247
Anal. Chem. 1991, 63, 247-250 (38) Garoff, S.; Stephens, R. 8.; Handon, C. D.;Granilla, T. J.; Sorenson, G. K. Opt. Commun. 1982, 47. 257. (39) LennardJones, J. E. Trans Faraday SOC.1932, 28, 333. (40) Appleyard, E. T. S. Roc. Phys. SOC. 1937, 49, 118. (41) Roward, P.; Bouquet, P. In Progress of Optics: Wolf, E., Ed.; North Holland: Amsterdam, 1965. (42) McCarthy, S. L. J . Vac. Sci. 1978, 73, 135. (43) Maxwe1K;arnett. J. C. Phil. Tr8ns. R. Soc. 1904, 203A, 385. (44) Garoff, S.; Weitz, D. A.; Gramila, T. J.; Hanson, C. D. Opt. Lett. 1981, 6.245. (45) Kotter, 2.; Nitzan, A. J . Phys. Chem. 1982, 86, 2011. (46) Craighead. H. G.: Glass, A. M. Opt. Lett. 1981, 6 , 248.
(47) Garoff, S.; Stephens, R. B.; Hanson, C. D.: Sorenson, G. K. Opt. Commun. 1982, 41, 257. (48) Eagen, C. F Appl. Opt. 1981, 77, 3035. (49) Wetzel, H.; Gericher. H. Chem. Phys. Lett. 1980, 76, 460.
RECEIVED for review October 29,1990. Accepted November 5, 1990. We are grateful for the support of the National Institutes of Health (GM 35108). T.M.C. is the recipient of a NIEHS Career Development Award (ES10069).
Differentiation of Isomeric Conjugated Bile Acids Using Positive-Ion B E Linked Scans Karl V. Wood,*J Yiping Sun,2and Robert G . E l k i d
Departments of Chemistry, Medicinal Chemistry and Pharmacognosy, and Animal Sciences, Purdue University, West Lafayette, Indiana 47907
The main objectlve of this work was to use positlve-Ion fast atom bombardment mass spectrometry (FAB-MS) B / E llnked scan spectra to Investigate the possibiilty of differentiating positional Isomers of various authentic glycine- and taurineconjugated bile aclds. Sodium salts of 14 conjugated bile acids were indlvlduaiiy ionized by FABMS and characterized by scanning simultaneously the magnetic field B and the electric sector field E such that B / E remained constant throughout the scan. The dominant fragment Ions could be related to cleavage of the aliphatic side chain with charge retentlon on the conjugated end of the bile acids. However, fragment ions arlslng from rlng cleavages were also observed and could be used to distinguish the positions of substituent hydroxyl groups. For example, ring cleavage of conjugated dihydroxy bile acids at C-7/C-8 and C-WC-10 permitted the differentiation of chenodeoxycholyltaurine(3a,7a-substltution pattern) from deoxychoiyltaurine (3a,12a-substitutlon pattern) based on the presence of fragment Ions at m / z 388 or m / z 404, which were Indicative of hydroxyl group substitutions at either the 7- or 12-positlons, respectively. I t was concluded that B / E ilnked scans can be used to dlscrlmlnate positional isomers of conjugated bile aclds.
INTRODUCTION Bile acids, which are synthesized from cholesterol in the liver, play an important role in the intestinal solubilization and absorption of lipids ( I , 2). In vertebrate animals, the most common bile acids are derivatives of 5p- or 5a-cholan-24-oic acid, which are N-acyl conjugated with taurine or glycine at the C-24 position ( I , 3). Although placental mammals can synthesize both glyco- and tauro-conjugated bile acids ( 3 4 the remainder of the animal kingdom conjugates bile acids exclusively with taurine (5). With regard to the positions of substituent hydroxyl groups, mammalian bile acids are most commonly hydroxylated in
* To whom all correspondence should be addressed.
