Evaluation of 2-(2-thiophenecarboxy)benzoic acid and related active metabolites in biological samples.

This paper describes a quantitative assay of 2-(2-thiophenecarboxy)benzoic acid and its systemic metabolites, namely salicylic acid and thiophenecarboxylic acid, by high-performance liquid chromatography with ultraviolet detection at 254 nm. Analytes were extracted from acidified samples with tert.-butylmethyl ether and separated with an RP-18 column (150 mm x 3 mm I.D., 5 microns particle size) with a mixture of 0.01 M potassium phosphate and methanol (70:30, v/v) at pH 3.1. The method proved to have the validation required for pharmacokinetic investigations in animals and humans.     

Validation of liquid chromatography and liquid chromatography/tandem mass spectrometry methods for the determination of etoricoxib in pharmaceutical formulations

Reversed-phase liquid chromatography (LC) and LC/tandem mass spectrometry (LC/MS/MS) methods were developed and validated for the determination of etoricoxib in pharmaceutical dosage forms. The LC method was performed by reversed-phase chromatography on a Synergi fusion C18 column (150 x 4.6 mm id) maintained at ambient temperature. The mobile phase consisted of 0.01 M phosphoric acid, pH 3.0-acetonitrile (62 + 38, v/v) at a flow rate of 1.0 mL/min, and photodiode array detection at 234 nm was used. The chromatographic separation was obtained within 7.0 min, and calibration curves were linear in the concentration range of 0.02-150 microg/mL. The LC/MS/MS method was performed on a Luna C18 column (50 x 3.0 mm id). The mobile phase consisted of acetonitrile-water (95 + 5)-0.1% acetic acid (90 + 10, v/v). Detection was performed by positive electrospray ionization in the multiple reaction monitoring mode, monitoring the transitions 359.3 > 280.0 and 332.0 > 95.0 for etoricoxib and piroxicam (internal standard), respectively. The chromatographic separation was obtained within 2.0 min, and calibration curves were linear in the concentration range of 1-5000 ng/mL. Validation parameters, such as specificity, linearity, precision, accuracy, and robustness, were evaluated, which gave results within the acceptable range for both methods. Moreover, the proposed methods were successfully applied for routine quality control analysis of pharmaceutical products and showed significant correlation (r = 0.9999) of the results.     

Advantages of pentafluorophenylpropyl stationary phase over conventional C18 stationary phase--application to analysis of triamcinolone acetonide

A pentafluorophenylpropyl (PFPP) stationary phase was for the first time tested for the simultaneous determination of triamcinolone acetonide, its degradation product triamcinolone and two preservatives, methylparaben, and propylparaben. A new simple isocratic reversed phase HPLC method with UV detection, using estradiol hemihydrate as an internal standard, has been developed and validated. Chromatography was performed on a Discovery HS F5 column (150 mm x 4.6 mm, 5 microm) using a binary mobile phase composed of acetonitrile and water 45:55 (v:v). The flow-rate was 0.6 mL/min, the column temperature 25 degrees C and the UV detection was accomplished at 240 nm. The chromatography results using PFPP stationary phase were compared with those obtained using conventional C18 columns.     

High throughput screening of active pharmaceutical ingredients by UPLC

Ultra performance LC (UPLC) was evaluated as an efficient screening approach to facilitate method development for drug candidates. Three stationary phases were screened: C-18, phenyl, and Shield RP 18 with column dimensions of 150 mm x 2.1 mm, 1.7 microm, which should theoretically generate 35,000 plates or 175% of the typical column plate count of a conventional 250 mm x 4.6 mm, 5 microm particle column. Thirteen different active pharmaceutical ingredients (APIs) were screened using this column set with a standardized mobile-phase gradient. The UPLC method selectivity results were compared to those obtained for these compounds via methods developed through laborious trial and error screening experiments using numerous conventional HPLC mobile and stationary phases. Peak capacity was compared for columns packed with 5 microm particles and columns packed with 1.7 microm particles. The impurities screened by UPLC were confirmed by LC/MS. The results demonstrate that simple, high efficiency UPLC gradients are a feasible and productive alternative to more conventional multiparametric chromatographic screening approaches for many compounds in the early stages of drug development.     

Simultaneous determination of caffeine and its primary demethylated metabolites in human plasma by high-performance liquid chromatography.

