Capillary Gas Chromatography/Negative Ion Chemical Ionization Mass Spectrometry for the Quantification of Bile Acids -Including 1B-Hydroxylated and Unsaturated Bile Acids in Serum and Urine Takako Miyara, Noriko Shindo, Masahiko Tohmat, and Kimie Murayamat Division of Biochemical Analysis, Central Laboratory of Medical Sciences, Juntendo University School of Medicine, Bunkyo-ku, Tokyo, 113, Japan
12 bile acids, including lp-hydroxylated and unsaturated bile acids, have been quantified by capillary gas chromatographyhegative ion chemical ionization mass spectrometry, using the trimethylsilyl(TMS) ether derivatives of bile acid pentafluorobenzyl(PFB) esters. The analysis time is 12 min and the minimum measurable amount is 100 fg for each bile acid. Bile acids in 200 pL of serum and 50 pL of urine from healthy human adults were measured. These small sample sizes enhance the practicality of using this method as a screening test for bile acids in the serum and urine of human infants, where small sample size is a major problem.
~
~~~
INTRODUCTION The concentration of bile acids in serum, whether measured during fasting or postprandially, can be a sensitive and specific indicator of hepatobiliary dysfunction (Bloomer et al., 1976; Barnes et al., 1975). A urinary bile acid profile has been used in the diagnosis of disease caused by genetic defects leading to the incomplete metabolism of bile acids (Koopman et al., 1987a), and the defective synthesis of bile acids (Almt5 et al., 1977). Bile acids hydroxylated at the C-1 position, such as 1~,3a,12a-trihydroxycholanoicacid and unsaturated bile acids, such as 3P-hydroxy-5-cholenoic acid, have been found in the urine of patients with liver cirrhosis (AlmC et al., 1977) and hepatobiliary disease (Makino et al., 1971), respectively. Bile acid screening can be a significant diagnostic tool and there is a need for a simple and sensitive screening test for bile acids in biological materials. Four techniques are currently used to measure bile acids: gas chromatography/mass spectrometry (GC/MS) (e.g., Miyazaki et al., 1978; Goto et al., 1987); high performance liquid chromatography (HPLC) (e.g., Batta et al., 1985); enzymatic assay (e.g., Koopman et al., 1987b); and radioimmunoassay (e.g., Miller et al., 1981). The most specific method for the screening of both normal and unusual bile acids is GC/MS (Tohma et al., 1987; Mahara et al., 1987). The TMS ether derivative of the bile acid methyl ester has been used widely for electron ionization mass spectrometry (EIMS) and positive ion chemical ionization mass spectrometry (PICIMS). The detection limit is t Faculty of Pharmaceutical Sciences, Higashi-Nippon-Gakuen University, Ishikari-Tobetsu, Hokkaido, 061-01, Japan. Author to whom correspondence should be addressed. Abbreuiarions: CA, cholic acid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; LCA, litocholic acid; UDCA, ursodeoxycholic acid; HCA, hyocholic acid.
*
greater than 1OOpg for each bile acid with PICIMS (Tohma et al., 1987). The sensitivity of negative ion chemical ionization mass spectrometry (NICIMS) is 1000 times higher than that of PICIMS. The negative ion chemical ionization mass spectra of the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of bile acids were reported by Goto et al. (1987). They preferred the DMES derivatives to the TMS derivatives for trace analysis of bile acids because the former were stable and had fewer fragment ions. They reported the relative retention times of 26 authentic bile acids, excluding the 1P-OH substituted and unsaturated bile acids. However, these were not completely resolved in all cases. This paper presents a quantitative estimation of the PFB ester/TMS ether derivatives of 12 bile acids, including 1P-OH substituted and A5-3P-OHand A5-3P,12a-OH cholenoic acids using capillary GC/NICIMS.
EXPERIMENTAL ~
~~
Reagents. Cholic acid (CA), chenodeoxycholic acid (CDCA), lithocholic acid (LCA), ursodeoxycholic acid (UDCA) and hyocholic acid (HCA) were from Sigma Chemical Co. (St Louis, MO, USA). Deoxycholic acid (DCA) was from Gaschro Kogyo Co. Ltd (Tokyo, Japan). 3P-Hydroxy-5-cholenoic acid (A5-3P-0H)was from Calbiochem-Behring Co. (La Jolla, CA, USA). 3P,12a-Dihydroxy-5-cholenoicacid (A5-3P,12a-OH), Ip,3a,7a,l2a-tetrahydroxy-5P-cholanoicacid (CA-I@-OH), Ip,3a,7a-trihydroxy-5P-cholanoic acid (CDCA-lP-OH), I~,3a,l2a-trihydroxy-5P-cholanoicacid (DCA-lP-OH), 1P,3a-dihydroxy-5P-cholanoic acid (LCA-1p-OH) and [d,]cholic acid (d,-CA), were synthesized (Tohma et al., 1986a,b). 0-Methylhydroxylamine hydrochloride (analytical grade) and the other solvents were from Wako Pure Chemical (Tokyo, Japan). Pentafluorobenzyl bromide (PFB-Br), N, Ndiisopropylethylamine ( DIPEA), dimethylethylsilylimidazole
CCC-O269-3879/90/0056-0060 $2.50
56 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 2, 1990
@ Heyden & Son Limited, 1990
G C / N l C I M S QUANTIFICATION OF BILE ACIDS I N S E R U M A N D U R I N E
10 ng d,-CA in 100 pL methanol was pipetted into a glass tube and dried under a stream of nitrogen. 50pL urine or 200 p L serum and 1 mL H,O were added. The mixture was applied to a Sep-Pak C,* cartridge. The cartridge was washed with 1 0 m L water and eluted with 1 0 m L methanol+ chloroform (9: 1, v/v). The eluate was evaporated under a stream of nitrogen. The residue was dissolved in 2 mL ethanol + acetone (1 :9, v/v), adjusted to pH 1.0 with 2 M HCI, at 38 "C for 1 h, and then the organic solvent was evaporated. The residue was hydrolysed with 4 m L 4~ NaOH methanol (1 : 1, v/v) at 80 "C for 16 h and then concentrated to 2 mL using a concentrator (Model EC-10, Tomy Seiko Co. Ltd, Japan). After eliminating neutral lipids with 2 mL hexane, the residue was acidified to p H 1.0 with 6 M HCI at below 4 "C and applied to a Sep-Pak c18 cartridge. All bile acids were also eluted with 10 mL of methanol+ chloroform and dried under a stream of nitrogen.