*Department of Chemistry. *Department of Medicinal Chemistry and Pharmacognosy. Department of Animal Sciences.
the 3-, 6-, 7-, or 12-positions ( I ) , although C-Chydroxylated bile acids have recently been identified in human fetal gallbladder bile (6, 7). The bile acids of domestic fowl are hydroxylated primarily in the 3-, 7-, and 12-positions (8),although recent work (8-10) has shown that duck bile contains significant amounts of phocaecholyltaurine (PCT), a 23hydroxylated bile acid. In a previous study from our laboratory (8),the biliary bile acid profiles of chickens, turkeys, and ducks were determined by using reversed-phase high-performance liquid chromatography (HPLC) and positive-ion fast atom bombardment mass spectrometry (FAB-MS). In addition, we were able to rapidly (and semiquantitatively) assess the relative proportions of taurine-conjugated monohydroxy-, dihydroxy-, and trihydroxycholanoates present in avian bile extracts by direct FAB-MS (extracts not subjected to prior HPLC analysis). Although direct FAB-MS has potential as a screening procedure, definitive characterization of the bile acids present in a sample subjected to direct FAB-MS would require further analysis by another technique (11). B / E linked scans, obtained by scanning simultaneously the magnetic field B and the electric sector field E such that BIE remains constant throughout the scan, provide a convenient method of determining daughter ions formed from a chosen parent ion (I2,13). Thus, the objective of the present study was to determine if positional isomers of authentic conjugated bile acids could be differentiated by direct positive-ion FAB-MS in conjunction with B I E linked scans.
EXPERIMENTAL SECTION Instrumentation. FAJ3-MS was carried out by using a K r a h MS 50 mass spectrometer (Kratos Analytical Inc., Ramsey, NJ)
equipped with a Kratos FAB ion source. The atom gun used xenon and produced a neutral atom beam at 7-8 keV with an ion current of approximately 1mA. Conventional FAB spectra were obtained by scanning up to m/z 1200 with an INCOS data system. The B / E linked scan spectra of selected ions were recorded with an oscillographicrecorder. The mass scale was calibrated with CsI and glycerol. All spectra were recorded in the positive mode at a resolution of 1000-2000. Reagents. Sodium salts of all conjugated bile acid standards, except for phocaecholyltaurine (PCT) and sulfolithocholyltaurine (SLCT), were purchased from Calbiochem (San Diego, CA). Authentic sodium salts of PCT and SLCT were obtained from
0003-2700/91/0363-0247$02.50/00 1991 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 63, NO. 3, FEBRUARY 1, 1991
248
negative-ion FAl3-MS collision-activated decomposition (141, facilitated structural identification. The spectra of cholyltaurine ( C T [M + Na]+ = 560; Figure 2) and cholylglycine (CG; [M + Na]+ = 510; Figure 3) were representative of all tauro- and glyco-conjugated bile acids, respectively, with few exceptions. The dominant fragment ions in both CT and CG corresponded to cleavage of the aliphatic side chain of cholic acid. This resulted in fragment ions having m / z 211 (cleavagebetween C-22 and C-23; Figure 1,fragmentation A) and m/z 224/225 (cleavage between C-20 and C-22; Figure 1,fragmentation B) for tauro-conjugates and fragment ions having m/z 161 and 174/175 for corresponding glyco-conjugates. The reason for the differences in mass between similarly fragmented tauro- and glyco-conjugated bile acids is that the fragment retaining the charge included the conjugated amino acid. A less abundant fragment ion resulted from side-chain cleavage between C-17 and C-20, producing a fragment ion at m/z 252 and 202 (Figure 1,fragmentation C) for tauro- and glyco-conjugates, respectively. In addition to side-chain cleavages, the linked scan spectra of CT and CG revealed a number of ring cleavages. These included fragment ions for CT a t m/z 280 (cleavage between C-13/C-17 and C-15/C-16; Figure 1,fragmentation D), m / z 293 (cleavage between C-13/C-17 and C-14/C-15; Figure 1, fragmentation E), m / z 334 (cleavage between C-8/C-14 and C-12/C-13; Figure 1, fragmentation F), and m/z 404 (cleavage between C-7/C-8 and C-9/C-10; Figure 1, fragmentation G). Corresponding fragment ions for CG were observed at m / t 230, 243, 284, and 354. The occurrence of fragment ions at m / z 404 and 354 for CT and CG, respectively, representing cleavage between C7/C-8 and C-9/C-10 (Figure 1, fragmentation G), was par-
R, I H
Figure 1. Fragmentation patterns of glycine- and taurine-conjugated bile acids. See Table I for substituent groups R,-Re and X. L. R. Hagey, Department of Medicine, University of California at San Diego, La Jolla. Procedure. Several milligrams of each conjugated bile acid was individually dissolved in approximately0.5 mL of methanol, and 1 WL of each sample was mixed with 1 pL of the glycerol matrix on the stainless steel probe tip.
