Analytica Chimica Acta 493 (2003) 159–166
Simple and sensitive fluorimetric liquid chromatographyfor simultaneous analysis of chenodiol and ursodiol in
pharmaceutical formulations�
Ming-Chun Lin, Hsin-Lung Wu∗, Hwang-Shang Kou,Shou-Mei Wu, Su-Hwei Chen
Graduate Institute of Pharmaceutical Sciences, Kaohsiung Medical University, Kaohsiung 807, Taiwan
Received 1 May 2003; received in revised form 8 July 2003; accepted 8 July 2003
Abstract
Chenodiol and ursodiol are diastereomeric bile acids and widely used as anticholelithogenic. A sensitive method was estab-lished for the simultaneous determination of chenodiol and ursodiol by fluorigenic derivatization and liquid chromatography.The analytes were derivatized with 2-(2-naphthoxy)ethyl 2-(piperidino)ethanesulfonate (NOEPES) catalyzed by 18-crown-6ether (18-crown-6) and potassium hydrogen carbonate. The resulting derivatives were analyzed by isocratic HPLC with fluori-metric detection (excitation at 235 nm and emission at 350 nm). The linear range for the analysis of the drugs was 1.0–30.0�Mwith the detection limits (S/N = 3) of 0.4 and 0.2�M, respectively, for chenodiol and ursodiol each based on an injectionvolume of 10�l sample. The method was demonstrated to the analysis of chenodiol in capsules and ursodiol in tablets. Theresults indicate that the method is sensitive and selective.© 2003 Elsevier B.V. All rights reserved.
Keywords: Derivatization; LC; Fluorimetry; Chenodiol; Ursodiol
1. Introduction
Chenodiol and ursodiol are important bile acidswidely used for gall stone dissolution and cholestaticliver diseases, and the combination of chenodiol andursodiol is better for the treatment of cholesterol gall-stone than either drug alone[1,2]. Chenodiol and ur-sodiol are diastereomers; ursodiol can biologically bederived from the epimerization of chenodiol by intesti-nal bacteria[1].
� This work was presented in part at the 62nd InternationalCongress FIP, Nice, France, 31 August–5 September 2002.
∗ Corresponding author. Fax:+886-73159597.E-mail address: [emailprotected] (H.-L. Wu).
Fig. 1 shows that chenodiol and ursodiol have nopractical chromophores for being monitored in thecommon UV range. This limits the sensitivity of con-ventional method for the direct analysis of the drugs.Detection of the drugs at short UV wavelength tendsto be interfered with sample matrix having short UVchromophores.
A variety of methods have been used for the analysisof chenodiol and ursodiol in bulk or in pharmaceuticalpreparations, including alkalimetry[3,4], potentiom-etry [5,6], supercritical fluid chromatography (SFC)[7,8], micellar eletrokinetic chromatography (MEKC)[9], and HPLC[10–12].
The alkalimetry is non-specific for acidic drugs;the potentiometry reported[5,6] is not available for
0003-2670/$ – see front matter © 2003 Elsevier B.V. All rights reserved.doi:10.1016/S0003-2670(03)00871-7
160 M.-C. Lin et al. / Analytica Chimica Acta 493 (2003) 159–166
OO
S
O
ON
OO
C
O
R
H
COOHH3C
H H
HO OH
CH3
CH3
H
H
H
COOHH3C
H H
HO
CH3
CH3
H
H
OH
CDCA UDCA
R-COOH +
NOEPES
Derivative
Fig. 1. Simplified reaction scheme for the derivatization of chenodiol (CDCA) and ursodiol (UDCA). RCOOH as the general species ofchenodiol and ursodiol with carboxyl function and the structures of CDCA and UDCA also shown. NOEPES indicates derivatizing reagentand derivative for the general structure of the CDCA and UDCA derivatives.
the simultaneous analysis of chenodiol and ursodiol;the SFC with UV[7] or with evaporative light scat-tering detection[8] is capable of differentiating thedrugs, but the apparatus of SFC seems not very com-mon as compared to that of HPLC used in routinelaboratory.
The MEKC with indirect UV detection is usedto the analysis of ursodiol in formulation with-out chenodiol [9]. HPLC with refractometry[10]or UV (210 nm) [11] is applied to the analysis ofchenodiol and ursodiol, the sensitivities of the meth-ods are not high; fluorimetric HPLC[12] is usedfor the analysis of chenodiol by derivatization with2-bromoacetyl-6-methoxynaphthalene with increasingsensitivity (a limit of quantitation (LOQ) of 27.5�Mfor chenodiol), but ursodiol is not studied. Sophisti-cated GCMS method[13,14] has been used for theanalysis of the analytes in biosamples. In this work, ahighly sensitive method is described for the simulta-neous analysis of chenodiol and ursodiol based on thefluorigenic derivatization with 2-(2-naphthoxy)ethyl2-(piperidino)ethanesulfonate (NOEPES)[15,16].
