The Identification of Bile Acid Methyl Esters by Gas Chromatography Methane Chemical Ionization Mass Spectrometry? Gary M. Muschik,l Lynn H. Wright§ and James A. Schroerll Carcinogenesis Program, NCI Frederick Cancer Research Center, Frederick, Maryland 21701, USA
Methane chemical ionization mass spectrometric data and gas chromatography retention parameters are reported for 31 bile acid methyl esters. The group includes both saturated and unsaturated compounds with up to three keto or hydroxyl groups. Major ions include the quasimolecularions and fragments resulting from the loss of neutral species such as methanol, water and methane. The combination of gas-liquid chromatography retention and mass spectral data is sufficient for an unequivocal distinction among all the bile acids studied. Application of this data to the identification of bile acid metabolites is cited.
major fragment ions. These data can unequivocally distinguish all the bile acid methyl esters studied.
INTRODUCTION A number of workers have reported the characterization of bile acids by mass spectrometry or by gas chromatography mass spectrometry.'-' Gas chromatography mass spectrometry (GLC MS) has allowed the simultaneous separation and identification of bile acids and their metabolites from biological sources. These studies have dealt primarily with electron impact (EI) ionization. Following the introduction of the chemical ionization (CI) technique by Munson and Field,6 it became apparent that this mode of ionization generally resulted in simplified mass spectra. In addition, greater ion abundances in the region of the molecular ion are obtainable relative to EIMS. Szczepanik et d7reported GLC EIMS and isobutane CIMS spectra of 52 acetylated bile acid methyl esters. However, they reported only the CIMS intensities for [M+H]+ and fragments resulting from loss of acetic acid from the quasimolecular ions. We have obtained methane CI mass spectral data for 31 bile acid methyl esters having the structure and numbering shown by compound 1. We report here the relative retention ( a )values for the methyl esters and relative intensities of their quasimolecular ions and
EXPERIMENTAL Mass spectrometry The bile acid methyl esters were analyzed on a Finnigan 3300 CI/EI quadrupole mass spectrometer interfaced with a Varian 1400 series gas chromatograph and Finnigan 6000 data system. Spectra were collected under CI conditions, using methane as both GLC carrier gas and CIMS reagent gas. A source pressure of 1000 p m (uncorrected) of methane was maintained during data collection, which enabled use of an a proximate GLC carrier gas flow rate of 30mlmin-'. The mass spectrometer was operated at a sensitivity of A V-', a filter setting corresponding to 100 amu s-', an emission current of 0.50 mA, and an electron energy of 125 eV. A data acquisition rate of 8 ms amu-' was used, and the base peak of each spectrum was checked for saturation before obtaining a hard copy of the spectrum. Ion intensities were rounded off to the nearest integer. Gas chromatography conditions
1
i. Presented in part at the American Society for Mass Spectrometry, 24th Annual Conference on Mass Spectrometry and Allied Topics, San Diego, California (1976). $ Author to whom correspondence should be addressed. 5 Present address: Environmental Protection Agency, HERL (MD-69) Research Triangle Park, North Carolina 27711, USA. I/Present address: Bruker Instruments, Inc., Manning Park, Billerica, Massachusetts 01821, USA.
Gas chromatographic separations were carried out on a 183 cm (6 ft) X 2 mm i.d. glass column packed with 1% QF-1 on 100/120 mesh HP Chromosorb W. Operating temperatures of the column, interface/transfer lines, and ion source were 240, 250 and 120 "C, respectively. On-column injections were made with a Hamilton 701N 10 pl syringe (sampie sizes ranging from 1-2.5 pl), utilizing the solvent flush technique. Glass distilled solvents (Burdick and Jackson, Muskegon, Michigan) were used throughout. Two internal standards (compounds 7 and 23 in Table 1) were chromatographed with each bile acid ester. The internal standards were used to normalize and minimize variances in the GLC retention data (relative
CCC-0306-042X/79/0006-0266$02.50 266
BIOMEDICAL MASS SPECTROMETRY, VOL. 6, NO. 6, 1979
0Heyden & Son Ltd,
1979
BILE ACID METHYL ESTERS
to compound 23) and to verify the accuracy and performance of the mass spectrometer.