RESULTS AND DISCUSSION FAB-MS B / E linked scan spectra of each of the 14 conjugated bile acids (sodium salts) listed in Table I were obtained from the sodiated ion. The sodiated ion was chosen because previous work showed that of the two ionic forms, protonated and sodiated, the latter predominated in FAB-MS spectra of taurine-conjugated biliary bile acids from domestic fowl (8). Positive-ion FAB-MS B / E linked scans were chosen for this study because the increased degree of fragmentation of conjugated bile acids obtained, relative to that reported for
CT LINKED SCAN m/z 560
>
t m
z W
+ z 1
r
m /z
340
250
_---
- r
240
I
DCT
m /z
I
3JO
450
m /z
I
411
m/z
m /z
Flgure 2. Positive-ion FAB-MS B / E linked scan spectra of cholyltaurine (CT), phocaecholyltaurine (PCT), chenodeoxycholykaurine (CDCT), ursodeoxycholyltaurine(UDCT), and deoxycholyltaurine (DCT).
ANALYTICAL CHEMISTRY, VOL. 63, NO. 3, FEBRUARY 1, 1991
249
Table I. Names and Structures of Conjugated Bile Acids Analyzed by Positive-Ion Fast Atom Bombardment Mass Spectrometry B / E Linked Scans bile acid
abbreviation
3 (R,)
lithocholyltaurine sulfolithocholyltaurine deoxycholyltaurine chenodeoxycholyltaurine ursodeoxycholyltaurine cholyltaurine phocaecholyltaurine dehydrocholyltaurine lithocholylglycine deoxycholylglycine chenodeoxycholylglycine ursodeoxycholylglycine hyodeoxycholylglycine cholylglycine
LCT SLCT DCT CDCT UDCT CT PCT DHCT LCG DCG CDCG UDCG HDCG CG
a-OH SOIH a-OH a-OH CY-OH a-OH a-OH
=O CY-OH a-OH @-OH a-OH a-OH wOH
position and/or orientation of substituent groups" 6 (Rb) 7 (R,) 12 (Rd) 23 (Re) H H H
H H H H H H H H H H H H a-OH H
CY-OH @-OH a-OH CY-OH
H H wOH H H CY-OH H
=O
=O
H H wOH
H CY-OH H H H a-OH
@-OH H CY-OH
24 (X)
H H H H H H OH (R) H H H H H H H
taurine taurine taurine taurine taurine taurine taurine taurine glycine glycine glycine glycine glycine glycine
"See Figure 1.
I
CG LINKED SCAN m/z 510
L
I
do
I
200
300 I
m/z
430
m/z
338
m/z
UDCGl
I
338
m/z
I
HDCGI
m/z
Figure 3. Positlve-ion FAB-MS B I€ linked scan spectra of cholylglycine (CG), chenodeoxycholylglycine(CDCG), ursodeoxycholylglycine (UDCG), deoxycholylglycine (DCG), and hyodeoxycholylglycine (HDCG).
Figure 4.
Positive-ion FAB-MS B I E linked scan spectra of dehydrocholyltaurine (DHCT) and sulfolithocholyltaurine (SLCT).
