an in vivo microdialysis coupled with liquid chromatography/tandem mass spectrometry study of...

Analytical Biochemistry 381 (2008) 214–223

Contents lists available at ScienceDirect

Analytical Biochemistry

journal homepage: www.elsevier .com/ locate /yabio

An in vivo microdialysis coupled with liquid chromatography/tandem massspectrometry study of cortisol metabolism in monkey adipose tissue

Li Sun a,*, Julie A. Stenken b, Janice E. Brunner a, Kimberly B. Michel a, Jennifer K. Adelsberger a,Amy Y. Yang a, Jamie J. Zhao a, Donald G. Musson a

a WP75A-303, DMPK, Merck Research Laboratories, West Point, PA 19486, USAb Department of Chemistry&Biochemistry, University of Arkansas, Fayetteville, AR 72701, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 28 March 2008Available online 2 July 2008

Keywords:MicrodialysisLC/MS/MS11b-HSD1Adipose tissueCortisoneCortisolStable isotope-labeledAnalysisMetabolismEnzyme

0003-2697/$ - see front matter � 2008 Elsevier Inc. Adoi:10.1016/j.ab.2008.06.036

* Corresponding author. Fax: +1 215 652 4524.E-mail address: [email protected] (L. Sun).

1 Abbreviations used: 11b-HSD1, 11b-hydroxysteroiendoplasmic reticulum; SIL, stable isotope-labeled; MS,chromatography; GC, gas chromatography; MS/MS, tanultraviolet; CAD, collision-activated dissociation; SRM,LLOQ, lower limit of quantification; QC, quality contro

It is postulated that elevated tissue concentrations of cortisol may be associated with the development ofmetabolic syndrome, obesity, and type 2 diabetes. The 11b-hydroxysteroid dehydrogenase type 1 (11b-HSD1) enzyme regenerates cortisol from inactive cortisone in tissues such as liver and adipose. To betterunderstand the pivotal role of 11b-HSD1 in disease development, an in vivo microdialysis assay coupledwith liquid chromatography/tandem mass spectrometry (LC/MS/MS) analysis using stable isotope-labeled (SIL) cortisone as a substrate was developed. This assay overcomes the limitations of existingmethodologies that suffer from radioactivity exposure and analytical assay sensitivity and specificityconcerns. Analyte extraction efficiencies (Ed) were evaluated by retrodialysis. The conversion of SIL-cor-tisone to SIL-cortisol in rhesus monkey adipose tissue was studied. Solutions containing 100, 500, and1000 ng/mL SIL-cortisone were locally delivered through an implanted 30-mm microdialysis probe inadipose tissue. At the delivery rate of 1.0 and 0.5 lL/min, Ed values for SIL-cortisone were between58.7 ± 5.6% (n = 4) and 72.7 ± 1.3% (n = 4), whereas at 0.3 lL/min Ed reached nearly 100%. The presenceof 11b-HSD1 activities in adipose tissue was demonstrated by production of SIL-cortisol during SIL-cor-tisone infusion. This methodology could be applied to cortisol metabolism studies in tissues of othermammalian species.

� 2008 Elsevier Inc. All rights reserved.

11b-Hydroxysteroid dehydrogenase type 1 (11b-HSD1)1-regu-lated cortisol metabolism has drawn tremendous interest, asincreasing evidence supports the hypothesis that increased tissueintracellular cortisol levels caused by elevated 11b-HSD1 activitymay contribute to pathologies of metabolic diseases in humans, suchas obesity, diabetes, and metabolic syndrome [1–4]. 11b-HSD1 andits isozyme 11b-HSD2 are two key players in regulating cortisol tis-sue concentrations [5–7]. 11b-HSD1, an endoplasmic reticulum (ER)-localized enzyme, is predominantly an oxoreductase that convertsthe biologically inactive glucocorticoid cortisone to the active gluco-corticoid cortisol in tissues such as liver, adipose tissue, and brain[8]. 11b-HSD2 is expressed mainly in kidney, and is responsible forinactivating cortisol by converting it back to cortisone. Thus far,the correlation of 11b-HSD1 activity in specific tissues with diseasestates and the mechanism of disease development are not clearly de-

ll rights reserved.

d dehydrogenase type 1; ER,mass spectrometry; LC, liquiddem mass spectrometry; UV,selected reaction monitoring;

l.

fined. Therefore, an assay that could elucidate the tissue-specificin vivo events related to 11b-HSD1-regulated cortisol metabolismis imperative [9].

The available methods for determination of 11b-HSD1 activity,such as systemic measurements of cortisone, cortisol, and theirmetabolites from plasma and urine samples [10,11], share somecommon limitations because plasma and urine are far away fromthe site of action, e.g., adipose tissue. Endogenous cortisol is se-creted from adrenal cortex and distributed ubiquitously in almostall tissues in humans [12,13]. Furthermore, cortisol is metabolizedby various steroid-metabolizing enzymes, including A ring reduc-tases and cytochrome P450s [14]. Concentrations of cortisol ob-tained from plasma and urine suffer from the inability todistinguish the activities of all the different enzymes that are in-volved and isolate which tissues contribute to cortisol metabolicpathways. Tissue cortisol concentrations have been determinedby measurement from in vitro biopsies [2,4,15]. Arteriovenoussampling was developed to measure in vivo cortisol [16], but thistechnique suffers from large variability and accounts only for sys-temic cortisol production.

Microdialysis infusion has been used for studying drug metab-olism and activities of various enzymes in different tissues and

mailto:[email protected]

http://www.sciencedirect.com/science/journal/00032697

http://www.elsevier.com/locate/yabio

dialDiffusion

tissuetissueDiffusion MMESEESS ⎯⎯⎯ →⎯+→−↔+⎯⎯⎯ →⎯inf

Capillary Removal Capillary Removal

MetabolismMetabolism

TISSUE

Scheme 1. An illustration of the sequence of mass transfer during a microdialysisinfusion study of drug metabolism and relevant kinetic processes. Sinf and Stissue

denote the substrate in the infusion fluid and tissue, respectively; E represents theenzyme; E-S represents the complex of the enzyme and substrate; Mtissue and Mdial

represent the metabolite in tissue and in dialysate, respectively.