An improved high-performance liquid chromatographic method for the simultaneous determination of caffeine and its three primary metabolites (theophylline, theobromine and paraxanthine) in human plasma is described. The four substances were separated on a reversed-phase column (5 microns TSK gel ODS-80TM, 150 mm x 4.6 mm I.D.) by use of the mobile phase methanol-0.1 M NaH2PO4 (30:70, v/v) with a flow-rate of 0.8 ml/min. Absorbance was monitored at 274 nm. The detection limit was 5 ng/ml for theobromine and caffeine and 10 ng/ml for paraxanthine and theophylline. The linearity and reproducibility were sufficient for drug monitoring of caffeine and its primary methylxanthines.     

Use of on-line hydrogen/deuterium exchange to facilitate metabolite identification

Biotransformation studies performed on an investigational compound (I, represented by R1-CH(NH(2))-CO-N(R2)-CH(2)-S-R3) led to the identification of five metabolites (M1-M5). Based on LC/MS (liquid chromatography/mass spectrometry) analysis which included the use of H(2)O and D(2)O in the mobile phases, they were identified as the sulfoxide (M1), sulfone (M2), carbamoyl glucuronide (M3), N-glucuronide (M4), and N-glucoside (M5) metabolites, respectively. The structure of M3, a less commonly seen carbamoyl glucuronide metabolite, was established using on-line H/D (hydrogen/deuterium) exchange experiments conducted by LC/MS. H/D exchange experiments were also used to distinguish the S-oxidation structures of M1 and M2 from hydroxylation. Herein, the application of deuterium oxide as the LC/MS mobile phase for structural elucidation of drug metabolites in biological matrices is demonstrated.     

Simultaneous determination of triclabendazole and its sulphoxide and sulphone metabolites in bovine milk by high-performance liquid chromatography

A simple method has been developed for the simultaneous determination of triclabendazole and its metabolites (sulphoxide and sulphone) in bovine milk by reversed-phase high-performance liquid chromatography (HPLC). A milk sample was homogenized with sodium sulfate anhydrous and acetonitrile, and centrifuged. The supernatant was isolated, rinsed with n-hexane saturated with acetonitrile, and evaporated. The residue was dissolved with 0.1 M potassium dihydrogenphosphate, and 0.1 M sodium hydrogencarbonate, and then cleaned up on a Bond Elut C18 cartridge. The three compounds were separated on a Capcell Pak C18 UG 120 (5 microm, 150x4.6 mm I.D.) column and determined by UV detection at 295 nm. The mobile phase was acetonitrile-0.05 M ammonium acetate (50:50), and the flow-rate was 0.8 ml/min at 40 degrees C. The mean recoveries (n=4) were 89.1-95.0% with a relative standard deviation of 1.1-2.6%. The detection limits were 0.004-0.006 microg/g in milk. The proposed method was used to monitor raw milk samples for the market, and applied to the analysis of milk samples from 10 cows which had been administered with triclabendazole to control the liver fluke. The confirmation of the triclabendazole and its metabolites in the above milk sample has been achieved by electrospray LC-MS for the first time.     

Quantitative determination of coenyzme Q10 by liquid chromatography and liquid chromatography/mass spectrometry in dairy products

The dietary sources of CoQ10 and the evaluation of CoQ10 in dairy products were characterized. For quantitation of CoQ10 in food samples, 2 liquid chromatography (LC) methods with UV and mass spectrometry (MS) detections were developed. LC with UV detection was performed at 25 degrees C on a Hyperclone ODS 5 microm 150 x 4.6 mm column with mobile phase consisting of methanol-ethanol-2-propanol (70 + 15 + 15, v/v/v). Flow rate was 1.0 mL/min. Retention time of CoQ10 was 10.9 +/- 0.1 min. The method was sensitive [limit of detection (LOD) = 0.2 mg/kg], reproducible [relative standard deviation (RSD) = 3:0%), and linear up to 25 mg/kg (R > 0.999). LC/MS analysis was performed on a LUNA C18 3 microm, 150 x 4.6 mm column, using mobile phase consisting of ethanol-dioxane-acetic acid (9 + 1 + 0.01, v/v/v), flow rate was 0.6 mL/min, and the retention time of CoQ10 was 4.1 +/- 0.1 min. Identification and quantitation were performed with a Finnigan-LCQ mass detector in positive atmospheric pressure chemical ionization mode. Mass spectra were obtained in selected-ion monitoring mode; molecular mass (M+H)+ m/z 863.4 +/- 1 was used for quantitative determination. MS detection is more sensitive than UV detection (LOD = 0.1 mg/kg), less reproducible (RSD = 4.0%), and linear in selected range. Analytical recoveries are 75-90% and depend on the ratio between the amount of fat in the matrix and the concentration of CoQ10 in the sample. Some soybean milk products were analyzed together with different cow, goat, and sheep milk products. Concentrations obtained with LC and LC/MS were compared with a few accessible results available from the literature. Concentrations varied from 0 ppm in soybean milk to nearly 2 ppm in fresh milk from local farms.     

