Direct analysis of crude plasma samples by turbulent flow chromatography/tandem mass spectrometry

Summary Turbulent flow chromatography coupled to tandem mass spectrometry (TFC-MS-MS) has recently emerged as a potentially fast, sensitive and specific technique for the direct analysis of pharmaceutical compounds from crude plasma. TFC-MS-MS removes the need for time-consuming sample preparation procedures such as solid-phase extraction (SPE) or liquid-liquid extraction (LLE). A relatively high flow rate combined with the use, of an HPLC column with large porous particles allows the on-line clean up and quantification of compounds in plasma samples. Until, now, the amount of plasma directly injected into TFC systems has rarely exceeded 30 μL in order to prevent rapid column degradation. Increasing the injection volume also induces high carry-over levels, particularly for drugs with basic and/or lipophilic properties. This paper describes the first genetic TFC-MS-MS method developed in a 96-well format, which allows the direct injection of 200 μL of 1∶1 diluted plasma (equivalent to 100 μL neat plasma). An average, of 390 injections was carried out with each extraction column. More than 2000 dog plasma samples were injected into the system without any sign of carryover. The method was fully validated over a 5–500 ng mL⁻¹ range for three basic compounds: doxazosin, CP122,288 and dofetilide. The imprecision was 1.2 to 8.3% for doxazosin, 1.5 to 4% for CP122,288 and 1.6 to 9.2% for dofetilide. The inaccuracy ranged from 6% to 7.9%. This generic methodology was then used to assay two structurally unrelated development compounds, showing that the method accuracy and sensitivity were adequate for the early pharmacokinetic (PK) studies in animals.     

Sequential Injection System with Modified Glass Capillary for Automation in Immunoassay of Chondroitin Sulfate

Sequential injection was introduced to perform a multi-step immunoassay. Modified low cost hematocrit glass capillary was employed as the immobilization surface for a competitive immunoassay of chondroitin sulfate (CS), a potential biomarker for cancer. Glass capillary is low cost and adapts well to the flow system without causing back pressure. The analysis time per sample run with automation of the multi-step immunoassay is improved as compared to the conventional batch-wise micro-plate format. The performance of the sequential injection capillary immunoassay (SI-CI) system for CS was evaluated with spiked human serum samples.     

High performance displacement chromatography of proteins: Separation of β-Lactoglobulins A and B

Summary β-Lactoglobulins A and B were separated by high performance displacement chromatography on an anion-exchanger column with chondroitin sulfate as the displacer. A sample of 100 mg containing a mixture of the two β-lactoglublins was separated on a column of 75×7.5 mm in a single chromatographic run. The separation process followed the rules of the classical displacement development: the two proteins formed contiguous rectangular bands and their concentrations were dependent on the displacer concentration. The results demonstrate that high performance displacement chromatography is a useful technique for the separation of proteins in preparative amounts with columns and instrumentation typically used in analytical HPLC. Furthermore, it has the potential to become the method of choice in large scale separation of proteins.     

Development and validation of an ion‐exchange chromatography method for heparin and its impurities in heparin products

An anion‐exchange liquid chromatography method for the determination of heparin and its impurities (dermatan sulfate and oversulfated chondroitin sulfate) was developed using chemometric‐assisted optimization, including multivariate experimental design and response surface methodology. The separation of heparin, dermatan sulfate, and oversulfated chondroitin sulfate (R_s above 2.0) was achieved on a Dionex RF IC IonPac AS22 column with a gradient elution of 10–70% of 2.5 M sodium chloride and 20 mM Tris phosphate buffer (pH 2.1) at a flow rate of 0.6 mL/min and UV detection at 215 nm. Method validation shows good linearity (r > 0.99), acceptable precision (%relative standard deviations <11.4%) and trueness (%recovery of 92.3–103.9%) for all analytes. The limits of detection for dermatan sulfate and oversulfated chondroitin sulfate are equivalent to 0.11% w/w (10.5 μg/mL) and 0.07% w/w (7.2 μg/mL), while the limits of quantification are 0.32% w/w (31.5 μg/mL) and 0.22% w/w (22.0 μg/mL) relative to heparin, respectively. The method is specific for heparin, dermatan sulfate, and oversulfated chondroitin sulfate without interference from mobile phase and sample matrices and could be used for accurate quantitation the drug and its impurities in a single run. Applications of the method reveal contents of heparin between 90.3 and 97.8%. Dermatan sulfate and oversulfated chondroitin sulfate were not detected in any of the real‐life samples.     