Detection of 11b-HSD1 activity by microdialysis / L. Sun et al. / Anal. Biochem. 381 (2008) 214–223 215

organs such as brain, liver, and kidney [17–20]. In vivo microdialy-sis sampling has been utilized for various applications in cortisolresearch, and has shown great utility in obtaining regional cortisolactivity information that other techniques cannot achieve [21–23].A microdialysis assay for detecting tissue 11b-HSD1 activity wasreported by Sandeep et al. [24], in which adipose tissue 11b-HSD1 activity during cabenoxolone inhibition and insulin infusionwas monitored by microdialysis infusion of radioisotope-labeledtracer [1,2-3H2]cortisone. Later, Wake et al. applied this radiola-beled method to detect both reductase and dehydrogenase activityof 11b-HSD1 during insulin and intralipid infusions in humans[25]. This method requires dialysates to be processed using lengthysteps, including extraction in ethyl acetate, separation by thin-layer chromatography, fraction collection based on ultraviolet(UV) detection, followed by liquid scintillation counting for quan-tification. Radioactivity exposure and elaborate sample prepara-tion are the major drawbacks of this assay. An additionalcomplication with the radiolabeled approach is that endogenouscortisol cannot be quantified in a single sample.

The utilization of stable-isotope-labeled (SIL) cortisone and cor-tisol has the great advantage of avoiding exposure to radioactivity.Also, the SIL analogs can be distinguished from endogenous corti-sone and cortisol by their different masses. In the past, SIL analogshave been used for microdialysis sampling with mass spectromet-ric detection [26,27]. However, SIL analogs have not been reportedfor use in in vivo microdialysis studies of cortisol metabolism. Amajor obstacle is probably the lack of a sensitive analytical assayfor quantification of SIL analytes in dialysates. Mass spectrometry(MS) has great potential for detecting both endogenous and SILcortisone and cortisol with high levels of sensitivity and specificity.The quantification of microdialysates with liquid chromatography/mass spectrometry (LC/MS) was recently reviewed by Lanckmanset al. [28]. The coupling of microdialysis with LC/MS could be anideal solution that enables the study of tissue-specific 11b-HSD1activity, and potentially, the tissue distribution of endogenous glu-cocorticoids and their metabolites can be elucidated simulta-neously. In the past, gas chromatography/mass spectrometry(GC/MS) has been used for the analysis of cortisone and cortisolin human plasma [11,29]. However, GC/MS methods requireextraction of the analytes to an organic solvent and a lengthy pre-column derivatization. Recently, we have developed a microdialy-sis infusion of SIL M + 4 cortisone coupled with a direct injectionLC/tandem MS (MS/MS) assay to measure 11b-HSD1 activity in li-ver microsomes [30]. The analytical assay requires minimal samplepreparation, and reaches a limit of quantification as low as 0.1 ng/mL for cortisone and cortisol, showing significant improvementover existing assays for dialysate analysis.

For in vivo microdialysis, the extraction fraction Ed is a criticalparameter that relates analyte dialysate concentration to its con-centration surrounding the probe, and indirectly reflects the bio-logical kinetic events in the probe vicinity. By definition, Ed is thefraction of analyte that diffuses across the dialysis membrane fol-lowing its concentration gradient; hence, it can be expressed as

Ed ¼Cd � Ci

Cs � Ci; ð1Þ

Where Cd, Ci, and Cs are the concentrations of analytes in the dialy-sate, perfusion fluid, and un-disturbed external sample medium,respectively. Factors influencing in vivo Ed include analyte diffusiv-ity in tissue, tissue trauma during probe implantation, analyte ex-change rate between extracellular fluid and microvascular fluid,and between intracellular and extracellular fluid, analyte metabolicclearance rate, etc. [31]. Given all the variables involved in microdi-alysis sampling, careful assessment and interpretation of the micro-dialysis data are crucial. For tissue 11b-HSD1 activitydetermination, the use of in vitro Ed for analyte concentration cor-

rection [21] may not account for the effects of in vivo biologicalevents, such as metabolic removal and capillary exchange, whichcan have significant roles in the amount of metabolite collected.

In this study, the rhesus monkey was chosen as an animal mod-el for studying 11b-HSD1 activity. It has been reported that themonkey has 11b-HSD1 enzyme structure and catalytic propertiessimilar to those of humans [32], and monkey liver microsomeswere shown to have consistent cortisol formation with human li-ver microsomes [30]. In vitro microdialysis and in vivo retrodialysisof SIL cortisone and cortisol were used to characterize analyte Ed.The previously described LC/MS/MS assay for the measurementof 11b-HSD1 activity in liver microsomes was used to quantifydialysates [30]. Retrodialysis is frequently used as a probe calibra-tion method for in vivo microdialysis [33–35]. In retrodialysisexperiments, a calibrator compound is included in the perfusate,and the fraction of loss across the membrane to the surroundingtissue is measured, based on which analyte Ed is calculated. A ret-rodialysis calibrator can be the analyte itself, an analog structurallysimilar to the analyte of interest, or an isotope-labeled analog. Typ-ically, analogs are chosen to have similar mass transport character-istics. Thus, isotope-labeled derivatives are an ideal choice forretrodialysis experiments. During these in vivo experiments, SIL-cortisone delivery Ed was monitored to gain information regardingthe probe performance, the time to reach substrate delivery steadystate, the metabolite production capability as a function of infusedsubstrate concentration, etc.

The microsomal evaluation results verified that this in situmicrodialysis assay can detect variation of 11b-HSD1 enzymeactivity in vitro [30]. However, in vivo microdialysis in tissue is sub-ject to variables different from those in in vitro studies, due to thespecific tissue properties and dynamic biological events at theprobe area. The interplay of the elimination kinetics (e.g., meta-bolic and capillary removal) of the substrate and metabolite(s) isillustrated in Scheme 1. Microdialysis experimental conditions,such as the infusion concentration and flow rate, could also con-tribute to the experimental outcome [36,37]. Evaluation of micro-dialysis conditions is of great importance for understanding thesources of experimental variability, identifying the major determi-nants for experimental outcome, and optimizing the microdialysisparameters to improve assay reliability.