Determination of chlorpyrifos and its metabolites in rat brain tissue using coupled-column liquid chromatography/electrospray ionization tandem mass spectrometry

A method has been developed to quantify chlorpyrifos (O,O-diethyl-O-[3,5,6,-trichloro-2-pyridyl] phosphorothionate) and its metabolites chlorpyrifos-oxon (O,O-diethyl-O-[3,5,6,trichloro-2-pyridinyl] phosphate) and TCP (3,5,6,-trichloro-2-pyridinol) in rat brain tissue by coupled-column liquid chromatography/electrospray ionization tandem mass spectrometry (LC/LC/ESI-MS/MS). Rat brains were homogenized and treated by protein precipitation using ice-cold acetonitrile. The supernatant was directly injected onto the coupled-column system. Sample clean-up was achieved on a Zorbax Extend-C(18) column (2.1 x 50 mm, 5 microm) using a mobile phase of acetonitrile/water with 0.0025% formic acid (40:60, v/v). The compounds were separated isocratically on a Zorbax Eclipse XDB C(8) column (2.0 x 150 mm, 5 microm) using a mobile phase of acetonitrile/water with 0.0025% formic acid (75:25, v/v). Chlorpyrifos and chlorpyrifos-oxon were detected in positive ion mode using multiple reaction monitoring (MRM). TCP was detected in negative ion mode using precursor-to-precursor transition monitoring. The method was validated and the specificity, linearity, limit of quantitation (LOQ), precision, accuracy, stability, and recoveries were determined. Calibration curves for all three analytes yielded correlation coefficients of 0.993 or greater. The LOQs were 25.3 ng/g for chlorpyrifos and 6.3 ng/g for chlorpyrifos-oxon and TCP. All precision relative standard deviations (RSDs) were less than 16% for the LOQ and less than 11% for the other QC samples. This method was successfully applied to six rats that were injected subcutaneously with chlorpyrifos.     

A simple and sensitive assay for the quantitative analysis of paclitaxel and metabolites in human plasma using liquid chromatography/tandem mass spectrometry

A sensitive and specific LC-MS/MS assay for the determination of paclitaxel and its 3'p- and 6-alpha-hydroxy metabolites is presented. A 200 microL plasma aliquot was spiked with a 13C6-labeled paclitaxel internal standard and extracted with 1.0 mL tert-butylmethylether. Dried extracts were reconstituted in 0.1 M ammonium acetate-acetonitrile (1:1, v/v) and 25 microL volumes were injected onto the HPLC system. Separation was performed on a 150 x 2.1 mm C18 column using an alkaline eluent (10 mm ammonium hydroxide-methanol, 30:70, v/v). Detection was performed by positive ion electrospray followed by tandem mass spectrometry. The assay quantifies a range for paclitaxel from 0.25 to 1000 ng/mL and metabolites from 0.25 to 100 ng/mL using 200 microL human plasma samples. Validation results demonstrate that paclitaxel and metabolite concentrations can be accurately and precisely quantified in human plasma. This assay is now used to support clinical pharmacologic studies with paclitaxel.     

Direct injection versus liquid-liquid extraction for plasma sample analysis by high performance liquid chromatography with tandem mass spectrometry.