Quantification of trovafloxacin in serum by high-performance liquid chromatography with on-line solid-phase extraction

Summary A rapid, simple, accurate and sensitive liquid chromatographic assay with on-line solid-phase extraction is described for determination of trovafloxacin in human serum. Samples were deproteinized with acetonitrile and injected on to an NH₂ extraction column for sample clean-up. Thereafter, an on-line column-switching system was used for quantitative transfer of the drug to a C₁₈ analytical column. Separation was performed by ion-pair chromatography and detection was by ultraviolet absorbance at 275 nm. Recovery was 98.5%. The linear range was from 0.25 to 20μg mL⁻¹, with a correlation coefficient of 0.999. Detection limit was 0.1 μg mL⁻¹ from extraction of 25 μL serum.     

Combination of a two-step fluorescence assay and a two-step anti-Factor Xa assay for detection of heparin falsifications and protein in heparins

There are several methods for sensitive detection of oversulfated chondroitin sulfate (OSCS) in heparin. Although contamination with OSCS is unlikely to be repeated, use of other compounds to counterfeit heparin must be considered. We have previously developed a two-step fluorescence microplate assay (two-step FI assay) for detection of OSCS. First, the heparin sample is incubated with heparinase I, then its increasing effect on the fluorescence intensity (FI) of the sensor molecule Polymer-H is measured (PolyH assay). The high sensitivity of the assay is shown to be based on heparinase I inhibition by OSCS. The objective of this study was to evaluate another assay option — indirect quantification of OSCS after heparinase I incubation by means of the anti-Factor Xa (aXa) activity of the remaining undegraded heparin (two-step aXa assay). We also examined, whether other heparin mimetics (HepM), direct Factor Xa inhibitors (DXI), and protein impurities are detectable by use of these assays. Heparin was spiked with different amounts of HepM including OSCS, pentosan polysulfate, dextran sulfate, curdlan sulfate, the natural contaminant dermatan sulfate, the DXI rivaroxaban, and BSA as a protein. These samples were compared with pure heparin in the two-step FI assay, the two-step aXa assay, and in the PolyH assay and the aXa assay without heparinase I incubation. Both two-step assays sensitively measured contamination with all the HepM (LOD ≤ 0.5%, LOQ ≤ 0.7%). The two-step aXa assay also detected rivaroxaban (LOD 0.3%, LOQ 0.4%), whereas the two-step FI assay was shown to be suited to determination of protein impurities (LOD 0.11%, LOQ 0.13%). Use of two different heparinase I inactivation procedures enabled clear differentiation between protein, HepM, and both contaminants. Finally, with the aXa assay the heparin potency can be determined in the same assay run, whereas the FI increase in the PolyH assay was shown to be useful for identification. In conclusion, both the two-step FI assay and the two-step aXa assay are sensitive, rapid, and simple tests for the detection of counterfeit heparin. Comprehensive information about heparin quality can be obtained by their combined use and the parallel measurement of non-incubated heparin samples.     

A simple and fast liquid chromatography–tandem mass spectrometry method for measurement of underivatized <Emphasis Type="SmallCaps">l</Emphasis>-arginine,symmetric dimethylarginine,and asymmetric dimethylarginine and establishment of the reference ranges