Materials and methods

Chemicals

Stable isotope-labeled M + 4 cortisone and M + 4 cortisol (struc-tures shown in Scheme 2) were synthesized by Merck ResearchLaboratories. The chemical purity of each standard was P99.2%,as determined by HPLC-UV, and proton and 13C NMR. The isotopicpurities were greater than 99.9% for both analytes. Analytical assayinternal standards M + 9 cortisone and M + 9 cortisol (structuresshown in Scheme 2) were obtained from Dr. Kasuya’s laboratory,School of Pharmacy, Tokyo University of Pharmacy and Life Sci-

O

CH3

O

CH3

H

H

H

OHOH

12

35

14

8

9

7

12

11

10

13

6

4

15

1617

18

20

21

19

O

*

*

*

*

M+4 [1,2,4,19-13C4] CortisoneM+9 [1,2,4,19-13C4, 1,1,19,19,19-D5] Cortisone

(MW = 364)(MW = 369)

O

CH3

O

CH3

H

H

H

OHOH

12

35

14

8

9

7

12

11

10

13

6

4

15

1617

18

20

21

19

HO

*

*

*

*

M+4 [1,2,4,19-13C4] Cortisol

M+9 [1,2,4,19-13C4, 1,1,19,19,19-D5] Cortisol (MW = 371)

(MW = 366)

Scheme 2. Structures of SIL-cortisone and SIL-cortisol. M + 4 cortisone and cortisolwere used as the enzyme reaction substrate and product. M + 9 cortisone andcortisol were used as LC/MS/MS assay internal standards. *Locations of the 13C label.

216 Detection of 11b-HSD1 activity by microdialysis / L. Sun et al. / Anal. Biochem. 381 (2008) 214–223

ence. For in vivo experiments, perfusion fluid T1, which is a sterileisotonic fluid developed for use in peripheral tissue, was purchasedfrom CMA Microdialysis Inc. (North Chelmsford, MA). The T1 per-fusion fluid has an ion content of 147 mM Na+, 4 mM K+, 2.3 mMCa2+, and 156 mM Cl�. Ringer’s solution (Ringer’s Injection USP)was purchased from Bioanalytical Systems, Inc. (West Lafayette,IN), containing 147 mM Na+, 4 mM K+, 4.5 mM Ca2+, and 156 mMCl�.

In vitro microdialysis evaluations

The microdialysis pump (CMA 102) and microdialysis probes(CMA 60) were purchased from CMA Microdialysis Inc. (NorthChelmsford, MA 01863). The CMA 60 probe, sterilized at manufac-ture, has a polyamide membrane of 30 mm in length, with a molec-ular weight cutoff of 20 kDa. The microdialysis probe was placed inquiescent human plasma ultrafiltrate containing 10 ng/mL M + 4cortisone and M + 4 cortisol in a 37 �C water bath for the determina-tion of in vitro Ed at different flow rates. The perfusion fluid was Ring-er’s solution. The flow was driven by a CMA/102 microdialysis pumpwith a range of 0.3–2.0 lL/min. Dialysates were collected offline. A45-lL sample was collected at each time-point. Samples were storedfrozen at �70 �C. It should be noted that nonspecific binding ofhydrophobic cortisol or cortisone to the probe membrane and outlettubing was evaluated before the experiments were done, and the re-sults showed it was not a significant issue [30].

In vivo microdialysis

In vivo microdialysis was performed in the abdominal subcuta-neous adipose tissue of a female rhesus monkey (bodyweight 10kg) on 6 different days during a period of 6 weeks. All experimental

procedures and protocols were approved by the Institutional Ani-mal Care and Use Committee of Merck Research Laboratories andare consistent with the guidelines detailed in the NIH Guide forthe Care and Use of Laboratory Animals. In the morning of eachexperimental day, a new CMA 60 probe was inserted at the targettissue area after local anesthesia with 1% lidocaine, and secured byTegaderm tape. A portable CMA 106 microdialysis pump, pur-chased from CMA Microdialysis Inc. (North Chelmsford, MA), wasfixed to the monkey’s waist by a belt. To minimize the acute trau-ma effect caused by probe implantation, each microdialysis exper-iment started with a 60-min equilibration period with infusion ofperfusion fluid T1 at a flow rate of 1.0 lL/min. Dialysate was col-lected as a predose control, and these dialysates were used asblanks.

During the first week, in vivo Ed values of M + 4 cortisone andM + 4 cortisol were evaluated by retrodialysis. After the probewas implanted, perfusion fluid T1 was infused at a flow rate of1.0 lL/min. Following an equilibration period of 60 min, five dialy-sates were collected as endogenous control samples every 60 min.Then, using a manual liquid switch, a solution containing 5 ng/mLM + 4 cortisone and 5 ng/mL M + 4 cortisol prepared in Ringer’ssolution was infused at a flow rate of 1.0 lL/min and their concen-trations in the dialysates were measured. Given that the changes inin vivo M + 4 cortisone delivery Ed can provide a real-time indica-tion of mass transport variation at the probe vicinity [31], thisparameter was monitored throughout the following microdialysisexperiments.

Microdialysis experiments with M + 4 cortisone standard solu-tion prepared in Ringer’s solution as the perfusion fluid were con-ducted during the subsequent 5 weeks; details of the samplecollection are given in Table 1. First, T1 perfusion fluid was infusedas a pre-dose control for 60 min with concomitant dialysate collec-tion to determine endogenous analyte concentrations, after whichvarious M + 4 cortisone concentrations (100, 500, and 1000 ng/mL)and infusion flow rates (0.3, 0.5, and 1.0 lL/min) were tested forM + 4 cortisol production time–concentration profile.

There was a minimum washout period of 1 week between everytwo experiments to allow the monkey to recover and return to ba-sal physiological conditions, and for the clearance of M + 4 corti-sone and M + 4 cortisol. Probes from the same manufacturedbatch were used in these in vivo experiments to minimize probe-to-probe Ed variability.

Dialysates were collected offline and stored frozen at �70 �C.M + 4 cortisone, M + 4 cortisol, endogenous cortisone and cortisolwere monitored by LC/MS/MS. Storage stability at �70 �C wasevaluated.