(DMES) (Tokyo Kasei Kogyo Co., Tokyo, Japan) and N,Obis(trimethylsi1yl)acetamide (BSA) (Aldrich Chemical Co., Milwaukee, WI, USA) were used. Sep-Pak cartridges were purchased from Waters Associates (Milford, MA, USA). Derivatization. (1) Pentafluorobenzyl (PFB) esters. Bile acid mixtures were dried and treated with 20 p L 35% (v/v) PFB-Br in acetonitrile and 20 p L 10% (v/v) DIPEA in the same solvent at 40 "C for 20 min. Excess reagent was removed using a stream of nitrogen. (2) Trimethylsilyl (TMS) ethers. The residues were dissolved in 100 p L of BSA and held at room temperature for a minimum of 5 min. (3) Dimethylethylsilyl (DMES) ethers. The PFB esters of bile acids were dissolved in 1OOpL of dimethylethylimidazole in pyridine + hexane and allowed to stand at 60°C for 1 h. 1 p L reaction mixture was used for GC/MS analysis. GCIMS. The capillary GC/NICIMS was carried out with a Finnigan 4500 gas chromatograph/mass spectrometer with Incos Data System 2000 (Finnigan-MAT, San Jose, CA, USA). An SPB-1 fused silica capillary column (30 m X 0.25 mm ID, film thickness 0.25pm) from Supelco Japan Ltd (Tokyo, Japan) was used. The column head pressure was 151b/in2. The column was inserted directly into the mass spectrometer so that the end of the column was approximately 20 mm from the ion source. An all glass solid injector (Chrompack, Middelburg, The Netherlands) was used. Hydrogen was used as the carrier gas and methane as the chemical ionization gas (0.61 Torr). The column temperature was 290 "C. The injector and detector temperatures were both 310 "C, and the transfer line was 305 "C. The parameters of the mass spectrometer were: emission current, 0.30 A; electron energy, 110 eV; ionizer temperature, 120 "C; manifold temperature, 84 "C. The bile acids were monitored using (M PFB)- ions.
RESULTS AND DISCUSSION Comparison of the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of 12 bile acids. The NICI mass spectra of the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of 12 bile acids were obtained. Table 1 shows methane NICI mass spectral data for both derivatives. The prominent ions listed corresponded to the carboxylate anion formed by loss of the PFB moiety on both derivatives. Only small percentage losses of 3 xTMSOH, 2 xTMSOH and 1 x TMSOH from the (M-PFB)- ions on the PFB ester/TMS ether derivatives were observed, except for a significant loss of 1 x TMSOH from the (M - PFB)ion of LCA. However, loss of 1 xTMSOH from the ( M - PFB)- ions of all bile acids was observed in one reference (Goto et al., 1987). The PFB ester/DMES ether derivatives had few fragment ions. As a result, the base peak ion of each bile acid is an excellent monitoring ion on a selected ion monitoring (SIM) chromatogram for the quantification of bile aids. Figure 1 shows the SIM chromatograms of the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of 12 bile acids with d,-CA as an internal standard. When the column temperature was 290 "C, the total analysis times of the PFB ester/TMS ether and the PFB ester/DMES deriva-
Calibration curves. Each of five standard mixtures (100 ng, 10 ng, 1 ng, 100 pg and 10 pg) was pipetted into a tube together with 10 ng of the internal standard ([d,]CA), evaporated under a stream of nitrogen and converted to the PFB ester/TMS ether derivatives of bile acids. 1 pL of each reaction mixture was used for GC/MS analysis. Collection of samples. Serum was collected from 6 healthy human adults, aged 22-43 years. The samples were stored at -20 "C before analysis. Morning specimens of urine were collected from 10 healthy human adults, aged 23-58 years and also stored at -20 "C. Purification procedure for biological fluids. The work-up procedure described by Mahara et al. (1987) was used with small modifications. ~~
Table 1: Methane NICI mass spectral data for the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of 12 bile acids Relative intensity