Bile acid esters Diazomethane was used to convert the bile acids to their methyl esters prior to their analysis by GLC CIMS. All bile acids used in this study were kindly provided by D r M. I. Kelsey (NCI Frederick Cancer Research Center, Frederick, Maryland), with the exception of compounds 8,12,17,27,28and 30 listed in Table 1 (obtained from Steroloids Inc., Wilton, New Hampshire).
RESULTS AND DISCUSSION Mass spectral data The two principal features discernible from the methane CI mass spectra of bile acid esters were the formation of quasimolecular ions particularly for the monohydroxy and keto derivatives and the subsequent loss of neutral molecules from these ions (Table 2). Specific observations may be summarized from the data in Tables 1 and 2.
Di- and trihydroxy derivatives (compounds 23-26, 31) were found to yield weak to unobservable quasimolecular ions, presumably due to the facile loss of two or three molecules of water. The loss of water (often yielding the base peak) occurred more readily from [M + HI' than from [M - HI', for all hydroxy bile acids. However, with the exception of compounds 10 and 15 the monohydroxy compounds gave notable quasimolecular ion intensities. This allowed for the distinction of the monohydroxy derivatives (see compounds 3,4,6,8, 9, 11, 16, 18-22). It is interesting that differences in epimer configuration can also be discerned (see compounds 9 vs 10; 15 vs 16;18 vs 21). In contrast, Szczepanik et d7reported that acetylated bile acids methyl esters having the same number of acetate groups and the same carbon skeleton gave under isobutane CI conditions similar ion intensities regardless of acetate position or epimer configuration. Michnowicz and Munson' observed indistinguishable and weak [M + HI' intensities using methane CIMS for the 3alP-hydroxy epimers of a-androstane. In general, it was found that the relative intensity of the quasimolecular ions of a series of compounds increased with an increase in the oxidation state of the compound. The mono-ols 9 and 10 showed less intense quasimolecular ions than either of the analogous enols 6
Table 1. Base peaks and quasimolecular ions of substituted methyl 5P-cholanates Quasimolecular ions' Methyl 50-cholanate substituents
Compound
No.
1 2 3 4 5 6 7 8 9 10 11= 12 13 14 15b 16b 17 18 19" 20 21 22 23 24 25 26 27 28 29 30 31
None 3-Keto, 3a-OH, A7.9'"' 3a-OH, 3-Keto, A4 3p-OH, A5 3-Keto 12a-OH, A3 36-OH (isolithocholic) 3a-OH (lithocholic) 3/3-OH[5al 3,7-diketo 3,12-diketo 3.6-diketo 3a-OH[Et] (lithocholic) 3p-OHIEtI (isolithocholic) 3a-OH, 12a-OH, 3a-OH, 12-Keto 3a-OH, 6-Keto [5aI 12a-OH, 3-Keto 3p-OH, 12-Keto 3a-OH, 7-Keto 3a-OH, 12n-OH (deoxycholic) 3a-OH, 6a-OH 3a-OH, 7a-OH (Chenodeoxycholic) 3a-OH, 7p-OH 3,7,12-triketo 3a-OH, 7.12-diketo 3a-OH, 12a-OH, 7-Keto 3a-OH, 7a-OH, 12-Keto 3a-OH, 7a-OH, 120-OH (cholic)
Mol. wt
374 384 386 386 386 388 388 388 390 390 390 402 402 402 404 404 404 404 404 404 404 404 406 406 406 406 416 418 420 420 422
Base peak
375 385 369 369 387 371 37 1 37 1 373 373 373 403 37 1 403 387 387 369 405 387 387 405 387 37 1 371 37 1 37 1 417 419 403 353 369
LM + HI+
375(100) 385(100) 387(8) 387(30) 387(100) 389(67) 389(27) 389(4) 391(28) 391(-) 391(72) 403(100) 403(90) 403(100) 405( 405(44) 405( -1 405(100) 405(27) 405(17) 405(100) 405(5) 407( - 1 407( - ) 4071407( - ) 417(100) 419(100) 421(4) 421f -1 423( -
a Ions specified as mlz (relative intensity). Designations of ( - 1 indicate relative intensity
Methane chemical ionization mass spectrometric data and gas chromatography retention parameters are reported for 31 bile acid methyl esters. The group includes both saturated and unsaturated compounds with up to three keto or hydroxyl groups. Major ions include the quasimolecularions and fragments resulting from the loss of neutral species such as methanol, water and methane. The combination of gas-liquid chromatography retention and mass spectral data is sufficient for an unequivocal distinction among all the bile acids studied. Application of this data to the identification of bile acid metabolites is cited.