Analytical method

SIL-cortisone and cortisol were measured by our published LC/MS/MS assay [30]. Briefly, the LC separation was performed on aWaters Symmetry C18 column (4.6 � 50 mm, 3.5-lm particle size)with a MetaTrap precolumn 2.0-lm pore size filter. The mobilephase was 65/35 (v/v) methanol/ammonium acetate (1 mM, pH6.7). An aqueous wash for 3 min with 0.1% (v/v) formic acid fol-lowed by an organic wash for 3 min with 50/50 (v/v) acetoni-trile/methanol was programmed after isocratic elution with themobile phase for 7 min. The mobile phase and the wash solutionswere delivered by Perkin-Elmer Series 200 LC pumps at a flow rateof 300 lL/min and 600 lL/min, respectively. A PE Sciex API 4000triple quadrupole tandem mass spectrometer was coupled withthe LC system as the analyte detector. The ion source was a turboion spray operated under positive ionization conditions. The desol-vation temperature was 500 �C, and the ionization potential was5000 V. The nebulizer gas, heater gas, and curtain gas were sup-plied at 30, 60, and 20 psi, respectively (1 psi � 6.9 kPa). The colli-

Table 1In vivo sample collection time-table for five experiments

Sequence ofexperiment

M + 4 cortisone(ng/mL)

Flow rate(lL/min)

Collectioninterval (min)

Dialysate nominalvolume (lL)

1 100 0.3 120 362 100 1.0 60 603 500 0.5 90 454 500 1.0 60 605 1000 0.5 90 45

There was a wash-out period of at least a week between experiments.


sion-activated dissociation (CAD) gas was set at 4. The ion transi-tions monitored under selected reaction monitoring (SRM) modeinclude m/z 365.4–167.2 for M + 4 cortisone, m/z 367.3–125.1 forM + 4 cortisol, m/z 361.4–163.2 for cortisone, m/z 363.3–121.1 forcortisol, m/z 370.2�172.2 for M + 9 cortisone, and m/z 372.2–129.0 for M + 9 cortisol. A sample volume of 30 lL was injectedonto the LC/MS/MS system using a Leap autosampler (CTC PAL).A diverter valve was installed post-LC column to divert the sub-stances eluted early to waste. The diverter valve was inline withthe MS instrument at 3.0 min and offline at 7.0 min. The lower lim-it of quantification (LLOQ) for this LC/MS/MS assay was 0.1 ng/ml,and the calibration range was 0.1–500 ng/mL for M + 4 cortisoneand M + 4 cortisol.

Standard solutions were prepared in Ringer’s solution andstored at �20 �C. For sample analysis, a sample volume of 30 lLwas added to a 96-well plate, followed by addition of 5 lL of inter-nal standard solution (10 ng/mL M + 9 cortisone and 20 ng/mLM + 9 cortisol) and 10 lL of 100 mM ammonium acetate (pH 4.0).Quality control (QC) samples with five nominal concentrations ofM + 4 cortisone and M + 4 cortisol (0.3, 3, 30, 90, and 300 ng/mL)were prepared in Ringer’s solution and stored frozen at �70 �C.For each analytical assay, two replicates of QC at each of the fiveconcentrations were thawed and processed together with standardand unknown samples to ensure sample stability and to evaluatethe reproducibility of the assay. The LC/MS/MS conditions werefound to be selective for analytes (endogenous and M + 4 cortisoneand cortisol) and internal standard (M + 9 analogs). Tissue blankdialysate samples and internal standard working solutions didnot contain detectable interferences at the retention times of cor-tisone or cortisol.

Results and discussion

LC/MS/MS assay and analyte stability

Fig. 1A and B show the product ion scan mass spectra of M + 4cortisone and M + 4 cortisol. Both analytes were quantifiable from0.1 to 500 ng/mL in the LC/MS/MS assay. No interference was ob-served for the analytes or their internal standards in the ion chro-matograms of dialysate samples collected from microdialysisexperiments in the adipose tissue, including the tissue blanks.Endogenous and SIL cortisone eluted at 3.5 min. Endogenous andSIL cortisol eluted at 4.2 min. In the M + 4 cortisol chromatogram(Fig. 1C), another two peaks were observed at 3.2 min and3.5 min, respectively, similar to the chromatographic peaks ob-served in the microsomal studies [30]. The peak at 3.5 min is be-lieved to be the isotopic peak from M + 4 cortisone, as it has thesame retention time as M + 4 cortisone, and it was present in theion chromatograms of high concentration standards. The impuri-ties in SIL compound standard materials could come from the iso-topic labeling synthesis procedures and from the natural isotopicdistribution. The mass-to-charge ratio (m/z) of M + 4 cortisol istwo units higher than the m/z of M + 4 cortisone. They also sharethe same fragment ion of m/z 125.1. It was not surprising to seethe isotopic peak of M + 4 cortisone in M + 4 cortisol ion chromato-

grams, given that the concentration of M + 4 cortisone was approx-imately from 100 to 500 ng/mL in the dialysate. The small peak at3.2 min, which appeared only after 180 min or later during micro-dialysis infusion of 500 and 1000 ng/mL SIL-cortisone, was possiblythe metabolite M + 4 20b-dihydroxycortisone formed from M + 4cortisone due to the activity of 20b-oxidoreductase [14]. Since itwas not a focus of the study and resolvable, exact confirmationof this metabolite was not investigated further.

The assay accuracy, precision, and reproducibility for in vivodialysate analysis were evaluated by interday quality control(QC) sample precision and accuracy, as shown in Table 2. ForM + 4 cortisone, the measured concentrations were between100.1% and 107.3% of the nominal concentrations, and the CVwas between 2.8% and 5.1%; for M + 4 cortisol, the accuracy rangewas from 96.3% to 101.7%, and CV was between 3.2% and 8.6%. Thelinear regression coefficient was greater than 0.998 and the cali-bration curve slope interday variation was less than 12%. M + 4 cor-tisone and M + 4 cortisol were stable for at least 8 h in Ringer’ssolution at room temperature and in plasma ultrafiltrate at 37 �C.The storage stability at �70 �C was also established, and the ana-lytes were stable for at least one month.

Microdialysis probe Ed evaluation

Table 3 shows the in vitro Ed evaluation of CMA 60 probe forM + 4 cortisone and M + 4 cortisol at perfusion flow rates 0.3, 1.0,and 2.0 lL/min. There was no significant difference between theEd of M + 4 cortisone and M + 4 cortisol observed for these similarcompounds. The Ed range was between 46.8% and 85.8%, with thehigher flow rates corresponding to the lower Ed, as expected. Withthis Ed range, M + 4 cortisone and M + 4 cortisol were expected tobe detectable in vivo within the known analytical assay calibrationrange of 0.1 to 500 ng/mL. However, it should be noted that in vivoEd could be lower than in vitro Ed due to the higher mass transferresistance caused by the low volume fraction of interstitial fluidand the tortuous mass transfer pathways in tissue [31].

