This article was downloaded by: [New York University]

On: 19 October 2014, At: 23:13

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK

Analytical Letters
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/lanl20

Determination of Free Bile

Acids in Raw Materials and Bulk
Products by HPLC and GC
a b b
Santo Scalia , Umberto Cova , Marco Fogagnolo
b b
, Silvio Landi & Alessandro Medici
a
Istituto di Chimica Farmaceutica, Università di
Catania , Via A. Doria 6, 1-95100, Catania, Italy
b
Dipartimento di Chimica , Università di Ferrara ,
Via L. Borsari 46, 1-1-44100, Ferrara, Italy
Published online: 20 Aug 2006.

To cite this article: Santo Scalia , Umberto Cova , Marco Fogagnolo , Silvio Landi
& Alessandro Medici (1994) Determination of Free Bile Acids in Raw Materials
and Bulk Products by HPLC and GC, Analytical Letters, 27:9, 1789-1804, DOI:
10.1080/00032719408007436

To link to this article: http://dx.doi.org/10.1080/00032719408007436

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the
information (the “Content”) contained in the publications on our platform.
However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness,
or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the
Content should not be relied upon and should be independently verified with
primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages,
and other liabilities whatsoever or howsoever caused arising directly or
 indirectly in connection with, in relation to or arising out of the use of the
Content.

This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan,
sub-licensing, systematic supply, or distribution in any form to anyone is
expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions
Downloaded by [New York University] at 23:13 19 October 2014
 ANALYTICAL LETTERS, 27(9), 1789-1804 (1994)

DETERMINATION OF FREE BILE ACIDS IN RAW

MATERIALS AND BULK PRODUCTS BY HPLC AND GC
Downloaded by [New York University] at 23:13 19 October 2014

Key Words : Cholic Acid, Deoxycholic acid, HPLC, GC, bovine bile

Santo Scalia,l Umberto Cova,2 Marco Fogagnolo,2 Silvio Landi,2 and

Alessandro Medic@

1lstituto di Chimica Farmaceutica, Universita di Catania, Via A. Doria 6,

1-95100 Catania (Italy); 2Dipartimento di Chimica, Universith di
Ferrara, Via L. Borsari 46, 1-1-44100 Ferrara (Italy)

ABSTRACT
A procedure based on two chromatographic methods with different
selectivities (HPLC and GC) was developed for the quality control assay of
free bile acids in raw materials from animals and bulk products utilized
in the pharmaceutical industry. HPLC was carried out without
preliminary derivatization using an Ultrasphere ODS column with UV
detection at 210 nm and methanol-acetonitrile-acetate buffer as the
mobile phase. For GC, bile acids were converted into their
trifluoroacetyl-hexafluoroisopropyl derivatives and analysed on a SE-52
capillary column with flame-ionization detection. Bile acid levels in

1789

Copyright 0 1994 by Marcel Dekker, Inc.

 1790 SCALIA ET AL.

hydrolysed ox bile, in bulk cholic and deoxycholic acid determined by

HPLC correlated with results obtained by GC, with the exception of the
analytes present in low concentrations (less than 3% w/w) detectable
only by GC. HPLC-UV is the more suitable technique for routine analyses
of free bile acids in pharmaceutical matrices owing to its simplicity and
rapidity. However, because of the low sensitivity and specificity of the UV
detection, the accuracy of the HPLC assay should be verified by comparison
with GC.
Downloaded by [New York University] at 23:13 19 October 2014

INTROD UCTl ON
Chenodeoxycholic acid (CDCA) and especially ursodeoxycholic acid
(UDCA) (see Fig. 1) are widely used as therapeutic agents for the
dissolution of cholesterol gallstoneslf2 and in the therapy of bile reflux
gastritis.3 Moreover, the recent introduction of UDCA for the treatment of
cholestatic liver diseased-6 is emerging as a powerful new approach for
primary biliary cirrhosis,4 primary sclerosing cholangitis5 and cystic
fibrosis.6
To manufacture these two steroids, the pharmaceutical industry
utilizes raw materials with a high bile acid content and available in large
quantities at low costs. In particular, bovine bile, after alkaline
hydrolysis to release the free bile acids, is the most commonly employed
biological material.’ It contains mainly cholic acid (CA) and deoxycholic
acid (DCA) (see Fig. 1) with small amounts of CDCA and traces of
lithocholic acid (LCA)and their oxidized forms.8 Whereas CA is used as
the starting material for the synthesis of CDCA and UDCA,g DCA is
hepatotoxiclo and gastric mucosa damaging3 and it is employed for the
disruption of virus particles in the production of influenza vaccines1 and
in the synthesis of cortisone.12 Generally, the large scale isolation of CA
and DCA from bile is carried out by fractional crystallizations.7 In order
to control these purification processes and to monitor the presence of
 FREE BILE ACIDS 1791
Downloaded by [New York University] at 23:13 19 October 2014