Direct injection versus liquid-liquid extraction for post-dose human plasma sample analysis by high performance liquid chromatography with tandem mass spectrometry (LC/MS/MS) have been studied using a drug candidate compound. For the direct-injection method, an Oasis(R) HLB column (1 x 50 mm, 30 micrometer) was used as the on-line extraction column and a conventional Waters symmetry C18 column (3.9 x 50 mm, 5 micrometer) was used as the analytical column. Each plasma sample (100 microL) was mixed with 100 microL of a working solution of the internal standard in aqueous 0.05 M ammonium acetate (pH 6.9), and portions (10 microL) of these samples were then injected into the LC/MS/MS system. For the liquid-liquid extraction method, a YMC Basic C18 column (2.0 x 50 mm, 5 micrometer) was used as the analytical column. Each sample (0.5 mL) was extracted with methyl tert-butyl ether and the extract was reconstituted and injected into the LC/MS/MS system. The total analysis time for both methods was 2.0 min per sample. The accuracy, inter-day precision and intra-day precision obtained from the quality control samples were within 8% for both methods. The analysis results of post-dose human plasma samples showed that the deviations of 91% of the concentrations obtained using the direct-injection method were within +/-20% from the concentrations obtained using the liquid-liquid extraction method, and the overall average percentage deviation was -1.5%. The results showed that the two methods were equivalent in terms of total chromatographic run time, accuracy and precision. However, for a batch of 100 samples, the sample preparation time for the direct-injection method was only about 25% of the time required for liquid-liquid extraction. This decrease in sample preparation time resulted in the doubling of the overall sample analysis throughput.     

Enhanced resolution triple-quadrupole mass spectrometry for ultra-sensitive and quantitative analysis of the investigational anticancer agent EO9 (apaziquone) and its metabolite EO5a in human and dog plasma to support (pre)-clinical studies of EOquin given intravesically

A highly sensitive and selective liquid chromatography/tandem mass spectrometric (LC/MS/MS) method was developed to quantify the experimental anticancer agent EO9 and its metabolite EO5a in biological matrices. A 200-microL aliquot of human/dog plasma was spiked with a mixture of deuterated internal standards EO9-d3 and EO5a-d4 and extracted with 1.25 mL of ethyl acetate. Dried extracts were reconstituted in 0.1 M ammonium acetate/methanol (7:3, v/v) and 20-microL volumes were injected onto the LC system. Separation was achieved on a 150 x 2.1 mm C18 column using an alkaline eluent (1 mM ammonium hydroxide/methanol (gradient system)). The detection was performed by a Finnigan TSQ Quantum Ultra equipped with an electrospray ionization source operated in positive mode and enhanced mass resolution capability. It demonstrated improved sensitivity with a factor 10-20 for EO9 and EO5a over a 3-decades dynamic range, with acceptable accuracy and precision, when compared with the previously described assay for EO9 and EO5a, developed by our group, using an API 2000. The assay quantifies a range from 0.5 to 500 ng/mL for EO9 and EO5a using 200-microL human plasma and dog samples. The described mass resolution method was successfully applied for the evaluation of the pharmacokinetic profile of EO9 and its metabolite EO5a in human and dog plasma.     

Chromatographic and column stability at pH 7 of a C18 dimethylurea polar stationary phase

Chromatographic evaluations of a C18 dimethylurea phase in 150 mm x 3.9 mm HPLC columns were performed using the Tanaka and Engelhardt test mixtures. The applicability of the new C18 dimethylurea phase was also evaluated with a mixture of some herbicides and their metabolites. An artificial aging procedure was also performed by passing a potassium phosphate mobile phase buffered at pH 7.0 through C18 50 mm x 3.9 mm dimethylurea columns. The column stability was evaluated by means of the chromatographic parameters obtained for the separation of some compounds from the Neue test mixture, using apolar, polar and highly basic analytes.     

Simultaneous liquid chromatographic determination of ezetimibe and simvastatin in pharmaceutical products

A reversed-phase liquid chromatographic (LC) method was developed and validated for the simultaneous determination of ezetimibe and simvastatin in pharmaceutical dosage forms. The LC method was carried out on a Synergi fusion C18 column (150 mm x 4.6 mm id) maintained at 45 degrees C. The mobile phase consisted of phosphate buffer 0.03 M, pH 4.5-acetonitrile (35 + 65, v/v) run at a flow rate of 0.6 mL/min, and detection was made using a photodiode array detector at 234 nm. The chromatographic separation was obtained within 15.0 min, and calibration graphs were linear in the concentration range of 0.5-200 microg/mL. Validation parameters such as specificity, linearity, precision, accuracy, and robustness were evaluated, giving results within the acceptable range for both compounds. Moreover, the proposed method was successfully applied for the routine quality control analysis of pharmaceutical products.     