ticularly noteworthy because it indicated that hydroxyl group substitution at either C-7 or C-12 could be differentiated. The presence of a fragment ion at m / z 404 (tauro-conjugates) or m/z 354 (glyco-conjugates) was indicative of C-3/C-12 hydroxy substitution while a fragment ion at m / z 388 (tauro-conju-
gates) or 338 (glyco-conjugates) was indicative of C-3/C-6 or C-3/C-7 hydroxy substitution. Thus, deoxycholyltaurine (DCT; 30,12a-OH) could be distinguished from chenodeoxycholyltaurine (CDCT; 3a,7a-OH) or ursodeoxycholyltaurine (UDCT; 3a,7P-OH; Figure 2), and deoxy-
250
ANALYTICAL CHEMISTRY, VOL. 63, NO. 3, FEBRUARY 1, 1991
cholylglycine (DCG; 3a,12a-OH) could be differentiated from chenodeoxycholylglycine (CDCG; 3a,7a-OH), ursodeoxycholylglycine (UDCG; 3cu,7@-OH),or hyodeoxycholylglycine (HDCG; 3tu,6a-OH; Figure 3). HDCG could also be distinguished from other dihydroxy glycine-conjugated bile acids on the basis of a fragment ion at m / z 354 (cleavage between C-6/C-7 and C-9/C-10; Figure 1,fragmentation J) but the lack of a fragment ion at m / z 366 (Figure 3). In addition, fragment ions resulting from ring cleavages between C-5/C-6 and C-9/C-10 (Figure 1,fragmentation H) and between C-l/C-lO and C-4/C-5 (Figure 1,fragmentation I) were observed in the spectra (not shown) of most tauroand glyco-conjugated monohydroxy- and dihydroxycholanoates but were not discernible in the trihydroxycholanoate spectra. The reason for this phenomenon is not known. PCT (3n,7n,23R-OH) had a considerably different fragmentation pattern as compared to CT (Figure 21, a taurineconjugated trihydroxycholanoate in which all of the hydroxyl groups are ring substituted. Predominant PCT fragment ions in the mass range 200-400 were observed at m / z 205,207,297, 299, and 391. None of these ions were seen in the CT spectrum. The presence of a fragment ion a t m / z 205 could possibly be explained by PCT undergoing fragmentation A (Figure 1)with the concomitant replacement of a sodium atom by a hydrogen atom a t the taurine moiety. The ion at m / z 297 was consistent with fragmentation D (Figure l),while the presence of an ion a t m / z 391 was unexplained. The positive-ion FAB-MS B / E linked scan spectra of dehydrocholyltaurine (DHCT; m / z 554) and SLCT ( m / z 630) are shown in Figure 4. These compounds are structurally distinct from those found in Table I and Figures 2 and 3. Characterization of the general fragmentation pattern of these authentic conjugated bile acids could be of value with regard to the identification of unknown compounds that might be present in biological samples. DHCT is identical with CT except that it possesses keto groups, instead of hydroxyl groups, at (2-3, C-7, and C-12. The DHCT B / E linked scan spectrum (Figure 4) was similar to that observed for the tauro-conjugated bile acids listed in Table I. The dominant fragment ions were a t m / z 211 and 224/225. Evidence for the keto functionalities in place of the hydroxyl groups can be seen in the two mass unit shift from m / z 404 to 402 in the fragment ion resulting from cleavage between C-7/C-8 and C-9/C-10 (Figure 1,fragmentation G). This is the fragment ion that includes the C-12 keto group. Similarly, a four mass unit shift was observed from m / z 432 to 428 in the fragment ion resulting from cleavage between C-5/C-6 and C-9/C-10 (Figure 1, fragmentation H). This fragment includes both the C-7 and the C-12 keto groups. SLCT, which has a sulfate group in the 3-position, had the most unique positive-ion FAB-MS B / E linked scan spectrum of all of the bile acids analyzed (Figure 4), because the amino acid conjugate portion of the molecule no longer directed the fragmentation process. Rather, the sulfate group directed the fragmentation pathway as evidenced by the three most intense ions a t m / z 143 (Na2S04H),m / z 165 (Na2S04Na),and m / z
512 (M - NaS04). Two of the other significant fragment ions, m / z 285 and 344, resulted from cleavages observed for other conjugated bile acids, but in the case of SLCT, the charge was retained on the sulfated end of the molecule. These ions resulted from cleavage between C-7/C-8 and C-9/C-10 (Figure 1, fragmentation G) and cleavage between C-12/C-13 and C-8/C-14 (Figure 1,fragmentation F), respectively, with both sodium atoms on the charged fragment ion.