R R1 R2
CA OH H OH
CDCA OH H H
DCA H H OH
LCA H H H
UDCA H OH H

FIG 1. Chemical structures of free bile acids.

potentially toxic related impurities rapid, precise and simple analytical

techniques are required.
Despite the large number of chromatographic m e t h o d ~ l 3 8developed
~~
for the determination of these steroids in human fluids, procedures for
their assay in raw materials, obtained from animals and used in the
pharmaceutical industry, have not yet been reported.
This study was undertaken to investigate the potential use of high-
performance liquid chromatography (HPLC) and gas chromatography (GC)
for the rapid and accurate quantitative analysis of free bile acids in
partially purified raw materials and in the isolated bulk components. The
two techniques were validated and compared in terms of rapidity,
simplicity, specificity, accuracy and precision.
 1792 SCALIA ET AL.

EXPERIMENTAL
Materials
CA, CDCA, UDCA, DCA, LCA and hyodeoxycholic acid (HDCA) were
purchased from Sigma (St. Louis, MO, U.S.A.). Their purity was checked
by HPLC prior to use. Methanol, acetonitrile, water and sodium acetate
were all of HPLC-grade as supplied by Baker (Phillipsburg, NJ, U.S.A.).
Trifluoroacetic anhydride and hexafluoroisopropanol were from Aldrich
(Milan, Italy). All other chemicals were of analytical grade (Farmitalia
Downloaded by [New York University] at 23:13 19 October 2014

Carlo Erba, Milan, Italy). Ox bile raw materials and bulk CA and DCA
samples were obtained from I.C.E. (Reggio Emilia, Italy).

High Performance Liquid Chromatography

The HPLC apparatus consisted of a modular chromatographic system
(Model 880-PU pump, Model 880-02 solvent programmer and Model
875-UV variable-wavelength UVIVis detector: Jasco, Tokyo, Japan)
linked to an injection valve with a 20-pl sample loop (Rheodyne, Cotati,
CAI U.S.A.) and a chromatographic data processor (Chromatopac C-RSA;
Shimadzu, Kyoto, Japan ). The detector was set at 210 nm and 0.02 a.u.f.s.
Sample injections were made with a Model 802 RN syringe (10-pl:
Hamilton, Bonaduz, Switzerland).
Separations were performed on a 5-pm Ultrasphere ODs column
(150~4.6 mm i.d.; Beckman, Berkeley, CA, U.S.A.) fitted with a guard
column (LiChrospher RP-18, 5 pm particles , 4x4 mm i.d.; Merck,
Darmstadt, Germany) and eluted under isocratic conditions with
methanol-acetonitrile-0.02 M aqueous sodium acetate (60:20:20,
v/v/v) adjusted to pH 4.3 with phosphoric acid. The mobile phase was
filtered through type GV filters (0.22-pm; Millipore S.A., Bedford, MA,
U.S.A.) and deaerated on-line by a model ERC-3311 automatic solvent
degasser (Erma, Tokyo, .Japan). Chromatography was performed at
ambient temperature at a flow-rate of 1.1 mlhin.
 FREE BILE ACIDS 1793

The identity of the separated compounds was assigned by co-

chromatography with the authentic substances. Quantification was carried
out by integration of the peak areas using the external standardization
method.

Gas Chromatography
GC analyses were performed with a Mega HRGC gas chromatograph
(Model 5160; Carlo Erba Strumentazione) equipped with a split-splitless
Downloaded by [New York University] at 23:13 19 October 2014

injection system (split ratio, 1:15) and a flame ionization detector. A SE-
52 fused silica capillary column (25 m X 0.32 mm i.d.; Mega s.n.c.,
Milan) was used. The operating conditions were: injector port
temperature, 300" C; column temperature, 240 "C; detector
temperature 300 "C;carrier gas (helium) inlet pressure, 55 KPa. Peaks
were identified by comparison of their retention times with those of
standards. Quantification was on the basis of peak area ratios of analytes to
internal standard (HDCA).

Sample Preparation for HPLC Analysis

Bovine bile raw materials or purified products were accurately
weighted (200-300 mg), transferred to a 50-ml volumetric flask and
dissolved in methanol under ultrasonication (5 min). After dilution to
volume, a portion of the resulting suspension was filtered through 0.45-
pm membrane filters (Millipore) and assayed by HPLC.