Asymmetric dimethylarginine (ADMA) is an endogenous inhibitor of nitric oxide synthase and an established biomarker for endothelial function, while symmetric dimethylarginine (SDMA), an emerging biomarker for renal function, has been shown to outperform creatinine-based equations for estimated glomerular filtration rate. In order to study these analytes for clinical research, a fast and simple method for measuring arginine (ARG), SDMA, and ADMA in plasma by liquid chromatography–tandem mass spectrometry (LC-MS/MS) has been developed. Plasma (50 μL) was mixed with 50 μL of internal standard of ¹³C-arginine and d₇-ADMA followed by protein precipitation with methanol containing 1% ammonium acetate (300 μL). After centrifugation, the supernatant (100 μL) was mixed with 300 μL of acetonitrile with 1% formic acid, and the mixture was injected onto a silica column monitored by a mass spectrometer. The analytical cycle time was 5.0 min. The method was linear from 5.7 to 489.7 μM for ARG, 0.06 to 5.15 μM for SDMA, and from 0.34 to 5.65 μM for ADMA, with an accuracy of 99.0–120.0%. Total coefficients of variation for all analytes ranged from 2.7% to 7.7% for three concentration levels. The effects of hemolysis, lipemia, uremia, icterus, specimen tube types, storage at different temperature, and freeze/thaw were thoroughly investigated. Reference ranges were established using 51 well-defined reference subjects (12 men and 39 women, age 19–64 years): 53.1–129.7 μM for ARG, 0.32–0.65 μM for SDMA, and 0.36–0.67 μM for ADMA. In conclusion, the validated LC-MS/MS method described here offers a fast and reliable ARG, SDMA, and ADMA quantitation in plasma with minimum sample preparation.     

Detection of chondroitin sulfate-binding proteins on the membrane

Chondroitin sulfate is a ubiquitous component of proteoglycans that is present both in the extracellular matrix and at the cell surface of various tissues. Until recently, chondroitin sulfate has attracted less attention than heparan sulfate and dermatan sulfate, owing to the limited number of known chondroitin sulfate-binding proteins. To determine the biological function of chondroitin sulfate, biotinylated probes were prepared and used to search for binding proteins. Chondroitin sulfates A, C, D, and E were biotinylated through either the uronic acid or the residual core peptide. Lysates from mouse Lewis lung carcinoma (3LL) cells were blotted onto a nitrocellulose membrane after sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and the membrane was treated with the biotinylated chondroitin sulfates. Among the chondroitin sulfate variants, the E type showed the most intense bands upon visualization of the membrane with avidin-conjugated alkaline phosphatase and the appropriate substrates. The binding of chondroitin sulfate E to proteins in the cell lysate was not affected by the A, C or D variants but was reduced by treatment with dermatan sulfate. Lysates from 3LL cells were also treated with biotinylated chondroitin sulfate E and, after two-dimensional (2-D) electrophoresis and blotting, several chondroitin sulfate E-binding proteins including lamins and heterogeneous nuclear ribonucleoproteins were identified by mass spectrometry.     

Semi-micro size-exclusion chromatography: Molecular-weight measurement of poly (ethylene terephthalate) and separation of oligomers and prepolymers

Summary Several polystyrene gels of different pore sizes were packed into a 500 mm×2.1 mm I.D. column. Semi-micro size-exclusion chromatography (SEC) using these columns was carried out with a system consisting of a triple piston pump, a micro loop injector and a flow cell with 1.0-μl cell volume constructed for semi-micro HPLC, because the dead volume of the injector and the cell volume of flow cell for conventional HPLC caused a significant loss in column efficiency. The effects of sample amount, injection volume and mobile phase flow rate on column efficiency and retention volume were examined and the optimized operational variables of the sample amount (below 500 μg), the injection volume (less than 15 μl) and the flow rate range (30–70 μl/min) determiend for semi-micro SEC. Oligostyrene, epoxy resin, phenol-formaldehyde resin and phthalates were analyzed by the optimized semi-micro SEC system under the given conditions. In addition, molecular weight distribution of four different poly(ethylene terephthalate) films was successfully measured by using a mixture of chloroform and hexafluoroisopropanol as the eluent.     

Determination of Urinary Glucose by a Flow Injection Analysis Amperometric Biosensor and Ion-Exchange Chromatography

A practical biosensor system has been developed for the determination of urinary glucose using a flow-injection analysis (FIA) amperometric detector and ion-exchange chromatography. Glucose oxidase was immobilized onto porous aminopropyl glass beads via glutaraldehyde activation to form an immobilized enzyme column. On the basis of its negative charge at pH 5.5, endogenous urate in urine samples was effectively retained by an upstream anion-exchange resin column. The biosensor system possessed a sensitivity of 160 ±2.4 RU μM^-1 (RU or relative unit is defined as 2.86 μV at the detection output) for glucose with a minimum detection level of 10 μM. When applied for the determination of urinary glucose, the result obtained compared very well with that of the widely accepted hexokinase assay. The immobilized glucose oxidase could be reused for more than 1000 repeated analyses without losing its original activity. The reuse of the acetate anion-exchange column before replacement would be about 25–30 analyses. Acetaminophen and ascorbic acid were also effectively adsorbed by the acetate anion exchanger. The introduction of this type of anion exchanger thus greatly improved the selectivity of the FIA biosensor system and fostered its applicability for the determination of glucose in urine samples.     