2. Experimental
2.1. Chemicals and reagent solutions
Chenodiol, ursodiol and nonanoic acid (used as aninternal standard (IS)) (Sigma, St. Louis, MO, USA),2-(2-naphthoxy)ethyl 2-(piperidino)ethanesulfonate(synthesized at our laboratory)[15], 18-crown-6 ether(18-crown-6) (TCI, Tokyo, Japan), potassium hydro-gencarbonate and sulfuric acid (E. Merck, Darmstadt,Germany), and toluene (Tedia, Fairfield, OH, USA)were used without further treatment. Other chemi-cals were analytical reagent grade. Distilled waterpurified with the Ultrapure R/O water system (Mil-lipore, MA, USA) was used for preparing relatedaqueous solutions. Standard solutions of chenodioland ursodiol (500�M each) were prepared in ace-tonitrile (25 ml) and diluted with toluene for calibra-tion (1.0–30.0�M), and solution of IS (20�M) wasprepared in toluene; standard solutions of NOEPES(1–35 mM) and 18-crown-6 (1–30 mM) were preparedby dissolving the respective compounds in toluene.
M.-C. Lin et al. / Analytica Chimica Acta 493 (2003) 159–166 161
Solution of sulfuric acid (1.0 M) was prepared inwater.
2.2. HPLC conditions
A Waters LC system with a Model 515 HPLCpump, a Model 717 plus autosampler and a Model2475 fluorescence detector and a Millennium chro-matography manager was used. A Merck PurospherStar RP-18e column (250 mm× 4 mm I.D.; 5�M)and a mixed solvent of methanol/water (92:8 (v/v))at a flow rate of 1.0 ml/min were used. The columneluate was monitored at 235 nm for excitation and at350 nm for emission.
2.3. Sample pretreatment
Sample solutions of commercial ursodiol tablets (la-beled amount of 100 mg ursodiol per tablet) and chen-odiol capsules (labeled amount of 250 mg chenodiolper capsule) were prepared as follows: 20 tablets (totalweight about 3000 mg) of ursodiol or 20 capsules (to-tal weight about 5300 mg) of chenodiol content wereweighed and finely powdered in an agate mortar. Anaccurately weighed amount of the powder, equivalentto about 2.5 mg (6.3�mol) each of ursodiol or chen-odiol was transferred to a 25 ml volumetric flask andextracted with acetonitrile 22 ml for 20 min by son-ication, and the solution was diluted to the volumewith acetonitrile. A 0.4 ml aliquot of the solution waspipetted into a 10 ml volumetric flask and diluted tothe volume with toluene. A 200�l of the chenodiol orursodiol solution was subjected to the derivatizationas indicated under derivatization procedure.
2.4. Derivatization procedure
A 200�l aliquot of chenodiol and ursodiol solutionwas added to a 25 ml screw capped test tube contain-ing 100�l nonanoic acid (20�M), 200�l NOEPES(15 mM) in toluene, 100�l of 18-crown-6 in toluene(10 mM) and about 20 mg potassium bicarbonate. Thereactants were shaken at 95◦C for 60 min. After cool-ing, 500�l of the solution were transferred to a testtube and washed with 1.0 ml of H2SO4 (1 M) by vor-tex mixing for 30 s for removal of excess reagent(NOEPES). An aliquot of the acid-washed toluenelayer (100�l) was mixed with an equal volume of
methanol as sample solution for being compatible withthe mobile phase in chromatographic analysis. The re-sulting solution was analyzed by HPLC with an au-tosampler (sample size, 10�l).
3. Results and discussion
In order to optimize the derivatization conditionsfor chenodiol and ursodiol, several parameters affect-ing the derivatization of the analytes were studied.The amounts of chenodiol, ursodiol and IS used forthe study were 6, 6 and 2 nmol, respectively. Theeffects of the parameters on the formation of thederivatives were evaluated by peak areas of the result-ing derivatives. But the peak-area ratios of the druganalytes to the IS were used for the calibration andthe analysis of the drugs in formulation.