The in vivo Ed was evaluated by retrodialysis for M + 4 cortisoneand M + 4 cortisol. The results given in Table 4 show that these twocompounds had similar in vivo delivery Ed values. The average Ed

values were 68.1% for M + 4 cortisone and 72.0% for M + 4 cortisol,respectively, implying their similar diffusivity in the adipose tissue.The in vivo Ed was approximately 10% lower than the in vitro Ed val-ues in Table 3. Assuming passive diffusion is the primary form ofmass transfer in these experiments, the difference in in vivo andin vitro Ed may reflect the difference in mass transfer resistancefor in vitro solution and in vivo tissue. This test served as a controlexperiment for the comparison with results obtained at higherdoses of M + 4 cortisone.

In vivo M + 4 cortisol formation by microdialysis infusion of M + 4cortisone: effect of flow rate and substrate concentration

To assess the effects of flow rate and substrate concentration,solutions containing 100, 500 and 1000 ng/mL M + 4 cortisonewere infused at 0.3, 0.5, and 1.0 lL/min, and corresponding M + 4cortisone and M + 4 cortisol profiles in the dialysates were ob-tained. The reported Km for 11b-HSD1 oxidoreductase activity isat approximately the micromolar (sub-lg/mL to lg/mL) level,and Vmax is 70 pmol�min�1�mg�1 in human liver microsomes[38]. Slightly higher affinity (lower Km) for cortisone was reportedfor intact cells and perfused liver, compared with in vitro micro-somal experiment results, due to the microenvironmentmaintained by the intact cells [39,40]. To ensure the productionof the metabolite at quantifiable concentrations, the infusionconcentration range for M + 4 cortisone was 100–1000 ng/mL(0.27–2.75 lM), close to the reported Km values. Fig. 2 shows

Fig. 1. Representative mass spectra of stable-isotope-labeled (SIL) M + 4 cortisone and cortisol. (A) Product ion scan mass spectrum of protonated M + 4 cortisone ([M + H]+

365.2 m/z). (B) Product ion scan mass spectrum of protonated M + 4 cortisol ([M + H]+ 367.3 m/z). (C) Extracted ion current chromatogram of M + 4 cortisol (m/z 367.3 to125.1) detected in the in vivo dialysate collected at 420 min when 500 ng/mL M + 4 cortisone was infused at 1.0 lL/min.


M + 4 cortisone concentration in dialysates and calculated Ed as afunction of infusion time. It appears that Ed varied during the

course of experiment; however, the magnitude of variationstended to decrease slowly. The criterion for Ed steady-state was de-

Fig. 1 (continued)

Table 2Summary table of inter-day QC to evaluate assay reproducibility and sample stabilityfor M + 4 cortisone and M + 4 cortisol

Nominal concentration (ng/mL)0.3 3 30 90 300

Measured concentration (ng/mL)M + 4 cortisone

Mean (n = 10) 0.32 3.00 30.91 95.63 312.79%Accuracy 107.3 100.1 103.0 106.3 104.3%CV 3.5 4.4 2.8 4.6 5.1

M + 4 cortisolMean (n = 10) 0.31 2.89 29.70 90.04 301.15%Accuracy 101.7 96.3 99.0 100.0 100.4%CV 8.6 4.7 5.3 5.7 3.2

QC samples were prepared at the start of the study and stored frozen at �70 �C.

Table 3Mean in vitro probe Ed (n = 3) at different perfusion flow rates measured using a CMA60 probe with analyte concentration of 10 ng/mL in human plasma ultrafiltrate

Flow rate (lL/min)

M + 4 cortisone %Ed

(%CV)M + 4 cortisol %Ed

(%CV)Pvaluea

0.3 85.8 (5.5) 79.2 (10.3) 0.29b

1.0 81.9 (9.8) 74.9 (11.2) 0.36b

2.0 55.1 (6.4) 46.8 (9.8) 0.07b

a Calculation based on comparison between %Ed of M + 4 cortisone and M + 4cortisol using unpaired Student’s t test.

b Values considered not significantly different (P > 0.05).

Table 4Evaluation of in vivo Ed for M + 4 cortisone and M + 4 cortisol at monkey adiposetissue

Time (min) M + 4 cortisone %Ed M + 4 cortisol %Ed

40 82.2 85.180 67.3 71.2120 68.9 72.7

Infusion fluid was 5 ng/mL M + 4 cortisone and M + 4 cortisol. Microdialysis flowrate was 1.0 lL/min. Samples were collected every 40 min.


fined as a variation of no more than the analytical CV of 10% inM + 4 cortisone concentrations obtained from at least threesequentially collected dialysates. At a flow rate of 0.5 lL/min, stea-

dy-state Ed was 72.7 ± 1.3% (n = 4) and 70.5 ± 2.0% (n = 4) for the500 and 1000 ng/mL M + 4 cortisone infusions, respectively. Whileat a flow rate of 1.0 lL/min, steady-state Ed was 58.7 ± 5.6% (n = 4)and 58.8 ± 4.5% (n = 5) for the 100 and 500 ng/mL infusions, respec-tively. For 100 ng/mL M + 4 cortisone infusion at a flow rate of0.3 lL/min, the mean Ed was 96.4 ± 1.2% (n = 3) after 240 min.Apparently, for the same flow rate, the steady-state Ed values werenot concentration-dependent for the tested conditions and werereproducible from day to day. Compared with the in vitro Ed in Ta-ble 3, there was about 25% decrease for in vivo Ed at 1.0 lL/min, anda 10% increase at 0.3 lL/min. Compared with the data in Table 4,the steady-state Ed values were slightly lower for 5 ng/mL infusion,but the difference was not statistically significant, suggesting nosignificant tissue clearance effect on the substrate steady-stateEd. The time for establishing steady-state Ed was generally long,consistent with the low clearance rate (10�2 � 10�3 min�1) forthe analytes [41]. However, it increased from 80 min in the 5 ng/mL infusion to a range of 180–270 min in the 100–1000 ng/mLinfusion. This variation could be partly due to the different infusionflow rates. The penetration distance of the substrate after infusionmay also have a role, as it could be greater at higher concentrationsdue to the possible enzyme saturation nearby the probe.