Chromatographic performance of a new polar poly(ethylene glycol) bonded phase for the phytochemical analysis of Hypericum perforatum L

The aim of this study was to evaluate the chromatographic performance of a poly(ethylene glycol) (PEG) stationary phase for the HPLC analysis of the secondary metabolites (chlorogenic acid, flavonoids, phloroglucinols and naphthodianthrones) in methanolic extracts of Hypericum perforatum L. (St. John's Wort) flowering tops, herbal medicinal products and dietary supplements. A fast and reliable method was developed. The analyses were carried out on a Supelco Discovery HS PEG column (150 mm x 4.6 mm i.d., 5 microm). A gradient mobile phase, composed of 0.1 M aqueous acetic acid solution (pH 2.8) and methanol-acetonitrile (5:4, v/v), was used. The flow rate was 1 mL/min. The photodiode array detector monitored the eluent at 270 (for chlorogenic acid, flavonoids and phloroglucinols) and 590 nm (for naphthodianthrones). The column was maintained at room temperature. The total running time was 40 min. The method was validated and showed good linearity, precision, accuracy, sensitivity and specificity. Through the above described phytochemical markers, this technique allowed the unequivocal identification and standardization of H. perforatum plant material and phytoproducts. The quantification data highlighted the fact that the products on sale, in particular those labeled as dietary supplements, varied widely in the quantitative composition of the active constituents. The developed method could be considered suitable for the quality control of H. perforatum herb and derivatives.     

Considerations on the possibilities and limitations of comprehensive normal phase-reversed phase liquid chromatography (NPLC x RPLC)

A comprehensive normal phase system LC-reversed phase LC (NPLC x RPLC) was evaluated for the separation of a pharmaceutical mixture and citrus oil extracts. NPLC was performed on a 25 cm x 1 mm ID x 5 microm dp diol phase. In the second dimension, an RP 18 monolithic column (10 cm L x 4.6 mm ID x 2 microm macropore size) and an octadecyl silicagel-packed column (5 cm L x 4.6 mm ID x 3.5 microm dp) were applied for the analyses of the pharmaceutical sample and the citrus oil extracts, respectively. A two-position/ten-port switching valve was used as interface. Under optimised LC conditions, the high degree of orthogonality between NP and RP resulted in peak capacities of 300 for the pharmaceutical sample and of 450 for the citrus oil extract composed of lemon and orange oil. Despite the features of NPLC x RPLC, several shortcomings related with the solvent incompatibility between the two LC modes were identified and the practical consequences were discussed.     

Direct determination of endogenous melatonin in human saliva by column-switching semi-microcolumn liquid chromatography/mass spectrometry with on-line analyte enrichment

An analytical method that enables direct and sensitive determination of endogenous melatonin (MLT) in human saliva was developed by means of column-switching semi-microcolumn liquid chromatography (i.d.: 1-2 mm)/mass spectrometry (LC/MS). The system allows direct injection analysis of a 400-microL aliquot of saliva with minimal sample pretreatment (internal standard (IS) addition and vortex mixing) and a relatively short run-time (10 min). The system consists of three columns to attain large volume injection and on-line analyte enrichment. A pre-column packed with a silica-based mixed-functional C8 (4.0 mm i.d. x 20 mm) was used for on-line sample cleanup. MLT and an IS, the d7 isomer of MLT (d7-MLT), were heart-cut by valve switching and enriched at the top of the intermediate trapping column packed with a silica-based C18 (4.0 mm i.d. x 10 mm). Subsequently, the analytes were backflushed into a semi-micro C18 silica column (2.0 mm i.d. x 150 mm) for the final separation. MLT and IS were ascertained by positive electrospray ionization and selected ion monitoring (SIM). MLT was monitored based on its fragment ion at m/z 174.1 by in-source collision-induced dissociation (CID). The validation of this method revealed a detection limit of 2.5 pg mL(-1) at a signal-to-noise (S/N) ratio of 5. The linearity of the method was established in the ranges 5-250 and 100-2500 pg mL(-1) with a coefficient of determination of greater than 0.998. Accuracies, evaluated at five levels in the range 5-1000 pg mL(-1), were between 81 and 108% with a relative standard deviation (RSD) ranging from 1.3-20%. The method was successfully applied for the endogenous saliva MLT monitoring of two healthy subjects.     