Sample Preparation for GC Analysis

Prior to capillary GC, bile acids were converted into their
hexafluoroisopropyl ester trifluoroacetates (HFIP-TFA) as described by
Edenharden and Slemr,15 with minor modifications . In brief, the sample
and the internal standard (HDCA) were accurately weighted and dissolved
in 5 ml of chloroform-methanol (l:l,v/v). An aliquot of this solution
(corresponding to ca. 2 mg of sample) was evaporated under a nitrogen
 1794 SCALIA ET AL.

stream and the residue redissolved in 200 pl of trifluoroacetic anhydride-

hexafluoroisopropanoI (2:l .v/v). The mixture was incubated at 37 "C for
60 min and then reduced to dryness and redissolved in 1 ml of acetonitrile.
A portion of this solution (1 pl) was injected onto the GC column.

RESULTS AND DISCUSSION

Reversed-phase HPLC (RP-HPLC) is the method of choice for routine
quality control analyses of free bile acids in raw materials of animal
Downloaded by [New York University] at 23:13 19 October 2014

origin, since it exhibits several advantages over other chromatographic

techniques. In fact, RP-HPLC circumvents the derivatisation reactions
required in GC14 and, compared to thin-layer chromatography,l 6
achieves improved resolution between isomeric dihydroxy bile acids (e.g.
CDCA, DCA) which is of relevance for the assay of pharmaceutical
matrices.7$17
In a previous paper17 we have reported a method for the determination
of UDCA and CDCA in pharmaceutical dosage forms based on the use of a
C18 column and a binary eluent system (methanol-acetate buffer). In the
course of this study, it was found that addition of a third component to the
mobile phase (methanol-acetonitrile-acetate buffer) produced a more
efficient resolution of the free bile acids present in hydrolysed ox bile
within a shorter analysis time. The separation of the foregoing bile acids
presented in Figure 2 is faster than those typically obtained by the RP-
HPLC procedures reported in the literature.7~18-21 Linearity of the
above HPLC method was established over the range 1-180 pg on-column
weight, with correlation coefficients always higher than 0.997. In none of
the graphs was the intercept with the y-axis significantly different from
zero at the 95% confidence level. The detection limit ranged from 0.7 to
0.8 pg on-column weight with a signal-to-noise ratio of 3:l. The
sensitivity of the method although low, enables direct determination of the
major bile acid constituents present in pharmaceutical raw materials
without any pre-concentration step. Accordingly, sample pretreatment
 FREE BILE ACIDS 1795
Downloaded by [New York University] at 23:13 19 October 2014

I I

10 15 min

FIG 2. RP-HPLC separation of a standard mixture of free bile acids (10

pg of each component). Chromatographic conditions as described
under Experimental. Peaks: 1 = UDCA; 2 = CA; 3 = CDCA; 4 =
DCA; 5 = LCA.

involves simply the addition of methanol to the raw material or bulk

product, dissolution and filtration before injection onto the HPLC column.
This is the simplest possible sample preparation and an internal standard
is not needed.l7*22 Figures 3A and 4A illustrate respectively typical RP-
HPLC traces of CA- and DCA-enriched fractions obtained from the
hydrolysis and crystallization from ethanol of bovine bile. Representative
RP-HPLC chromatograms of purified bulk CA and DCA are shown in
Figures 5A and 6A, respectively.
Since free bile acids have low molar absorptivities, HPLC with
conventional UV detection suffers from limited sensitivity7123 a n d
requires the selection of short UV wavelengths7~17~20
which results in
increased interference from matrix constituents and reagents used during
 1796 SCALIA ET AL.

A
B

2
Downloaded by [New York University] at 23:13 19 October 2014

3 h c A i-
I
I I I
I I I

5 10 15 4 a 12 16 min

FIG 3. Traces of a CA-enriched fraction from hydrolysed ox bile. The

chromatograms were obtained by (A) RP-HPLC or (B) GC.
Conditions as described under Experimental. Peak identification
as in FIG 2. IS = internal standard (HDCA).
 FREE BILE ACIDS 1797

4
Downloaded by [New York University] at 23:13 19 October 2014

3
I
------ I
4, LI --rJI
I I I

5 10 15 4 8 12 min

FIG 4. Traces of DCA-enriched fraction from hydrolysed ox bile. The

chrornatograrns were obtained by (A) RP-HPLC or (B) GC.
Conditions and peak identification as in FIG 2.
 1798 SCALIA ET AL.