M + 4 cortisol formation profile confirmed the presence of 11b-HSD1 enzyme activity in the adipose tissue and showed time-dependent characteristics (Fig. 3). During the first 5 h, the M + 4cortisol concentration increased with time, suggesting continuousformation of M + 4 cortisol after the local delivery of M + 4 corti-sone. The kinetics of removal processes for M + 4 cortisol wereeither slow or negligible compared to the rate of M + 4 cortisol pro-duction. M + 4 cortisol production rate slowed, and reached a pla-teau or decreased after 5 h or later for most experiments. This canbe explained by the increased clearance rate of M + 4 cortisol, as aresult of its concentration increase following the infusion of M + 4cortisone.

The collected microdialysates containing M + 4 cortisol werehighly dependent on the initial M + 4 cortisone substrate concen-tration that was perfused through the probe. When the flow ratewas 0.5 or 1.0 lL/min, M + 4 cortisol concentrations in the dialy-sates increased when the substrate infusion concentration in-creased from 100 to 500 ng/mL. In contrast, at 0.5 lL/min, whenthe substrate concentration increased from 500 to 1000 ng/mL,

Fig. 2. Concentrations of M + 4 cortisone in the dialysates and M + 4 delivery Ed

when M + 4 cortisone (100, 500 and 1000 ng/mL) was infused at 0.3, 0.5, and 1.0 lL/min through the probe implanted at monkey adipose tissue. From �60 to 0 min,Ringer’s solution was infused. Infusion of M + 4 cortisone was started from 0 min.(A). The measured M + 4 cortisone dialysate concentrations. (B). The calculatedM + 4 cortisone delivery Ed.

Fig. 3. In vivo M + 4 cortisol formation at the monkey adipose tissue when M + 4cortisone was infused through the probe. M + 4 cortisol concentration was notcorrected by Ed. From �60 to 0 min, Ringer’s solution was infused. Infusion of M + 4cortisone was started from 0 min.


M + 4 cortisol production rate decreased twofold. Since there wasno significant difference in the Ed values between the 500 and1000 ng/mL infusions for the same flow rate, the decreased prod-uct formation rate probably indicates saturated or inhibited 11b-HSD1 enzyme activity due to an excess of substrate. A similarobservation was reported by Sandeep et al. [15]. Enzyme satura-tion for the tested substrate concentration was possible, and itwas possible also that other metabolic pathways were triggeredat 1000 ng/mL M + 4 cortisone infusion and may have contributedto the decrease in reaction rate.

When M + 4 cortisone was infused at 0.3 lL/min, M + 4 cortisolconcentrations were 0.27, 0.38, and 0.71 ng/mL at 240, 360, and480 min of M + 4 cortisone infusion, respectively, which weremuch lower than the concentrations at other flow rates. The con-centration kept increasing throughout the experimental period of8 h. Given that the 100 ng/mL M + 4 cortisone was delivered atsuch a low flow rate, and comparing with M + 4 cortisol profiles

from infusion experiments at much higher concentration (i.e.,500 ng/mL), it was unlikely that enzyme saturation occurred atthe probe area during the period of infusion. The low flow rate(i.e., 0.3 lL/min) was probably the major cause of the low concen-tration of M + 4 cortisol dialysate. Essentially, M + 4 cortisol con-centrations in the dialysates were affected by multiple dynamicevents at the probe area. First, the low infusion rate dictated lowmass delivery of the substrate M + 4 cortisone, which may resultin a low substrate tissue concentration after its tissue distribution,and consequently, a low production rate of M + 4 cortisol. Second,the lower flow rate corresponds to a longer analyte residence timein the probe lumen. Considering that the microdialysis probe com-petes with other analyte elimination pathways during the samplecollection, the long residence time of M + 4 cortisol in the probe in-creased the chances for it to diffuse out to the far field tissue whereits concentration was zero and/or be removed by other pathways(e.g., capillary and metabolic removal). Third, the axial distributionof M + 4 cortisol in the long probe (3-cm membrane length) couldalso result in diluted M + 4 cortisol concentrations in thedialysates.

The metabolism ratios for different experimental conditions aresummarized in Table 5, as determined from M + 4 cortisol dialysateconcentrations divided by the concentration loss of M + 4 cortisoneto the adipose tissue during microdialysis infusion. The metabo-lism ratio varied from 0.45 to 6.16%, indicative of the combined ef-fects of substrate delivery rate and sample collection rate on themetabolite collection in dialysates. Since the capillary removal rateis very low for the analytes [41], it is highly probable that the con-tribution of 11b-HSD1 activity to cortisone and cortisol concentra-tion profiles is detectable in dialysates. Our evaluation resultsshow a flow rate range of 0.5–1.0 lL/min is feasible. For compari-son, B.R. Walker’s group has reported two in vivo microdialysisstudies with [3H]cortisone infusion. In one study, 67 nM [3H]corti-sone was infused into human adipose tissue at 0.3 lL/min, and 10–15% conversion from [3H]cortisone to [3H]cortisol was detected[25]; while in another study in which 50 nM radiolabeled cortisonewas infused into human adipose tissue at 0.3 lL/min, the conver-sion ranged from 30 to 50% [24]. These infusion concentrationswere not examined for the monkey study in this report due to

Fig. 4. Endogenous cortisone and cortisol concentration in the dialysates when 0 to500 ng/mL M + 4 cortisone was infused by microdialysis at 1.0 lL/min. From -60 to0 min, Ringer’s solution was infused. Infusion of M + 4 cortisone was started from0 min. (A) Cortisone. (B) Cortisol.

Table 5Average mass delivery rate of M + 4 cortisone to the adipose tissue under different microdialysis conditions and the ratio of M + 4 cortisol collection and M + 4 cortisone deliveryas determined from microdialysis

[M + 4 cortisone] (ng/mL)/flow rate(lL/min)

Time of SSa establishment(min)

M + 4 cortisone delivery rated

(ng/min)Total M + 4 cortisonedelivered (ng)

Total M + 4 cortisolrecovered (ng)

Mfraction

(%)e

100/0.3 240b 0.027 (CV%, 16.7 ; n, 4) 12.8 0.05 0.45100/1.0 240c 0.060 (CV%, 16.4 ; n, 7) 25.2 1.16 6.16500/0.5 270c 0.178 (CV%, 11.5; n, 5) 80.0 1.66 2.80500/1.0 180c 0.290 (CV%, 14.5; n, 7) 121.9 2.61 2.711000/0.5 180c 0.346 (CV%, 4.81; n, 5) 155.5 0.68 0.52

a SS denotes M + 4 cortisone Ed steady state.b 33% CV in dialysate concentrations measured from 240, 360, and 480 min collections.c <10% CV in dialysate concentrations collected for three sequential collections of dialysates.d The rate of delivery was determined by dividing the absolute mass of M + 4 cortisone delivered to the tissue by the time of delivery.e Mfraction denotes the mean metabolism ratio, which was obtained by multiplying the ratio of the mean [M + 4 cortisol] in the dialysate and the mean [M + 4 cortisone] loss

from the probe to the tissue after steady-state Ed for M + 4 cortisone was established, by a factor of 100.


the concern that SIL-cortisol formation concentrations could fallbelow the limit of quantification.