Bile acid
CA CDCA D CA UDCA LCA A5-3P-OH A5-3P,12a-OH CA-lP-OH CDCA-lP-OH D CA- 1P-0 H LCA-1P-OH HCA
[ M - PFB]-
100 100 100 100
100 100 100
100 100 100 100 100
@ Heyden & Son Limited, 1990
TMS ether [ M - PFB [ M - PFB -TMSOH]-2TMSOHI-
3 2 5 2 16 3 1 1 1
2 3 5
5
(YO) DMES ether [ M - PFB [M- PFB
[ M -PFB -3TMSOHl-
2
[M-PFBI-
100 100 100 100 100 100 100 100 100 100 100 100
-DMESOH]-
-2DMESOHl-
[M-PFB -3DMESOHl-
1.29 0.87 0.47
BIOMEDICAL CHROMATOGRAPHY, VOL. 4,
NO. 2. 1990 57
T. MIYARA ET AL.
4
A , I
.
.
1
0.1 hount
of
10 CA
1000
100 injected
( pg
,
luDcA LCA.18.DH
I .
Retention Time
(min:sec)
Figure 1. Selected ion monitoring chromatograms of (a) the PFB ester/TMS ether and (b) the PFB ester/DMES ether derivatives of 12 bile acids. 1 pL of a mixture of 100 pg of each of the 12 authentic bile acids and 100 pg of [d,]CA was injected. A bonded SPB-1 fused silica capillarycolumn 30 m ~ 0 . 2 5mm ID with film thickness 0.25 p m was used. The column temperature was 290 "C. Hydrogen was used as a carrier gas.
tives were 12 and 24min, respectively. The PFB ester/DMES ether derivatives of DCA-1P-OH and HCA were coeluted, but all other bile acids were resolved in 24 min (Fig. l(b)). However, the PFB ester/TMS ether derivatives of the 12 bile acids were well separated in less than 12 min (Fig. l(a)). The calibration curves of the PFB ester/TMS ether derivatives of CA and DCA-1p-OH. The calibration curve of each
bile acid was drawn from the concentration of that bile acid, compared with the peak area of [d7]CA as the internal standard. Figure 2 shows the calibration curves
Amount
of
OCA-1P-OH
injected
( PI )
Figure 2. Calibration curves of (a) CA and (b) DCA-ID-OH, obtained from standard solutions containing 100 pg [d,]CA and increasing amounts of CA and DCA-1p-OH (0.1-1000 pg/pL), respectively.
of CA and DCA-1P-OH. Although the calibration curve of CA is slightly concave because of a small amount of CA in [d,]CA(0.076%), the lower unit of detection was below 100 fg. The calibration curve of DCA-lP-OH, determined by injecting 100 fg more closely, approached a straight line. The other calibration curves were also slightly concave and determined by injecting 100 fg, although they used [d,]CA as an internal standard instead of deuterium derivatives of each bile acid. Reproducibility of the method. The reproducibility of the
method was examined by working up four times the same 50 pL aliquot of normal human urine (Table 2). The use of a high CA level and [d7]CA as an internal
Table 2. Reproducibility of the method Amount Bile acid
CA CDCA DCA UDCA LCA A5-3P-OH A5-3P,12a-OH CA-ID-OH DCA-1P-OH
(ng/mL urine) mean i SD
*
1554 25 189*18 224*17 2051 16 295 1 138*26 13*2 27*2 121 12
*
cv (%) 1.9 10.9 8.6 9.1 5.2 22.1 13.9
1.5
638
-
d,-CA
8.6 11.o
50 KL aliquots of the same urine of a male subject were measured (n =4).
58
$ Y
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 2, 1990
Figure 3. Selected ion monitoring chromatograms of the PFB ester/TMS ether derivatives of bile acids from an extract of human urine (50 pL). (Chromatography details as in the legend to Fig. 1 .) *, Non-estimated bile acid peak.
@ Heyden & Son Limited, 1990
C C / N I C I M S QUANTIFICATION OF BILE ACIDS I N SERUM A N D URINE
Table 3: Urine bile acid concentration in three healthy males and seven females aged 23-58 years Bile acid concentration (ng/mL) CDCA
3p-OH
A5-3p, 12a-OH
CA
UDCA
1P-OH
lp-OH
DCA I@-OH
lD-OH
HCA
79
11 8 74 21 7 40 54 18 173 10
14 6 19 13 n.d. 11 12 30 78 13
3 n.d. 6 28 n.d. 4 n.d. n.d. 7 1
n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.
129 37 123 33 19 169 n.d. 86 349 180
n.d. n.d. n.d. n.d. n.d. n.d. n.d. 12 n.d. n.d.
27 3 19 34 n.d. 9 53 5 n.d. n.d.
A5-
Subject No
Sex
CA
CDCA
1 2 3 4 5 6 7 8 9 10
female female male female male male female female female female
139 72 889 810 134 823 504 47 110 30
28 21 98 65 22 66 255 15 139 10
DCA
LCA
55 24 109 16 48 269 8 239 1967 85
9 7 16 6 13 17 8 64 624 23
40 149 82 43 66 278 26 171 15
LCA
n.d., Not determined.