A B
Downloaded by [New York University] at 23:13 19 October 2014

I I I
- I I
-
1

5 10 15 4 8 12 min

FIG 5. Chromatograms of bulk CA. Traces were obtained by (A) RP-

HPLC or (B)GC. Conditions and peak identification as in FIG 2.
 FREE BILE ACIDS 1799

A B
Downloaded by [New York University] at 23:13 19 October 2014

. c -L I I
A
3

5 10 15 4 8 12 min

FIG 6. Chromatograms of bulk DCA. Traces were obtained by (A) RP-

HPLC or (B) GC. Conditions and peak identification as in FIG 2.
 1800 SCALIA ET AL.

the purification process. Consequently, problems may arise in the

detection of components present in trace amounts and the identification of
a peak based solely on its retention time on RP-HPLC may not be
sufficient t h u s requiring additional confirmatory methods. The most
rigorous way to validate an HPLC assay is to develop an independent
chromatographic system with an alternative retention mechanism.24
Consequently, the RP-HPLC procedure described above was compared with
a GC technique based on different separation selectivity (as demonstrated
Downloaded by [New York University] at 23:13 19 October 2014

by the elution patterns of Figs. 3-6 ) and detection system (i.e., flame-
ionization detection).
Since bile acids are not sufficiently volatile for GC, a preliminary
derivatization step is required. Among the derivatives reported in the
literature,l4 the hexafluoroisopropyl ester trifluoroacetates were
selected on the ground of their superior stability and since they are easily
and rapidly prepared by a one-step reaction, without artifact
form at io n .1 5
The values of bile acid concentrations determined by systematic
comparison of RP-HPLC and GC assays on various samples of raw
materials from hydrolysed and partially purified bovine bile and on the
final bulk products are presented in Table 1. The results show good
agreement between the two methods based on different chromatographic
and detection principles. However, it must be pointed out that the bile
acids present in small quantities (less than 3% w/w, i.e. CDCA) can be
quantified only by GC-FID (see Table 1) due to the inherent higher
sensitivity of this technique (0.5 ng on-column weight) compared to
HPLC-UV. Although the two methods exhibit comparable precision (see
Table l ) , the relative standard deviation values are higher using RP-
HPLC. This can be traced in part to the use of the internal standard (HDCA)
in the GC procedure to compensate the variability caused by the
derivatization reaction and the injection system. However, the selection of
an appropriate internal standard can be difficult, since this must not be
 Downloaded by [New York University] at 23:13 19 October 2014

TABLE 1.
k4m
Comparison of bile acid concentrations in bile raw materials and in bulk products determined by HPLC and GC zr
m
9
Sample Concentration ( % w/w ) c,
U
rz
CA DCA CDCA

HPLC GC HPLC GC HPLC GC

Bovine bile
 1802 SCALIA ET AL.

naturally present in the analysed biological material. Obviously, HDCA

would not be suitable as an internal standard for the assay of pig bile.25
The chromatograms obtained for the same sample by RP-HPLC (Figs. 3A-
6A) or GC (Figs. 3B-6B ) indicate that in addition to improved efficiency,
the capillary GC method produces traces with a lower background from
endogenous components. This can be ascribed to: (i) the derivatization
reaction which introduces specificity in the sample handling step and (ii)
to the fact that UV detection, unlike FID, is markedly affected by the
Downloaded by [New York University] at 23:13 19 October 2014

presence of impurities with much stronger molar absorptivities at 210

nm than bile acids. Even trace amounts of these impurities would generate
large peaks thus limiting the accuracy and precision of the UV detection
results (Table 1).

CONCLUSIONS
In this study, procedures based on HPLC with UV detection and GC-FID
have been evaluated for the assay of free bile acids in pharmaceutical raw
materials from ox bile and bulk products. The systematic comparison of
the obtained data indicates that both methods give generally equivalent
results. Consequently, the HPLC technique is more convenient for routine
quality control analyses since it allows the direct determination of bile
acids bypassing derivatization reactions. However, for evaluation of low
concentrations of analytes and for the more complex matrices, the higher
sensitivity and specificity of the GC procedure are more suitable. Since
HPLC utilizes a non-specific low-wavelength UV detector, the accuracy of
this type of assay should be verified by comparison with a GC technique.
This approach is more rigorous than those based on the use of alternative
detector^,^ since it provides additional assurance that impurities or
related compounds are not coeluted with the analytes or strongly retained
to a particular phase.
 FREE BILE ACIDS 1803

ACNOWLEDGEMENTS

We thank the M.U.R.S.T (Minister0 dell’UniversitA e delta Ricerca

Scientifica e Tecnologica, Rome) for financial support and the I.C.E.
(Industria Chimica Emiliana, Reggio Ernilia) for raw materials and bulk
products supplied.