major fragment ions. These data can unequivocally distinguish all the bile acid methyl esters studied.
INTRODUCTION A number of workers have reported the characterization of bile acids by mass spectrometry or by gas chromatography mass spectrometry.'-' Gas chromatography mass spectrometry (GLC MS) has allowed the simultaneous separation and identification of bile acids and their metabolites from biological sources. These studies have dealt primarily with electron impact (EI) ionization. Following the introduction of the chemical ionization (CI) technique by Munson and Field,6 it became apparent that this mode of ionization generally resulted in simplified mass spectra. In addition, greater ion abundances in the region of the molecular ion are obtainable relative to EIMS. Szczepanik et d7reported GLC EIMS and isobutane CIMS spectra of 52 acetylated bile acid methyl esters. However, they reported only the CIMS intensities for [M+H]+ and fragments resulting from loss of acetic acid from the quasimolecular ions. We have obtained methane CI mass spectral data for 31 bile acid methyl esters having the structure and numbering shown by compound 1. We report here the relative retention ( a )values for the methyl esters and relative intensities of their quasimolecular ions and
EXPERIMENTAL Mass spectrometry The bile acid methyl esters were analyzed on a Finnigan 3300 CI/EI quadrupole mass spectrometer interfaced with a Varian 1400 series gas chromatograph and Finnigan 6000 data system. Spectra were collected under CI conditions, using methane as both GLC carrier gas and CIMS reagent gas. A source pressure of 1000 p m (uncorrected) of methane was maintained during data collection, which enabled use of an a proximate GLC carrier gas flow rate of 30mlmin-'. The mass spectrometer was operated at a sensitivity of A V-', a filter setting corresponding to 100 amu s-', an emission current of 0.50 mA, and an electron energy of 125 eV. A data acquisition rate of 8 ms amu-' was used, and the base peak of each spectrum was checked for saturation before obtaining a hard copy of the spectrum. Ion intensities were rounded off to the nearest integer. Gas chromatography conditions
1
i. Presented in part at the American Society for Mass Spectrometry, 24th Annual Conference on Mass Spectrometry and Allied Topics, San Diego, California (1976). $ Author to whom correspondence should be addressed. 5 Present address: Environmental Protection Agency, HERL (MD-69) Research Triangle Park, North Carolina 27711, USA. I/Present address: Bruker Instruments, Inc., Manning Park, Billerica, Massachusetts 01821, USA.
Gas chromatographic separations were carried out on a 183 cm (6 ft) X 2 mm i.d. glass column packed with 1% QF-1 on 100/120 mesh HP Chromosorb W. Operating temperatures of the column, interface/transfer lines, and ion source were 240, 250 and 120 "C, respectively. On-column injections were made with a Hamilton 701N 10 pl syringe (sampie sizes ranging from 1-2.5 pl), utilizing the solvent flush technique. Glass distilled solvents (Burdick and Jackson, Muskegon, Michigan) were used throughout. Two internal standards (compounds 7 and 23 in Table 1) were chromatographed with each bile acid ester. The internal standards were used to normalize and minimize variances in the GLC retention data (relative
CCC-0306-042X/79/0006-0266$02.50 266
BIOMEDICAL MASS SPECTROMETRY, VOL. 6, NO. 6, 1979
0Heyden & Son Ltd,
1979
BILE ACID METHYL ESTERS
to compound 23) and to verify the accuracy and performance of the mass spectrometer.