The in vivo Ed of the collected M + 4 cortisol that is formed afterM + 4 cortisone infusion cannot be measured directly during theseexperiments. The M + 4 cortisol distribution profile surroundingthe probe would be heterogeneous, since this product had the po-tential to diffuse either toward or away from the probe, followingits concentration gradients. Experimental and modeling work hasbeen performed to define metabolite distribution profiles duringmicrodialysis infusion, and the correlation of metabolite concen-tration in the dialysate with kinetic parameters for tissue biologicalactivities [36,37]. Although it is difficult to define the absolute con-centration of M + 4 cortisol generated in the tissue at each selectedtime point, the relative changes in M + 4 cortisol production con-centrations obtained from the dialysates under control and duringpharmacological intervention can provide useful evidence for local11b-HSD1 enzyme activity. Potentially, when experimental condi-tions are carefully controlled, this assay could be used to test var-ious therapeutic modalities against 11b-HSD1 activity.

Cortisone and cortisol endogenous level

Theoretically, the reaction of SIL-cortisone would affect only thearea surrounding the probe, and 11b-HSD1 enzyme activity in theundisturbed tissue area would remain at a physiological level dur-ing microdialysis sampling. In order to confirm that the microdi-alysis probe implantation and M + 4 cortisone infusion did notcause any significant impact on the systemic circulating cortisoneand cortisol concentrations, endogenous cortisone and cortisolwere measured simultaneously with their SIL counterparts duringin vivo microdialysis. Only the results obtained at a flow rate of1 lL/min are plotted (Fig. 4) to exclude the effect of Ed variationat various flow rates. The cortisone and cortisol profiles obtainedby infusion of the perfusion fluid in the absence of M + 4 cortisonewere used as a control. The concentrations of cortisone and cortisolwere estimated from M + 4 cortisone and M + 4 cortisol standardcurves, based on the peak area ratios of cortisone and cortisol totheir corresponding analytical internal standard M + 9 cortisoneand M + 9 cortisol, assuming the responses of the MS instrumentto cortisone and cortisol are not significantly different from thoseto the M + 4 SIL counterparts. It is reasonable to make theseassumptions, since relative changes of endogenous cortisone andcortisol tissue concentrations were of major interest. Calculatedfrom the dialysate concentrations corrected by in vivo Ed obtainedby retrodialysis (Table 4), peak concentrations for cortisol and cor-tisone were 50 and 25 ng/mL, and trough concentrations were13.06 and 8.82 ng/mL, respectively. The reported endogenousvalues in human plasma are about 50–125 ng/mL for cortisol and

15–35 ng/mL for cortisone [42]. Fluctuations of endogenous con-centrations of cortisone and cortisol in dialysates could be causedby variations in microdialysis Ed and/or analyte tissue concentra-tions. In the beginning of microdialysis, analyte Ed fluctuated dueto variations of the diffusion and reaction rates. Diurnal variationhas been reported for cortisol in the rhesus monkey, caused mainly


by the episodic cortisol secretion from the adrenal cortex [43]. Theanalyte tissue concentration can be influenced by their distribu-tion, metabolism and elimination pathways. These physiologicalactivities might contribute to the higher concentrations of cortisolin the earlier stage of experiments (Fig. 4B). Endogenous concen-trations of both analytes were within a similar range of magnitudefrom day to day. Considering all the variables, it could be con-cluded that the microdialysis infusion of up to 1 lg/mL M + 4 cor-tisone for 7–8 h continuously did not cause any noticeableperturbation of systemic cortisone and cortisol balance or inhibi-tion of cortisol synthesis and/or secretion.

Conclusions

The in vivo microdialysis sampling of cortisone and cortisol wasdeveloped and used to study 11b-HSD1-catalyzed conversion ofM + 4 cortisone and M + 4 cortisol in the monkey adipose tissue.In vivo Ed values for both analytes were obtained by retrodialysis.M + 4 cortisone was infused locally through an implanted microdi-alysis probe to the monkey abdominal adipose tissue and the cor-responding M + 4 cortisol production profiles were obtained. Theeffects of the concentration and flow rate of substrate infusion onthe metabolite collection were determined. The probe was well tol-erated throughout the study. M + 4 cortisol formation was detectedin the monkey adipose tissue after infusion of 100, 500, and1000 ng/mL M + 4 cortisone at 0.3, 0.5, and 1.0 lL/min. The rateof conversion was influenced by the substrate delivery rate and en-zyme kinetic properties and by the microdialysis sampling param-eters. Endogenous cortisone and cortisol were also monitored bymicrodialysis. M + 4 cortisone infusion did not cause any signifi-cant change in the systemic concentrations of cortisone and corti-sol. This study demonstrates an example of using a microdialysisprobe as both an infusion device and a sampling device to studythe local tissue metabolism. This methodology could be appliedto in vivo 11b-HSD1 activity studies in tissues of interest in variousanimal species, including humans. Furthermore, the in vivo resultscould aid the study design of other local metabolism studies usingmicrodialysis infusion.

Acknowledgments

We are grateful for helpful discussions and support from Dr. Mi-chael Dobrinska, Dr. Eseng Lai, Dr. Julie Stone, and Dr. Larissa Wen-ning. We thank Michael Schwartz, Dr. Eric Woolf, and Dr. AllenJones for their comments on the manuscript.

References

[1] H. Masuzaki, J. Paterson, H. Shinyama, N.M. Morton, J.J. Mullins, J.R. Seckl, J.S.Flier, A transgenic model of visceral obesity and the metabolic syndrome,Science 294 (2001) 2166–2170.