Table 4. Serum bile acid concentration in two healthy males and four females aged 2 2 4 3 years Bile acid concentration (ng/rnL)
A5-
A5-3p.
No
Sex
CA
CDCA
DCA
LCA
UDCA
3@-OH
12u-OH
CDCA 1P-OH
CA IP-OH
DCA 1P-OH
LCA 1@-OH
HCA
1 2 3 4 5 6
female male male female female female
12 108 83 74 312 108
98 135 258 221 426 171
187 408 834 327 4 12
7 12 71 3 1 2
46 17 51 26 190 22
10 n.d. 6 2 3 6
3 3 3 n.d. 2 2
n.d. n.d. n.d. n.d. n.d. n.d.
n.d. n.d. n.d. n.d. n.d. n.d.
26 19 24 29 3 4
n.d. n.d. n.d. n.d. n.d. n.d.
6 2 1 n.d. n.d. 3
Subject
n.d., Not determined.
standard may cause a small margin of error for CA. Degrees of error for other bile acids, except A5-3P, are generally permissible when [d,]CA is used as an internal standard for them. However, the deuterium labelled derivatives for each bile acid should be used for microanalysis. Quantification of bile acids in normal human urine. Figure 3 shows the SIM chromatograms of bile acids from 50 p L of authentic bile acids and normal human urine of subject three in Table 3. CA, CDCA, DCA, UDCA, AS-3P-OH and DCA-1P-OH were the main urinary bile acids observed. Minor bile acids such as LCA, As3P,12a-OH, CA-1P-OH and HCA were also present. There are many unknown peaks such as m / z 533, m / z 535, and m / z 623. The peak at m / z 533 seems to be a dihydroxycholenoic acid. The peaks at m / z 535 and m / z 623 may be di and trihydroxycholanoic acids, respectively, as there are fragmented ions such as (M - PFB - TMS0H)-, (M - PFB - 2TMSOH)- and (M - PFB - 3TMSOH)- on the same retention time of each unknown peak. Table 3 shows bile acid concentrations in urine from healthy human adults. The total amounts of the bile acids present varied by specimen. CA was predominant
in specimens 1-7. DCA and LCA were higher than CA, and CDCA in specimens from 8-10. DCA-1P-OH was the most prevalent bile acid in specimen 10. Quantification of bile acids in normal human serum. Table 4
shows the bile acids present in the serum of six healthy human adults. Although the total amounts of bile acids varied in each specimen, they tended to fall into two groups. In the first group, in which secondary bile acids were major in specimens 1-4, DCA was predominant. In the second group, where primary bile acids were significant, CDCA was predominant in specimens 5 and 6 . The secondary bile acids, such as DCA and LCA are produced from CA and CDCA, respectively, in the intestine by bacteria. The minor components of bile acids in normal serum are As-3P-OH, A5-3P,12a-OH, DCA-1P-OH and HCA. This method allows the use of samples as small as 200 p L of serum and 50 FL of urine for the quantification of bile acids. Serum samples as small as 50 p L can be used, but reliability is less than that for 200 pL. These small sample sizes enhance the practicality of using this method as a screening test for bile acids in serum and urine of human infants, where sample size is a major problem.
REFERENCES Alme, B., Bremmelgaard, A., Sjovall, J. and Thomassen, P. (1977). J. Lipid Res. 18, 339. Barnes, S., Gallo, G. A., Trash, D. 8. and Morris, J. S. (1975). J. Clin. Parhol. 28, 506.
@ Heyden & Son Limited, 1990
Batta. A. K.. Shefer, S., Batta, M. and Salen, G. (1985). J. Lipid Res. 26, 690. Bloomer. J. R., Allen, R. M. and Klatskin, G. (1976). Arch. hrern. Med. 136, 57.
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 2, 1990 59
T. MIYARA E T A L . Kooprnan, B. J., Wolthers. B. G.. van der Molen, J. C., Nagel, G. T. and Kruizinga, W. (1987a). Biochem. Biophys. Acta, 917, 238. Kooprnan, B. J., van der Molen, J. C. and Wolthers, B. J. (1987b). Clin. Chem. 33, 142. Goto, J., Watanabe, K., Miura, H., Narnbara, T. and lida, T. (1987). J. Chromatogr. 388,379. Mahara, R., Takeshita, H., Kurosawa. T., Ikegawa, S.and Tohma, M. (1987). Anal. Sci. 3, 449. Makino, I.,Sjovall, J., Norman, A. and Strandvike, B. (1971). FEBS Lett. 15, 161. Miller, P., Weiss, S., Cornell, M. and Pockery. J. (1981). Clin. Chem. 27, 1698.
60 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 2, 1990
Miyazaki. H., Ishibashi, M. and Yamashita, K. (1978). Biomed. Mass Spec. 5, 469. Tohrna, M., Mahara. R., Takeshita, H. and Kurosawa, T. (1986a). Steroids. 48. 331. Tohrna, M.. Mahara, R., Takeshita, H., Kurosawa, T. and Ikegawa, S . (1986b). Chem. fharm. But/. 34, 2890. Tohma, M., Takeshita, H., Mahara, R. and Kurosawa, T. (1987). J. Chromatogr. Biomed. Appl. 421, 9.