Bile acid esters Diazomethane was used to convert the bile acids to their methyl esters prior to their analysis by GLC CIMS. All bile acids used in this study were kindly provided by D r M. I. Kelsey (NCI Frederick Cancer Research Center, Frederick, Maryland), with the exception of compounds 8,12,17,27,28and 30 listed in Table 1 (obtained from Steroloids Inc., Wilton, New Hampshire).
RESULTS AND DISCUSSION Mass spectral data The two principal features discernible from the methane CI mass spectra of bile acid esters were the formation of quasimolecular ions particularly for the monohydroxy and keto derivatives and the subsequent loss of neutral molecules from these ions (Table 2). Specific observations may be summarized from the data in Tables 1 and 2.
Di- and trihydroxy derivatives (compounds 23-26, 31) were found to yield weak to unobservable quasimolecular ions, presumably due to the facile loss of two or three molecules of water. The loss of water (often yielding the base peak) occurred more readily from [M + HI' than from [M - HI', for all hydroxy bile acids. However, with the exception of compounds 10 and 15 the monohydroxy compounds gave notable quasimolecular ion intensities. This allowed for the distinction of the monohydroxy derivatives (see compounds 3,4,6,8, 9, 11, 16, 18-22). It is interesting that differences in epimer configuration can also be discerned (see compounds 9 vs 10; 15 vs 16;18 vs 21). In contrast, Szczepanik et d7reported that acetylated bile acids methyl esters having the same number of acetate groups and the same carbon skeleton gave under isobutane CI conditions similar ion intensities regardless of acetate position or epimer configuration. Michnowicz and Munson' observed indistinguishable and weak [M + HI' intensities using methane CIMS for the 3alP-hydroxy epimers of a-androstane. In general, it was found that the relative intensity of the quasimolecular ions of a series of compounds increased with an increase in the oxidation state of the compound. The mono-ols 9 and 10 showed less intense quasimolecular ions than either of the analogous enols 6
Table 1. Base peaks and quasimolecular ions of substituted methyl 5P-cholanates Quasimolecular ions' Methyl 50-cholanate substituents
Compound
No.
1 2 3 4 5 6 7 8 9 10 11= 12 13 14 15b 16b 17 18 19" 20 21 22 23 24 25 26 27 28 29 30 31
None 3-Keto, 3a-OH, A7.9'"' 3a-OH, 3-Keto, A4 3p-OH, A5 3-Keto 12a-OH, A3 36-OH (isolithocholic) 3a-OH (lithocholic) 3/3-OH[5al 3,7-diketo 3,12-diketo 3.6-diketo 3a-OH[Et] (lithocholic) 3p-OHIEtI (isolithocholic) 3a-OH, 12a-OH, 3a-OH, 12-Keto 3a-OH, 6-Keto [5aI 12a-OH, 3-Keto 3p-OH, 12-Keto 3a-OH, 7-Keto 3a-OH, 12n-OH (deoxycholic) 3a-OH, 6a-OH 3a-OH, 7a-OH (Chenodeoxycholic) 3a-OH, 7p-OH 3,7,12-triketo 3a-OH, 7.12-diketo 3a-OH, 12a-OH, 7-Keto 3a-OH, 7a-OH, 12-Keto 3a-OH, 7a-OH, 120-OH (cholic)
Mol. wt
374 384 386 386 386 388 388 388 390 390 390 402 402 402 404 404 404 404 404 404 404 404 406 406 406 406 416 418 420 420 422
Base peak
375 385 369 369 387 371 37 1 37 1 373 373 373 403 37 1 403 387 387 369 405 387 387 405 387 37 1 371 37 1 37 1 417 419 403 353 369
LM + HI+
375(100) 385(100) 387(8) 387(30) 387(100) 389(67) 389(27) 389(4) 391(28) 391(-) 391(72) 403(100) 403(90) 403(100) 405( 405(44) 405( -1 405(100) 405(27) 405(17) 405(100) 405(5) 407( - 1 407( - ) 4071407( - ) 417(100) 419(100) 421(4) 421f -1 423( -
a Ions specified as mlz (relative intensity). Designations of ( - 1 indicate relative intensity