3.1. Optimization of the derivatization
In light of the derivatizing reaction for chenodioland ursodiol (Fig. 1), the main parameters affectingthe derivatization were evaluated and optimized asfollows.
The effects of NOEPES in the concentration range(0–35 mM) on the derivatization of chenodiol and ur-sodiol were studied. Plateau formation of the deriva-tives is attainable using NOEPES at concentrations≥15 mM. Derivatization of chenodiol and ursodiol at95◦C for 1 h gave better yield than that derivatiza-tion at 70◦C for 1 h with respective yield (based onpeak-area) of 100 and 28.5. Plateau formation of thederivatives of chenodiol and ursodiol is attainable at95◦C for 1 h, but it cannot be obtained at 70◦C for2.5 h. Optimum amounts of potassium hydrogencar-bonate (0–35 mg tested) and 18-crown-6 (0–30 mMtested) were≥10 mg and 5 mM, respectively. For be-ing easily taken, the amount of potassium bicarbonateused was about 20 mg (18–22 mg).
Fig. 1 shows the derivatization of chenodiol andursodiol. The resulting derivatives having a substi-tuded alkoxy moiety (auxochrome) attached to thefluorophore/chromophore naphthalene system, result-ing in favorable spectrophotometric properties. Inaddition, excess reagent (NOEPES with a tertiaryamino function) can be removed as a water-solubleammonium species by protonation with an acid. The
162 M.-C. Lin et al. / Analytica Chimica Acta 493 (2003) 159–166
optimized conditions for the derivatization was for-mulated inSection 2.4.
3.2. Stability of the derivative
The stabilities of the derivatives of chenodiol,ursodiol and the IS were studied by observing thepeak-area of each analyte for 24 h after derivatization.No significant changes of peak-area from each analytewere found, indicating that the derivatives are suffi-ciently stable for the time required for their analysis.
3.3. Analytical calibration
Based on the optimized derivatization conditions,quantitative applicability of the method for the de-termination of chenodiol and ursodiol was evaluatedat five different concentrations of each drug over therange 1.0–30.0�M. Calibration graphs were estab-lished with y for the peak-area ratios of chenodiolor ursodiol to the IS (nonanoic acid) andx for theconcentration (�M) of respective drug. The linearregression equations obtained are as follows: forchenodiol,y = (0.088± 0.002)x − (0.091± 0.009)with r = 0.999 (n = 5); for ursodiol, y =(0.098± 0.002)x − (0.045± 0.061) with r = 0.999(n = 5). The detection limits (S/N = 3; sample size,10�l) of chenodiol and ursodiol are 0.4�M (4 pmol)and 0.2�M (2 pmol), respectively.
The intra- and inter-day precisions (relative stan-dard deviations (R.S.D.)) of the method were studiedbased on the peak-area ratios for the analysis of chen-odiol and ursodiol each at three levels of 5.0, 15.0and 25.0�M. The analytical results inTable 1indicatethat the percentage R.S.D. and relative errors (R.E.)for the intra-day (n = 5) and inter-day (n = 5) arenumerically all below 2.0.
3.4. Selectivity of the method
The selectivity of the method was tested on theseparation of a standard mixture of bile acids cov-ering chenodiol, cholic acid, dehydroxycholic acid,deoxycholic acid, lithocholic acid and ursodiol eachat 25�M with IS at 20�M. The bile acid mix-ture was derivatized by the derivatization procedure(Section 2.4) with an increment of the derivatizationreagent (15 mM, 200�l). The results indicated that
Table 1Precision and accuracy for the determination of chenodiol andursodiol
Concentrationknowna (�M)
Concentrationfound (�M)
R.S.D.(%)
R.E.b
(%)
ChenodiolIntra-day
5.0 4.9± 0.02 0.4 −2.015.0 15.2± 0.06 0.4 1.325.0 25.3± 0.05 0.2 1.2
Inter-day5.0 4.9± 0.06 1.2 −2.0
15.0 15.1± 0.15 1.0 0.725.0 25.4± 0.34 1.3 1.6
UrsodiolIntra-day
5.0 5.0± 0.02 0.4 0.015.0 15.1± 0.07 0.5 0.725.0 25.0± 0.06 0.2 0.0
Inter-day5.0 5.1± 0.07 1.4 2.0
15.0 15.0± 0.13 0.9 0.025.0 25.3± 0.49 1.9 1.2
a Intra-day assay variance from triplicate analysis of respectivedrug at five interval on a single day and inter-day assay variancefrom triplicate analysis of respective drug on five consecutive days.