Received 4 September 1989; accepted (revised) 11 December 1989.
@ Heyden & Son Limited, 1990
12 bile acids, including lp-hydroxylated and unsaturated bile acids, have been quantified by capillary gas chromatographyhegative ion chemical ionization mass spectrometry, using the trimethylsilyl(TMS) ether derivatives of bile acid pentafluorobenzyl(PFB) esters. The analysis time is 12 min and the minimum measurable amount is 100 fg for each bile acid. Bile acids in 200 pL of serum and 50 pL of urine from healthy human adults were measured. These small sample sizes enhance the practicality of using this method as a screening test for bile acids in the serum and urine of human infants, where small sample size is a major problem.
~
~~~
INTRODUCTION The concentration of bile acids in serum, whether measured during fasting or postprandially, can be a sensitive and specific indicator of hepatobiliary dysfunction (Bloomer et al., 1976; Barnes et al., 1975). A urinary bile acid profile has been used in the diagnosis of disease caused by genetic defects leading to the incomplete metabolism of bile acids (Koopman et al., 1987a), and the defective synthesis of bile acids (Almt5 et al., 1977). Bile acids hydroxylated at the C-1 position, such as 1~,3a,12a-trihydroxycholanoicacid and unsaturated bile acids, such as 3P-hydroxy-5-cholenoic acid, have been found in the urine of patients with liver cirrhosis (AlmC et al., 1977) and hepatobiliary disease (Makino et al., 1971), respectively. Bile acid screening can be a significant diagnostic tool and there is a need for a simple and sensitive screening test for bile acids in biological materials. Four techniques are currently used to measure bile acids: gas chromatography/mass spectrometry (GC/MS) (e.g., Miyazaki et al., 1978; Goto et al., 1987); high performance liquid chromatography (HPLC) (e.g., Batta et al., 1985); enzymatic assay (e.g., Koopman et al., 1987b); and radioimmunoassay (e.g., Miller et al., 1981). The most specific method for the screening of both normal and unusual bile acids is GC/MS (Tohma et al., 1987; Mahara et al., 1987). The TMS ether derivative of the bile acid methyl ester has been used widely for electron ionization mass spectrometry (EIMS) and positive ion chemical ionization mass spectrometry (PICIMS). The detection limit is t Faculty of Pharmaceutical Sciences, Higashi-Nippon-Gakuen University, Ishikari-Tobetsu, Hokkaido, 061-01, Japan. Author to whom correspondence should be addressed. Abbreuiarions: CA, cholic acid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; LCA, litocholic acid; UDCA, ursodeoxycholic acid; HCA, hyocholic acid.
*
greater than 1OOpg for each bile acid with PICIMS (Tohma et al., 1987). The sensitivity of negative ion chemical ionization mass spectrometry (NICIMS) is 1000 times higher than that of PICIMS. The negative ion chemical ionization mass spectra of the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of bile acids were reported by Goto et al. (1987). They preferred the DMES derivatives to the TMS derivatives for trace analysis of bile acids because the former were stable and had fewer fragment ions. They reported the relative retention times of 26 authentic bile acids, excluding the 1P-OH substituted and unsaturated bile acids. However, these were not completely resolved in all cases. This paper presents a quantitative estimation of the PFB ester/TMS ether derivatives of 12 bile acids, including 1P-OH substituted and A5-3P-OHand A5-3P,12a-OH cholenoic acids using capillary GC/NICIMS.
EXPERIMENTAL ~
~~
Reagents. Cholic acid (CA), chenodeoxycholic acid (CDCA), lithocholic acid (LCA), ursodeoxycholic acid (UDCA) and hyocholic acid (HCA) were from Sigma Chemical Co. (St Louis, MO, USA). Deoxycholic acid (DCA) was from Gaschro Kogyo Co. Ltd (Tokyo, Japan). 3P-Hydroxy-5-cholenoic acid (A5-3P-0H)was from Calbiochem-Behring Co. (La Jolla, CA, USA). 3P,12a-Dihydroxy-5-cholenoicacid (A5-3P,12a-OH), Ip,3a,7a,l2a-tetrahydroxy-5P-cholanoicacid (CA-I@-OH), Ip,3a,7a-trihydroxy-5P-cholanoic acid (CDCA-lP-OH), I~,3a,l2a-trihydroxy-5P-cholanoicacid (DCA-lP-OH), 1P,3a-dihydroxy-5P-cholanoic acid (LCA-1p-OH) and [d,]cholic acid (d,-CA), were synthesized (Tohma et al., 1986a,b). 0-Methylhydroxylamine hydrochloride (analytical grade) and the other solvents were from Wako Pure Chemical (Tokyo, Japan). Pentafluorobenzyl bromide (PFB-Br), N, Ndiisopropylethylamine ( DIPEA), dimethylethylsilylimidazole
CCC-O269-3879/90/0056-0060 $2.50
56 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 2, 1990
@ Heyden & Son Limited, 1990
G C / N l C I M S QUANTIFICATION OF BILE ACIDS I N S E R U M A N D U R I N E
10 ng d,-CA in 100 pL methanol was pipetted into a glass tube and dried under a stream of nitrogen. 50pL urine or 200 p L serum and 1 mL H,O were added. The mixture was applied to a Sep-Pak C,* cartridge. The cartridge was washed with 1 0 m L water and eluted with 1 0 m L methanol+ chloroform (9: 1, v/v). The eluate was evaporated under a stream of nitrogen. The residue was dissolved in 2 mL ethanol + acetone (1 :9, v/v), adjusted to pH 1.0 with 2 M HCI, at 38 "C for 1 h, and then the organic solvent was evaporated. The residue was hydrolysed with 4 m L 4~ NaOH methanol (1 : 1, v/v) at 80 "C for 16 h and then concentrated to 2 mL using a concentrator (Model EC-10, Tomy Seiko Co. Ltd, Japan). After eliminating neutral lipids with 2 mL hexane, the residue was acidified to p H 1.0 with 6 M HCI at below 4 "C and applied to a Sep-Pak c18 cartridge. All bile acids were also eluted with 10 mL of methanol+ chloroform and dried under a stream of nitrogen.
