., and Shida, T. (1979) J. Chromatogr. 175, 339-342. 20. Mita, H., Yasueda, H., and Shida, T. (1980) J. Chromatogr. 181, 153-159. 21. Mita, H., Yasueda, H., and Shida, T. (1980) J. Chromatogr. 221, l-7. 22. Roberts, L. J., II, and Oates, J. A., submitted for publication.