b (value found− value known)/value known.
the elution order (with increasing retention time) ofthe bile acids and IS is dehydrocholic acid< ursodiol< cholic acid < IS < chenodiol, deoxycholic acid< lithocholic acid. But baseline resolution cannot beobtained for dehydrocholic acid, ursodiol and cholicacid in addition to the coelute of chenodiol and deoxy-cholic acid. Namely, complete resolution of the bileacids is unattainable by the present isocratic LC witha mobile phase of high methanol content (92% (v/v))(data not shown). A good separation of the bile acidsis obtainable by isocratic LC with a mobile phase oflower organic solvent content (Fig. 3). HPLC withgradient elution could be a reasonable approach forthe analysis of various bile acids in a short run time.
3.5. Structural analysis of the derivatives
The derivative of chenodiol or ursodiol was pre-pared by scaling up the amount of respective drug(0.13 mmol) with similar procedure to that indicatedin Section 2.4without adding the IS. The purifiedderivatives was examined by electron impact MS
M.-C. Lin et al. / Analytica Chimica Acta 493 (2003) 159–166 163
Time (min)
0 2 4 6 8 10 12 14
Time (min)
0 2 4 6 8 10 12 14
1
2
3
(A)
(B)
Fig. 2. Liquid chromatograms for (A) reagent blank and (B) chenodiol and ursodiol standards derivatized with NOEPES. Peak 1: ursodiolderivative; peak 2: IS; and peak 3: chenodiol derivative, LC conditions seeSection 2.
164 M.-C. Lin et al. / Analytica Chimica Acta 493 (2003) 159–166
Time (min)0 5 10 15 20 25 30 80
6
Time (min)0 5 10 15 20 25 30 80
2 1
3
4 5
7
(A)
(B)
Fig. 3. Liquid chromatograms for (A) reagent blank and (B) a standard mixture of six bile acids (each at 25�M with IS at 20�M). Peak1: dehydroxycholic acid; peak 2: cholic acid; peak 3: ursodiol; peak 4: chenodiol; peak 5: deoxycholic acid; peak 6: IS; and peak 7:lithocholic acid. Mobile phase: acetonitrile/water (75:25 (v/v)) at flow rate of 1.5 ml/min.
(JEOL-SX102A mass spectrometer with an ionizationenergy of 70 eV). The mass spectra obtained exhib-ited molecular ion atm/z 562 (M) for the derivativeof chenodiol or ursodiol and a base ion peak atm/z419, equivalent to the molecular ion (M) minus anaphthoxy fragment (C10H7O). The retention times
of peaks 1 and 3 inFig. 2 are respectively identicalto that of the ursodiol and chenodiol derivatives. Awide separation between the derivatives of ursodioland chenodiol is shown inFig. 2. It is interesting toknow that the 3�,7�-dihydroxy species (chenodiolderivative) and the 3�,7�-dihydroxy species (ursodiol
M.-C. Lin et al. / Analytica Chimica Acta 493 (2003) 159–166 165
derivative) mainly differ in the orientation of the hy-droxyl groups leading to a significant difference oftheir affinity to the liphophilic stationary phase (C18)giving better affinity of 3�,7�-dihydroxy species tothe stationary phase (Fig. 3).
3.6. Application
The method was applied to the analysis of chen-odiol in capsules and ursodiol in tablets. The resultsin Table 2indicate that all the analytical values fellwithin 90–110% of the labeled range usually requiredby a pharmacopeia.
The recoveries of the method were briefly studiedby spiking and mixing known amounts of chenodiol orursodiol to suitable amount of finely triturated tabletsor capsule contents. After extraction and dilution asindicated inSection 2.3, the resulting sample solu-tions were prepared to contain known levels of addedchenodiol or ursodiol each at 5.0, 10.0 and 15.0�M
Table 2Assay results for chenodiol and ursodiol in pharmaceutical prod-ucts
Producta Amountfoundb (mg)
R.S.D.(%)
Claimedcontent (%)
ChenodiolA1 245.0± 1.2 0.5 98.0A2 244.1± 2.0 0.8 97.6A3 251.0± 1.9 0.8 100.4A4 237.2± 3.1 1.3 94.9A5 252.2± 2.3 0.9 100.9A6 242.3± 2.5 1.0 96.9
Mean 98.1
S.D. 2.2
UrsodiolB1 98.5± 1.1 1.1 98.5B2 98.2± 1.0 1.0 98.2B3 97.5± 1.8 1.8 97.5B4 100.6± 1.4 1.4 100.6B5 97.6± 0.7 0.7 97.6B6 97.5± 1.5 1.5 97.5
Mean 98.3
S.D. 1.2
a Chenodiol capsules (labeled amount 250 mg per capsule) andursodiol tablets (labeled amount 100 mg per tablet) from two drugcompanies in Taiwan.
b Mean± S.D. (n = 3).