CONCLUSIONS In the context of conjugated bile acid analysis, the major disadvantage of FAB-MS is the inability of the technique to distinguish stereoisomers or positional isomers (11). T o the authors’ knowledge, this is the first report in which positive-ion FAB-MS B / E linked scan spectra were used to differentiate positional isomers of authentic glycine- and taurine-conjugated bile acids. The significance of this work is that linked scan procedures may have potential application as rapid techniques for the identification of bile acid isomers present in biological samples. ACKNOWLEDGMENT We thank David L. Smith for making the facilities of his laboratory available for this study. Appreciation is also extended to Kelly J. Stringham for her excellent technical assistance. Registry No. LCT, 516-90-5; LCT-Na, 6042-32-6; SLCT, 15324-65-9; SLCT-Na, 64936-83-0; DCT, 516-50-7; DCT-Na, 1180-95-6; CDCT, 516-35-8; CDCTSNa, 6009-98-9; UDCT, 14605-22-2;UDCTeNa, 35807-85-3;CT, 81-24-3;CTSNa, 145-42-6; PCT, 85482-99-1;PCT-Na, 130931-36-1;DHCT, 517-37-3;DHCTaNa, 57011-24-2; LCG, 474-74-8; LCG-Na, 24404-83-9; DCG, 360-65-6; DCG-Na, 16409-34-0; CDCG, 640-79-9; CDCG-Na, 16564-43-5;UDCG, 64480-66-6; UDCG-Na,92411-07-9;HDCG, 13042-33-6;HDCGSNa, 38411-84-6;CG, 475-31-0; CGeNa, 863-57-0. LITERATURE CITED (1) Hofmann, A. F. In The Liver: Biology and Pathobiology, 2nd ed.; Arias, I. M., Jakoby, W. B., Popper, H., Schachter, D., Shafritz, D. A., Eds.; Raven Press: New York, 1988; Chapter 32. (2) Borgstrom, B.; Barrowman, J. A,; Lindstrom, M. In Sterols and Ell8 AcMs; Danielsson, ti., Sjovaii, J., Eds.; Elsevier: New York, 1985; Chapter 14. (3) Whiting, M. J. Adv. Clin. Chem. 1986, 2 5 , 169-232. (4) Vessey, D. A. Biochem. J . 1978, 774, 621-626. (5) Elliott, W. H. In Sterols and Bile AcMs; Danielsson. H., Sjovall, J., Eds.; Elsevier: New York, 1985; Chapter 11. ( 6 ) Setchell, K. D. R.; Dumaswala, R.; Colombo, C.; Ronchi, M. J . Bo/. Chem. 1988, 263, 16637-16644. (7) Dumaswala, R.; Setchell, K. D. R.; Zimmer-Nechemias, L.; Iida, T.; Goto, J.; Nambara, T. J . Lipki Res. 1889, 3 0 , 847-856. (8) Elkin, R. G.; Wood, K. V.; Hagey, L. R. Comp. Eiochem. Physiol. E : Comp. B/ocochem. 1990, 9 6 8 , 157-161. (9) Kiinot, J.; Jirsa, M.; Klinotova, E.; Ubik, K.; Protiva, J. Collect. Czech. Chem. Commun. 1988, 51, 1722-1730. (10) Jirsa, M.; Klinot, J.; Klinotova, E.; Ubik, K.; Kucera, K. Comp. Biochem. Physiol. E : Comp. Biochem. 1989, 9 2 8 , 357-380. (11) Setcheil, K. D. R.; Street, J. M. Semin. Liver Dls. 1987, 7, 85-99. (12) Bruins, A. P.; Jennings, K. R.; Evans, S. Int. J . Mass Spectrom. Ion Phys. 1978, 2 6 , 395-404. (13) Jennings, K. R.; Mason, R . S. In Tandem Mass Spectrometry; Mclafferty, F. W., Ed.; John Wiley & Sons: New York, 1983; Chapter 9. (14) Tomer, K. B.; Jensen, N. J.; Gross, M. L.;WhRney, J. Eiomed. Environ. Mass Spectrom. 1986, 13. 265-272.
RECEIVED for review July 23,1990. Accepted October 29,1990.