Determination of sulfation pattern in brain glycosaminoglycans by chip-based electrospray ionization ion trap mass spectrometry

Chondroitin sulfate (CS) and dermatan sulfate (DS) glycosaminoglycans display variability of sulfation in their constituent disaccharide repeats during chain elongation. Since a large proportion of the extracellular matrix of the central nervous system (CNS) is composed of proteoglycans, CS/DS disaccharide degree and profile of sulfation play important roles in the functional diversity of neurons, brain development, and some of its pathological states. To investigate the sulfation pattern of CS/DS structures expressed in CNS, we introduced here a novel method based on an advanced system encompassing fully automated chip nanoelectrospray ionization (nanoESI) in the negative ion mode and high capacity ion trap multistage mass spectrometry (MS²–MS³) by collision-induced dissociation (CID). This method, introduced here for the first time in glycomics of brain glycosaminoglycans, was particularly applied to structural investigation of disaccharides obtained by β-elimination and digestion with chondroitin B and AC I lyase of hybrid CS/DS chains from wild-type mouse brain. Screening in the chip-MS mode of DS disaccharide fraction resulting after depolymerization with chondroitin B lyase revealed molecular ions assigned to monosulfated disaccharide species having a composition of 4,5-Δ-[IdoA-GalNAc]. By optimized CID MS²–MS³, fragment ions supporting the localization of sulfate ester group at C4 within GalNAc were produced. Chip ESI MS profiling of CS disaccharide fraction obtained by depolymerization of the same CS/DS chain using chondroitin AC I lyase indicated the occurrence of mono- and bisulfated 4,5-Δ-[GlcA-GalNAc]. The site of oversulfation was determined by MS²–MS³, which provided sequence patterns consistent with a rare GlcA-3-sulfate–GalNAc-6-sulfate structural motif.      

Rapid quantification of tilidine,nortilidine, and bisnortilidine in urine by automated online SPE-LC-MS/MS

The opioid tilidine is a prodrug which is hepatically metabolized to active nortilidine and bisnortilidine. Due to the increasing abuse of tilidine by drug users and the lack of a specific immunoassay, we developed an analytical method for the quantification of tilidine, nortilidine, and bisnortilidine in urine suitable for screening. In a following step, this method was used to establish data on excretion kinetics of the substances in order to evaluate the time window of detection after a single oral dose of tilidine/naloxone and also was applied to authentic urine samples from correctional facilities. Urine samples were mixed with internal standard solution and extracted on a weak cation exchanger at pH 6 using a Symbiosis Pico system. The chromatographic separation was achieved within a 3.5-min run time on a Phenylhexyl column (50 × 2.0 mm, 5 μm) via gradient elution (methanol and 0.2% formic acid) at a flow rate of 0.50 mL/min. The ESI-MS/MS was performed on a QTrap 3,200 in positive multiple reaction monitoring mode using two mass transitions per analyte. Validating the method resulted in a lower limit of quantification of 1.0 μg/L followed by a linear calibration range to 100 μg/L for each analyte (r ² > 0.99). The analytical method allowed the detection of a single dose of a commercially available tilidine solution up to 7 days after administration. Using this highly sensitive method, 55 of 3,665 urine samples were tested positive.     

Short-end injection capillary electrophoresis for quantification of plasma chondroitin sulfate isomer disaccharides

We describe a new ultra-rapid capillary electrophoresis method with UV detection for analysis of the disaccharides obtained after enzymatic depolymerization of plasma chondroitin sulfates. The free reducing groups of the released carbohydrate molecules are derivatized with 2-aminoacridone by reductive amination in the presence of cyanoborohydride. The fluorotagged products can be separated by short-end injection capillary electrophoresis in a capillary with an effective length of 10.2 cm. The migration times of Δdi-0S and Δdi-4S were 0.95 and 1.81 min, respectively. We compared the proposed method with UV detection to a reference CE-LIF assay by measuring plasma chondroitin sulfate in 94 subjects. The described assay for total plasma CS measurement may, owing to the high throughput and the fast analytical times, be a good tool for routine studies both in research and in clinical applications.     