(DMES) (Tokyo Kasei Kogyo Co., Tokyo, Japan) and N,Obis(trimethylsi1yl)acetamide (BSA) (Aldrich Chemical Co., Milwaukee, WI, USA) were used. Sep-Pak cartridges were purchased from Waters Associates (Milford, MA, USA). Derivatization. (1) Pentafluorobenzyl (PFB) esters. Bile acid mixtures were dried and treated with 20 p L 35% (v/v) PFB-Br in acetonitrile and 20 p L 10% (v/v) DIPEA in the same solvent at 40 "C for 20 min. Excess reagent was removed using a stream of nitrogen. (2) Trimethylsilyl (TMS) ethers. The residues were dissolved in 100 p L of BSA and held at room temperature for a minimum of 5 min. (3) Dimethylethylsilyl (DMES) ethers. The PFB esters of bile acids were dissolved in 1OOpL of dimethylethylimidazole in pyridine + hexane and allowed to stand at 60°C for 1 h. 1 p L reaction mixture was used for GC/MS analysis. GCIMS. The capillary GC/NICIMS was carried out with a Finnigan 4500 gas chromatograph/mass spectrometer with Incos Data System 2000 (Finnigan-MAT, San Jose, CA, USA). An SPB-1 fused silica capillary column (30 m X 0.25 mm ID, film thickness 0.25pm) from Supelco Japan Ltd (Tokyo, Japan) was used. The column head pressure was 151b/in2. The column was inserted directly into the mass spectrometer so that the end of the column was approximately 20 mm from the ion source. An all glass solid injector (Chrompack, Middelburg, The Netherlands) was used. Hydrogen was used as the carrier gas and methane as the chemical ionization gas (0.61 Torr). The column temperature was 290 "C. The injector and detector temperatures were both 310 "C, and the transfer line was 305 "C. The parameters of the mass spectrometer were: emission current, 0.30 A; electron energy, 110 eV; ionizer temperature, 120 "C; manifold temperature, 84 "C. The bile acids were monitored using (M PFB)- ions.
RESULTS AND DISCUSSION Comparison of the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of 12 bile acids. The NICI mass spectra of the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of 12 bile acids were obtained. Table 1 shows methane NICI mass spectral data for both derivatives. The prominent ions listed corresponded to the carboxylate anion formed by loss of the PFB moiety on both derivatives. Only small percentage losses of 3 xTMSOH, 2 xTMSOH and 1 x TMSOH from the (M-PFB)- ions on the PFB ester/TMS ether derivatives were observed, except for a significant loss of 1 x TMSOH from the (M - PFB)ion of LCA. However, loss of 1 xTMSOH from the ( M - PFB)- ions of all bile acids was observed in one reference (Goto et al., 1987). The PFB ester/DMES ether derivatives had few fragment ions. As a result, the base peak ion of each bile acid is an excellent monitoring ion on a selected ion monitoring (SIM) chromatogram for the quantification of bile aids. Figure 1 shows the SIM chromatograms of the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of 12 bile acids with d,-CA as an internal standard. When the column temperature was 290 "C, the total analysis times of the PFB ester/TMS ether and the PFB ester/DMES deriva-
Calibration curves. Each of five standard mixtures (100 ng, 10 ng, 1 ng, 100 pg and 10 pg) was pipetted into a tube together with 10 ng of the internal standard ([d,]CA), evaporated under a stream of nitrogen and converted to the PFB ester/TMS ether derivatives of bile acids. 1 pL of each reaction mixture was used for GC/MS analysis. Collection of samples. Serum was collected from 6 healthy human adults, aged 22-43 years. The samples were stored at -20 "C before analysis. Morning specimens of urine were collected from 10 healthy human adults, aged 23-58 years and also stored at -20 "C. Purification procedure for biological fluids. The work-up procedure described by Mahara et al. (1987) was used with small modifications. ~~
Table 1: Methane NICI mass spectral data for the PFB ester/TMS ether and the PFB ester/DMES ether derivatives of 12 bile acids Relative intensity