REFERENCES
Downloaded by [New York University] at 23:13 19 October 2014

1. J. L. Thistle and A. F. Hofmann, New fngl. J. Med., 289 , 655

(1 973).
2. A. Ward, R. N. Brogden, R. C. Heel, T. M. Speight and G. S. Avery,
Drugs, 2 7 , 95 (1984).
3 . S. Scalia, P. Pazzi, G. Stabellini and M. Guarneri, J. Pharm. Biomed.
Anal. , 6 , 911 (1988).
4. R. Poupon, R. E. Poupon, Y. Calrnus, Y. Chretien, F. Ballet and F.
Darnis,Lancef , 1 , 834 (1987).
5. U. Beuers, U. Spengler, W. Kruis, U. Aydernir, B. Wiebecke, W.
Heldwein, M. Weinzierl, G. R. Pope, T. Sawerbruch and G.
Paumgartner, Hepafology, 16, 707 (1992)
6. C. Galabert, J. C. Montet, D. Lengrand, A. Lecuire, C. Sotta, C.
Figarella, and J. P. Chazalette, J. Pediat., 121, 138. (1992)
7 . A. Roda, R. Gatti, V. Cavrini, C. Cerre and P. Simoni, J. Pharm.
Biomed. Anal., 11, 751 (1993).
8. S. Mullner, R. Hofmann, K. Saar and B. Karbe-Thonges, J. Planar
Chromatogr. , 5, 408 (1992).
9. S. L. Hsia, The Bile Acids, Chemistry, Physiology and Methabolism; P.
P. Nair and D. Kritchevsky, Eds.; Plenum Press: New York, 1971; vol.
1, chapt. 4, p. 109.
10. J. Scholmerich, M. S. Becher, K. Schmidt, R. Schubert, B. Kremer, S.
Feldhaus and W. Gerok, Hepatology , 4 , 661 (1984).
 1804 SCALIA ET AL.

11. A. E. Duxbury, A. W Hampson and J. G. Severs, J. Immunol., 101,

62 (1968).
12. L. F. Fieser and M. Fiser, Steroids; Reinhold Publishing Corporation:
New York, 1959, p. 651.
13. T. Nambara and J. Goto, The Bile Acids, Chemistry, Physiology and
Methabolism; K. D. R. Setchell, P. P. Nair and D. Kritchevsky, Eds.;
Plenum Press: New York, 1988; vol. 4, chapt. 1, pp. 43-64.
14. J. Sjovell, A. M. Lawson, K. D. R. Setchell, Methods in Enzymology; J.
Downloaded by [New York University] at 23:13 19 October 2014

H. Low and H. C. Rillings, Eds.; Academic Press: London, 1985, vol.

111, p. 63.
15. R. Edenharder and J. Slemr, J. Chromatogr. , 222, 1 (1981).
16. G. Szepesi, K. Dudas, A. Pap, Z. Vkch, E. Mincsovics and T. Tyihak, J.
Chromatogr. , 237, 137 (1982).
17. S. Scalia, P. Pazzi and M. Guarneri, Anal. Lett. , 22, 915 (1989).
18. A. D. Reid and P. R. Baker,. J. Chromatogr. , 247, 149 (1982).
19. W. H. Elliott and R. Shaw, Methods in Enzymology; J. H. Law and H. C.
Rilling, Eds.; Academic Press: London, 1985, vol. 111, pp. 59-62.
20. S. Scalia, J. Chromatogr. , 431, 259 (1988).
21. X. X. Zhu and G. R. Brown, Anal. Lett. , 23, 2011 (1990).
22. P. Haefelfinger,. J. Chromatogr. , 218 , 73 (1981).
23. K. D. R. Setchell and C. H. Vestal, J. Lipid Res., 30, 1459 (1989).
24. J. Kirschbaum, S. Perlman, J. Joseph and J. Adamovics, J .
Chromatogr. Sci. , 22, 27 (1984).
25. D. Kritchevsky and P. P. Nair, The Bile Acids, Chemistry, Physiology
and Methabolism; P. P. Nair and D. Kritchevsky, Eds.; Plenum Press:
New York, 1971; vol. 1, chapt. 1, p. 19.

Received December 3, 1993

Accepted March 8 , 1994