Radiometric determination of sulfate

A radiometric procedure for the determination of sulfate based upon the precipitation of barium sulfate is described with a sensitivity of about 0.01 μg/ml. Carrier-free³⁵SO₄ is added to the sample to measure the chemical recovery. The sulfate is precipitated with an excess of barium having a known specific activity of¹³³Ba. The amount of¹³³Ba determined by gamma counting is directly related by stoichiometry to the amount of sulfate in the precipitate.     

ALTERATIONS OF PROTEOGLYCANS IN ULTRAVIOLET-IRRADIATED SKIN

Abstract— The effect of UVB exposure on the distribution and synthesis of dermal proteoglycans was measured in the skin of hairless mice. Two groups of mice were included: one was irradiated for 10 weeks; the other was kept as control. After intraperitoneal injection of sodium ³⁵S-sulfate, punch biopsies were taken for histology and proteoglycans were extracted from the remaining skin with 4 M guanidinium chloride, containing 3–[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (0.5%, weight per volume). Following proteolytic digestion, the glycosaminoglycan constituents were isolated and analyzed by quantitative cellulose acetate electrophoresis and enzymatic digestibility.
Under the influence of UVB radiation, newly synthesized proteoglycans measured by ³⁵SO₄ uptake increased as much as 60%. In addition, the irradiated skin had a higher average content of proteoglycan than had control skin (4981 μg vs 4134 μg/g dry weight). This could be ascribed to an increase in heparin (1400 vs 533 μ g/g dry weight) and heparan sulfate (472 vs 367 μg/g dry weight), whereas no change in the concentration of hyaluronic acid (1243 vs 1372 μg/g dry weight) and dermatan sulfate (1866 vs 1863 μg/g dry weight) was observed. The irradiated animals also exhibited a marked increase in the synthesis of heparan sulfate and heparin (62% and 71%, respectively). These results demonstrate that chronic doses of UVB altered proteoglycan metabolism through both quantitative and qualitative changes.     

Online preconcentration and determination of chondroitin sulfate,dermatan sulfate and hyaluronic acid in biological and cosmetic samples using capillary electrophoresis

Capillary electrophoresis with large‐volume sample stacking using an electroosmotic flow pump was developed for the determination of chondroitin sulfate, dermatan sulfate, and hyaluronic acid. Central composite design was used to simultaneously optimize the parameters for capillary electrophoresis separation. The optimized capillary electrophoresis conditions were 200 mM sodium dihydrogen phosphate, 200 mM butylamine, and 0.5% w/v polyethylene glycol as a background electrolyte, pH 4 and ‐16 kV. Exploiting large‐volume sample stacking using an electroosmotic flow pump, the sensitivity of the proposed capillary electrophoresis system coupled with UV detection was significantly improved with limits of detection of 3, 5, 1 mg/L for chondroitin sulfate, dermatan sulfate, and hyaluronic acid, respectively. The developed method was applied to the determination of chondroitin sulfate and hyaluronic acid in cell culture media, cerebrospinal fluid, cosmetic products, and supplementary samples with highly acceptable accuracy and precision. Therefore, the proposed capillary electrophoresis approach was found to be simple, rapid, and reliable for the determination of chondroitin sulfate, dermatan sulfate, and hyaluronic acid in cell culture media, cerebrospinal fluid, cosmetic, and supplementary samples without sample pretreatment.     

Base sequence- and <Emphasis Type="Italic">T</Emphasis><Subscript>m</Subscript>-dependent DNA oligomer separation by open tubular capillary columns carrying complementary DNA oligomers as probes