Bile acid
CA CDCA D CA UDCA LCA A5-3P-OH A5-3P,12a-OH CA-lP-OH CDCA-lP-OH D CA- 1P-0 H LCA-1P-OH HCA
[ M - PFB]-
100 100 100 100
100 100 100
100 100 100 100 100
@ Heyden & Son Limited, 1990
TMS ether [ M - PFB [ M - PFB -TMSOH]-2TMSOHI-
3 2 5 2 16 3 1 1 1
2 3 5
5
(YO) DMES ether [ M - PFB [M- PFB
[ M -PFB -3TMSOHl-
2
[M-PFBI-
100 100 100 100 100 100 100 100 100 100 100 100
-DMESOH]-
-2DMESOHl-
[M-PFB -3DMESOHl-
1.29 0.87 0.47
BIOMEDICAL CHROMATOGRAPHY, VOL. 4,
NO. 2. 1990 57
T. MIYARA ET AL.
4
A , I
.
.
1
0.1 hount
of
10 CA
1000
100 injected
( pg
,
luDcA LCA.18.DH
I .
Retention Time
(min:sec)
Figure 1. Selected ion monitoring chromatograms of (a) the PFB ester/TMS ether and (b) the PFB ester/DMES ether derivatives of 12 bile acids. 1 pL of a mixture of 100 pg of each of the 12 authentic bile acids and 100 pg of [d,]CA was injected. A bonded SPB-1 fused silica capillarycolumn 30 m ~ 0 . 2 5mm ID with film thickness 0.25 p m was used. The column temperature was 290 "C. Hydrogen was used as a carrier gas.
tives were 12 and 24min, respectively. The PFB ester/DMES ether derivatives of DCA-1P-OH and HCA were coeluted, but all other bile acids were resolved in 24 min (Fig. l(b)). However, the PFB ester/TMS ether derivatives of the 12 bile acids were well separated in less than 12 min (Fig. l(a)). The calibration curves of the PFB ester/TMS ether derivatives of CA and DCA-1p-OH. The calibration curve of each
bile acid was drawn from the concentration of that bile acid, compared with the peak area of [d7]CA as the internal standard. Figure 2 shows the calibration curves
Amount
of
OCA-1P-OH
injected
( PI )
Figure 2. Calibration curves of (a) CA and (b) DCA-ID-OH, obtained from standard solutions containing 100 pg [d,]CA and increasing amounts of CA and DCA-1p-OH (0.1-1000 pg/pL), respectively.
of CA and DCA-1P-OH. Although the calibration curve of CA is slightly concave because of a small amount of CA in [d,]CA(0.076%), the lower unit of detection was below 100 fg. The calibration curve of DCA-lP-OH, determined by injecting 100 fg more closely, approached a straight line. The other calibration curves were also slightly concave and determined by injecting 100 fg, although they used [d,]CA as an internal standard instead of deuterium derivatives of each bile acid. Reproducibility of the method. The reproducibility of the
method was examined by working up four times the same 50 pL aliquot of normal human urine (Table 2). The use of a high CA level and [d7]CA as an internal
Table 2. Reproducibility of the method Amount Bile acid
CA CDCA DCA UDCA LCA A5-3P-OH A5-3P,12a-OH CA-ID-OH DCA-1P-OH
(ng/mL urine) mean i SD
*
1554 25 189*18 224*17 2051 16 295 1 138*26 13*2 27*2 121 12
*
cv (%) 1.9 10.9 8.6 9.1 5.2 22.1 13.9
1.5
638
-
d,-CA
8.6 11.o
50 KL aliquots of the same urine of a male subject were measured (n =4).
58
$ Y
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 2, 1990
Figure 3. Selected ion monitoring chromatograms of the PFB ester/TMS ether derivatives of bile acids from an extract of human urine (50 pL). (Chromatography details as in the legend to Fig. 1 .) *, Non-estimated bile acid peak.
@ Heyden & Son Limited, 1990
C C / N I C I M S QUANTIFICATION OF BILE ACIDS I N SERUM A N D URINE
Table 3: Urine bile acid concentration in three healthy males and seven females aged 23-58 years Bile acid concentration (ng/mL) CDCA
3p-OH
A5-3p, 12a-OH
CA
UDCA
1P-OH
lp-OH
DCA I@-OH
lD-OH
HCA
79
11 8 74 21 7 40 54 18 173 10
14 6 19 13 n.d. 11 12 30 78 13
3 n.d. 6 28 n.d. 4 n.d. n.d. 7 1
n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.
129 37 123 33 19 169 n.d. 86 349 180
n.d. n.d. n.d. n.d. n.d. n.d. n.d. 12 n.d. n.d.
27 3 19 34 n.d. 9 53 5 n.d. n.d.
A5-
Subject No
Sex
CA
CDCA
1 2 3 4 5 6 7 8 9 10
female female male female male male female female female female
139 72 889 810 134 823 504 47 110 30
28 21 98 65 22 66 255 15 139 10
DCA
LCA
55 24 109 16 48 269 8 239 1967 85
9 7 16 6 13 17 8 64 624 23
40 149 82 43 66 278 26 171 15
LCA
n.d., Not determined.
Table 4. Serum bile acid concentration in two healthy males and four females aged 2 2 4 3 years Bile acid concentration (ng/rnL)
A5-
A5-3p.