Separation of a BMS drug candidate and acyl glucuronide from seven glucuronide positional isomers in rat plasma via high-performance liquid chromatography with tandem mass spectrometric detection

A high-performance liquid chromatography/tandem mass spectrometry (LC/MS/MS) method has been developed and validated for the determination of a BMS drug candidate and its acyl glucuronide (1-O-beta glucuronide) in rat plasma. A 50-microL aliquot of each plasma sample was fortified with acetonitrile containing the internal standard to precipitate proteins and extract the analytes of interest. After mixing and centrifugation, the supernatant from each sample was transferred to a 96-well plate and injected into an LC/MS/MS system. Chromatographic separation was achieved isocratically on a Phenomenex Luna C(18), 3 mm x 150 mm, 3 microm column. The mobile phase contained 0.075% formic acid in 70:30 (v/v) acetonitrile/water. Under the optimized chromatographic conditions, the BMS drug candidate and its acyl glucuronide were separated from its seven glucuronide positional isomers within 10 min. Resolution of the parent from all glucuronides and acyl glucuronide from its positional isomers was critical to avoid their interference with quantitation of parent or acyl glucuronide. Detection was by positive ion electrospray MS/MS on a Sciex API 4000. The standard curve, which ranged from 5 to 5000 ng/mL, was fitted to a 1/x(2) weighted quadratic regression model for both the BMS drug candidate and its acyl glucuronide. Whole blood and plasma stability experiments were conducted to establish the sample collection, storage, and processing conditions. The validation results demonstrated that this method was rugged and repeatable. The same methodology has also been used in mouse and human plasma for the determination of the BMS drug candidate and its acyl glucuronide.     

The novel assay method for nicotine metabolism to cotinine using high performance liquid chromatography

Nicotine is the primary psychoactive component in tobacco. It is taken into the body by tobacco smoking, and mainly metabolized to cotinine in the hepatic cytochrme P450 (CYP) 2A6. The objective of this study was to develop a sensitive method for the determination of nicotine metabolism to cotinine using HPLC. The internal standard, trans-4'-carboxycotinine methyl ester was synthesized with a simple method. The nicotine and cotinine were separated completely and detected by C(18) 5-μm analytical column (L-column Octa decyl silyl (ODS), 150 mm × 4.6 mm i.d.) equipped with a C(18) 5-μm guard column (L-column ODS, 10 mm × 4.6 mm i.d.) and ultraviolet detection at 260 nm. The detection limit of the assay was 0.05 μM for cotinine (n=5, R.S.D) and 0.1 μM for nicotine. Thus the present results provided a sensitive and useful method for the determination of nicotine metabolism catalyzed by CYP2A6.     

Simultaneous determination of swertiamarin and its metabolites (5Z)-5-ethylidene-8-hydroxy-3,4,5,6,7,8-hexahydro-1H-pyrano[3,4-c]pyridin-1-one and erythrocentaurin in broth of Aspergillus niger by HPLC

When cultivated with Aspergillus niger, swertiamarin, an important drug, is rapidly transformed into erythrocentaurin and (5Z)-5-ethylidene-8-hydroxy-3,4,5,6,7,8-hexahydro-1H-pyrano[3,4-c]pyridin-1-one (M(1)), a new compound with high anti-inflammatory activity. A simple and rapid HPLC method for simultaneous determination of swertiamarin and its two metabolites in broth of A. niger is described. The chromatographic separation was achieved on a C(18) ODS column (250 x 4.6 mm i.d.) by gradient elution with 0.04% formic acid in water and 0.04% formic acid in acetonitrile as the gradient mixtures. The flow rate was 1 mL/min, the detection wavelength was 237 nm and the column temperature was kept at 30 degrees C. The retention times of swertiamarin, erythrocentaurin and M(1) were 14.6, 16.8 and 24.8 min, respectively. The mean absolute recoveries of three analysts were over 96%. Quantification limits were 0.02 microg/mL for swertiamarin and 0.05 microg/mL for both of the two metabolites. The method was applied for the quantification of swertiamarin and its two metabolites during the fermentation process and the evaluation of the bioavailabilities in the Caco-2 monolayer.