DNA chips prepared on a flat glass surface have unavoidable drawbacks when used for quantitative analysis. In an attempt to overcome this problem, we constructed an HPLC-type system suitable for quantitative analysis that enables base sequence- and T _m-dependent DNA oligomer separation in a flow system. A small open tubular capillary column (300-mm × 100-μm I.D.) was used. The DNA oligomers used as probes had an amino group at the 5′-end and were immobilized on the inner silica surface of the capillary column which had been sequentially treated with 3-aminopropyltriethoxysilane, butyltrimethoxysilane, and disuccinimidylglutarate. Using the combination of probe-immobilized column placed in a column oven equipped with temperature gradient function, a nano-flow-controllable pump, a small sample-loading injector, and a capillary-fitted UV detector, we succeeded in separating complementary and non-complementary DNA oligomers in specific and quantitative modes. We also designed a temperature gradient strategy for efficient separation of target DNA oligomers in DNA mixture samples. Using a column carrying two different probes with similar T _m values, their complementary target DNA oligomers were also separated and detected. The developed DNA open tubular capillary column system investigated in the present study could be further improved as an alternative tool to DNA chips to be used for the quantitative analysis of DNA or mRNA samples. Kamakshaiah Charyulu Devarayapalli and Seung Pil Pack contributed equally to this paper.     

dual stage preconcentration system for flame atomic absorption spectrometry using flow injection on-line ion-exchange followed by solvent extraction

A dual stage preconcentration system based on flow injection on-line ion-exchange and solvent-extraction has been developed for flame atomic absorption spectrometry. Lead is taken as a model trace element. A column packed with Amberlite IRC-718 cation exchanger is incorporated into the FI manifold. The analyte is retained on the column by time-based sample loading and eluted by 1 mol/L HNO₃. The eluate is subsequently merged with potassium iodide and tetrabutylammonium bromide (TBABr), and isobutyl methyl ketone (IBMK). Lead is extracted on-line into IBMK as the ion-pair formed between the iodoplumbate anion and tetrabutylammonium cation. The organic phase is separated from the aqueous phase by a gravity phase separator. 50 μL of concentrate is introduced into the nebulizer-burner system of the spectrometer. An enhancement factor of 550 is achieved with a 30 mL sample consumption at a sampling frequency of 30/h. The precision (relative standard deviation) is 2.4% at 10 μg/L level and the detection limit is 0.3 μg/L (3 σ). The method was successfully applied to the determination of lead in water samples. Received: 19 March 1997 / Revised: 24 June 1997 / Accepted: 27 June 1997     

Molecularly imprinted monolith in-tube solid-phase microextraction coupled with HPLC/UV detection for determination of 8-hydroxy-2′-deoxyguanosine in urine

Urinary 8-hydroxy-2′-deoxyguanosine (8-OHdG) has been widely used as a biomarker of oxidative DNA damage. Measurements of 8-OHdG in urinary samples are challenging owing to the low level of 8-OHdG and the complex matrix. In this study, a novel molecularly imprinted polymer (MIP) monolithic column was synthesized with guanosine as a dummy template which was used as the medium for in-tube solid-phase microextraction (SPME). In-tube SPME coupled with HPLC/UV detection for extraction and determination of urinary 8-OHdG was developed. The synthesized MIP monolithic column exhibited high extraction efficiency owing to its greater phase ratio with convective mass transfer and inherent selectivity. The enrichment factor for 8-OHdG was found to be 76 and the limits of detection and quantification of the method for urinary samples were 3.2 nmol/L (signal-to-noise ratio 3) and 11 nmol/L (signal-to-noise ratio 10), respectively. The MIP^’s selectivity also made the sample preparation procedure and chromatographic separation much easier. The linear range of the proposed method was from 0.010 to 5.30 μmol/L (r = 0.9997), with a relative standard deviation of 1.1–6.8%, and the recovery for spiked urine samples was 84 ± 3%. The newly developed method was successfully applied to determine urinary samples of healthy volunteers, coking plant workers, and cancer patients. The 8-OHdG level in cancer patients was significantly higher than that in healthy people.     

Simple Liquid Chromatographic Determination of Minocycline in Brain and Plasma

A new high-performance liquid chromatography assay was developed for the determination of minocycline in plasma and brain. A solid–liquid extraction procedure was coupled with a reversed-phase HPLC system. The system requires a mobile phase consisting of acetonitrile:water:perchloric acid (26:74:0.25, v/v/v) adjusted to pH 2.5 with 5 M sodium hydroxide for elution through a RP8 column (250 × 3.0 mm, i.d.) with UV detection set at 350 nm. The method proved to be accurate, precise (RSD < 20%) and linear between 0.15–20 μg mL⁻¹ in plasma and 1–20 μg mg⁻¹ in brain. The method was successfully applied to a blood-brain barrier minocycline transport study.