No
Sex
CA
CDCA
DCA
LCA
UDCA
3@-OH
12u-OH
CDCA 1P-OH
CA IP-OH
DCA 1P-OH
LCA 1@-OH
HCA
1 2 3 4 5 6
female male male female female female
12 108 83 74 312 108
98 135 258 221 426 171
187 408 834 327 4 12
7 12 71 3 1 2
46 17 51 26 190 22
10 n.d. 6 2 3 6
3 3 3 n.d. 2 2
n.d. n.d. n.d. n.d. n.d. n.d.
n.d. n.d. n.d. n.d. n.d. n.d.
26 19 24 29 3 4
n.d. n.d. n.d. n.d. n.d. n.d.
6 2 1 n.d. n.d. 3
Subject
n.d., Not determined.
standard may cause a small margin of error for CA. Degrees of error for other bile acids, except A5-3P, are generally permissible when [d,]CA is used as an internal standard for them. However, the deuterium labelled derivatives for each bile acid should be used for microanalysis. Quantification of bile acids in normal human urine. Figure 3 shows the SIM chromatograms of bile acids from 50 p L of authentic bile acids and normal human urine of subject three in Table 3. CA, CDCA, DCA, UDCA, AS-3P-OH and DCA-1P-OH were the main urinary bile acids observed. Minor bile acids such as LCA, As3P,12a-OH, CA-1P-OH and HCA were also present. There are many unknown peaks such as m / z 533, m / z 535, and m / z 623. The peak at m / z 533 seems to be a dihydroxycholenoic acid. The peaks at m / z 535 and m / z 623 may be di and trihydroxycholanoic acids, respectively, as there are fragmented ions such as (M - PFB - TMS0H)-, (M - PFB - 2TMSOH)- and (M - PFB - 3TMSOH)- on the same retention time of each unknown peak. Table 3 shows bile acid concentrations in urine from healthy human adults. The total amounts of the bile acids present varied by specimen. CA was predominant
in specimens 1-7. DCA and LCA were higher than CA, and CDCA in specimens from 8-10. DCA-1P-OH was the most prevalent bile acid in specimen 10. Quantification of bile acids in normal human serum. Table 4
shows the bile acids present in the serum of six healthy human adults. Although the total amounts of bile acids varied in each specimen, they tended to fall into two groups. In the first group, in which secondary bile acids were major in specimens 1-4, DCA was predominant. In the second group, where primary bile acids were significant, CDCA was predominant in specimens 5 and 6 . The secondary bile acids, such as DCA and LCA are produced from CA and CDCA, respectively, in the intestine by bacteria. The minor components of bile acids in normal serum are As-3P-OH, A5-3P,12a-OH, DCA-1P-OH and HCA. This method allows the use of samples as small as 200 p L of serum and 50 FL of urine for the quantification of bile acids. Serum samples as small as 50 p L can be used, but reliability is less than that for 200 pL. These small sample sizes enhance the practicality of using this method as a screening test for bile acids in serum and urine of human infants, where sample size is a major problem.
REFERENCES Alme, B., Bremmelgaard, A., Sjovall, J. and Thomassen, P. (1977). J. Lipid Res. 18, 339. Barnes, S., Gallo, G. A., Trash, D. 8. and Morris, J. S. (1975). J. Clin. Parhol. 28, 506.
@ Heyden & Son Limited, 1990
Batta. A. K.. Shefer, S., Batta, M. and Salen, G. (1985). J. Lipid Res. 26, 690. Bloomer. J. R., Allen, R. M. and Klatskin, G. (1976). Arch. hrern. Med. 136, 57.
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 2, 1990 59
T. MIYARA E T A L . Kooprnan, B. J., Wolthers. B. G.. van der Molen, J. C., Nagel, G. T. and Kruizinga, W. (1987a). Biochem. Biophys. Acta, 917, 238. Kooprnan, B. J., van der Molen, J. C. and Wolthers, B. J. (1987b). Clin. Chem. 33, 142. Goto, J., Watanabe, K., Miura, H., Narnbara, T. and lida, T. (1987). J. Chromatogr. 388,379. Mahara, R., Takeshita, H., Kurosawa. T., Ikegawa, S.and Tohma, M. (1987). Anal. Sci. 3, 449. Makino, I.,Sjovall, J., Norman, A. and Strandvike, B. (1971). FEBS Lett. 15, 161. Miller, P., Weiss, S., Cornell, M. and Pockery. J. (1981). Clin. Chem. 27, 1698.
60 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 2, 1990
Miyazaki. H., Ishibashi, M. and Yamashita, K. (1978). Biomed. Mass Spec. 5, 469. Tohrna, M., Mahara. R., Takeshita, H. and Kurosawa, T. (1986a). Steroids. 48. 331. Tohrna, M.. Mahara, R., Takeshita, H., Kurosawa, T. and Ikegawa, S . (1986b). Chem. fharm. But/. 34, 2890. Tohma, M., Takeshita, H., Mahara, R. and Kurosawa, T. (1987). J. Chromatogr. Biomed. Appl. 421, 9.
Received 4 September 1989; accepted (revised) 11 December 1989.
@ Heyden & Son Limited, 1990