Table 3Analytical results for recovery of chenodiol and ursodiol in phar-maceutical products
Samplea Concentrationspiked (�M)
Concentrationfound (�M)
R.S.D.(%)
Recoveryb
(%)
ChenodiolA 0.0 9.4± 0.1 1.1 −B 5.0 14.4± 0.1 0.7 100.0C 10.0 19.2± 0.3 1.6 98.0D 15.0 24.5± 0.3 1.2 100.7
UrsodiolA 0.0 9.1± 0.1 1.1 −B 5.0 14.2± 0.1 0.7 102.0C 10.0 19.3± 0.2 1.0 102.0D 15.0 24.5± 0.3 1.2 102.7
a Chenodiol capsules and ursodiol tablets from different sourceof drug companies in Taiwan; chenodiol A and ursodiol A sam-ples, respectively prepared from finely powdered chenodiol andursodiol products without addition of the known levels of the drugsexpressed as 0�M spiked.
b Recovery defined as the term of (concentration found−concentration spiked) divided by the concentration found in thezero spiked sample (in sample A for chenodiol or ursodiol); mean±S.D. (n = 3).
in addition to the unknown level of chenodiol or urso-diol existed in the formulated product. The analyticalresults (Table 3) indicate that the average recoveriesfor the analysis of the spiked drug analytes are allabove 98%.
In conclusion, a fluorimetric HPLC method was de-veloped for the analysis of chenodiol and ursodiol inpharmaceutical formulations. The method is sensitiveand selective. Further development of the method forthe trace analysis of various bile acids in biosamplecould be attractive.
Acknowledgements
H.L. Wu is grateful to the National Science Coun-cil (Taiwan) for financial support of this work (NSC91-2113M-037-0221, NSC 92-2113M-037-028).
References
[1] J.G. Hardman, L.E. Limbird, Goodman and Gilman’s: ThePharmacological Basis of Therapeutics, 10th ed., PergamonPress, New York, 2001, p. 1053.
[2] S. Nakagawa, I. Makino, T. Ishizaki, I. Dohl, Lancet 20(1977) 367.
166 M.-C. Lin et al. / Analytica Chimica Acta 493 (2003) 159–166
[3] Chinese Pharmacopeia, fifth ed., Department of Health,Executive Yuan, Taiwan, 2000, p. 1326.
[4] Japanese Pharmacopeia (with explanatory notes), 14th ed.,Hirogawa, Tokyo, 2001, p. C318.
[5] L. Campanella, L. Sorrentino, M. Tomassetti, Analyst 108(1983) 1490.
[6] O.A. De Fuentes, L. Campanella, G. Crescentini, A. Falcioni,M.P. Sammartino, M. Tomassetti, J. Pharm. Biomed. Anal.23 (2000) 89.
[7] S. Scalia, D.E. Grames, J. Pharm. Sci. 82 (1993) 44.[8] V. Villette, B. Herbreteau, M. Lafosse, M. Dreux, J. Liq.
Chromatogr. Relat. Technol. 19 (1996) 1805.[9] V.G. Rodrıguez, S.E. Lucangioli, G.C. Fernández Otero, C.N.
Carducci, J. Pharm. Biomed. Anal. 23 (2000) 375.
[10] US Patent 25/NF 20, United States PharmacopoeialConvention, Rockville, MD, 2002, p. 1787.
[11] S. Scalia, P. Pazzi, M. Guarneri, Anal. Lett. 22 (1989)915.
[12] R. Gatti, V. Cavrini, P. Roveri, Chromatographia 33 (1992)13.
[13] J. Goto, H. Miura, M. Inada, T. Nambara, T. Nagakura, H.Suzuki, J. Chromatogr. 452 (1988) 119.
[14] S. Scalia, R. Scagliarini, P. Pazzi, Arzneim. Forsch. 50 (2000)129.
[15] C.H. Lu, H.L. Wu, S.H. Chen, H.S. Kou, Chromatographia51 (2000) 315.
[16] C.H. Lu, H.L. Wu, S.H. Chen, H.S. Kou, S.M. Wu, J. Agric.Food Chem. 50 (2002) 71.