[2] D.J. Wake, E. Rask, D.E.W. Livingstone, S. Soderberg, T. Olsson, B.R. Walker,Local and systemic impact of transcriptional up-regulation of 11b-hydroxysteroid dehydrogenase type 1 in adipose tissue in human obesity, J.Clin. Endocrinol. Metab. 88 (2003) 3983–3988.

[3] T.M. Stulnig, W. Waldhausl, 11b-Hydroxysteroid dehydrogenase type 1 inobesity and type 2 diabetes, Diabetologia 47 (2004) 1–11.

[4] R.S. Lindsay, D.J. Wake, S. Nair, J. Bunt, D.E.W. Livingstone, P.A. Permana, P.A.Tataranni, B.R. Walker, Subcutaneous adipose 11b-hydroxysteroiddehydrogenase type 1 activity and messenger ribonucleic acid levels areassociated with adiposity and insulinemia in Pima Indians and Caucasians, J.Clin. Endocrinol. Metab. 88 (2003) 2738–2744.

[5] I. Bujalska, M. Shimojo, A. Howie, P.M. Stewart, Human 11b-hydroxysteroiddehydrogenase: Studies on stably transfected isoforms and localization of thetype 2 isozeme within renal tissue, Steroids 62 (1997) 77–82.

[6] M. Quinkler, B. Kosmale, V. Bähr, W. Oelkers, S. Diederich, Evidence forisoforms of 11b-hydroxysteroid dehydrogenases in the liver and kidney of theguinea pig, J. Endocrinol. 153 (1997) 291–298.

[7] L. Schipper, B. Spee, J. Rothuizen, F.W. Nijnanten, J. Fink-Gremmels,Characterization of 11b-hydroxysteroid dehydrogenases in feline kidney andliver, Biochim. Biophys. Acta 1688 (2004) 68–77.

[8] J.R. Seckl, B.R. Walker, Minireview: 11b-Hydroxysteroid dehydrogenase type1—a tissue-specific amplifier of glucocorticoid action, Endocrinology 142(2001) 1371–1376.

[9] P. Putignano, F.P. Giraldi, F. Cavagnini, Tissue-specific dysregulation of 11b-hydroxysteroid dehydrogenases type 1 and pathogenesis of the metabolicsyndrome, J. Endocrinol. Invest. 27 (2004) 969–974.

[10] R. Best, B.R. Walker, Additional value of measurement of urinary cortisone andunconjugated cortisol metabolisms in assessing the activity of 11b-hydroxysteroid dehydrogenase in vivo, Clin. Endocrinol. 47 (1997) 231–236.

[11] R. Andrew, K. Smith, G.C. Jones, B.R. Walker, Distinguishing the activities of11b-hydroxysteroid dehydrogenases in vivo using isotopically labeled cortisol,J. Clin. Endocrinol. Metab. 87 (2002) 277–285.

[12] G.P. Chrousos, E. Charmandari, T. Kino, Glucocorticoid action networks—anintroduction to system biology, J. Clin. Endocrinol. Metab. 89 (2004) 563–564.

[13] G.P. Chrousos, Is 11b-hydroxysteroid dehydrogenase type 1 a good therapeutictarget for blockade of glucocorticoid actions?, Proc Natl. Acad. Sci. USA 101(2004) 6329–6330.

[14] S.M. Abel, D.J. Back, J.L. Maggs, B.K. Park, Cortisol metabolism by human liverin vitro—I. Metabolite identification and inter-individual variability, J. SteroidBiochem. Mol. Biol. 43 (1992) 713–719.

[15] T.C. Sandeep, J. Yau, A. MacLullich, J. Noble, I.J. Deary, B.R. Walker, J.R. Seckl,11b-Hydroxysteroid dehydrogenase inhibition improves cognitive function inhealthy elderly men and type 2 diabetics, Proc. Natl. Acad. Sci. USA 101 (2004)6734–6739.

[16] J.R. Katz, V. Mohamed-Ali, P.J. Wood, J.S. Yudkin, S.W. Coppack, An in vivostudy of the cortisol-cortisone shuttle in subcutaneous abdominal adiposetissue, Clin. Endocrinol. 50 (1999) 63–68.

[17] D.O. Scott, C.E. Lunte, In vivo microdialysis sampling in the bile, blood and liverof rats to study the disposition of phenol, Pharm. Res. 10 (1993) 335–342.

[18] K.V. Belle, S. Sarre, G. Ebinger, Y. Michotte, Brain, liver and blood distributionkinetics of carbamazepine and its metabolic interaction with clamipramine inrats: a quantitative microdialysis study, J. Pharmacol. Exp. Ther. 272 (1995)1217–1222.

[19] A.L. Freed, J.D. Cooper, M.I. Davies, S.M. Lunte, Investigation of the metabolismof substance P in rat striatum by microdialysis sampling and capillaryelectrophoresis with laser-induced fluorescence detection, J. Neurosci.Methods 109 (2001) 23–29.

[20] T. Kajiro, Y. Nakajima, T. Fukushima, K. Imai, A method to evaluate the rennin-angiotensin system in rat renal cortex using a microdialysis techniquecombined with HPLC-Fluorescence detection, Anal. Chem. 74 (2002) 4519–4525.

[21] C. Venero, J. Borrell, Rapid glucocorticoid effects on excitatory amino acidlevels in the hippocampus: a microdialysis study in freely moving rats, Eur. J.Neurosci. 11 (1999) 2465–2473.

[22] C.J. Cook, Measuring of extracellular cortisol and corticotrophin-releasinghormone in the amygdale using immunosensor coupled microdialysis, J.Neurosci. Methods 110 (2001) 95–101.

[23] J.W. Tomlinson, M. Sherlock, B. Hughes, S.V. Hughes, F. Kilvington, W. Bartlett,R. Courtney, P. Rejto, W. Carley, P. Stewart, Inhibition of 11b-hydroxysteroiddehydrogenase type 1 activity in vivo limits glucocorticoid exposure to humanadipose tissue and decreases lipolysis, J. Clin. Endocrinol. Metab. 92 (2007)857–864.

[24] T.C. Sandeep, R. Andrew, N.Z.M. Homer, R.C. Andrews, K. Smith, B.R. Walker,Increased in vivo regeneration of cortisol in adipose tissue in human obesityand the effects of 11b-hydroxysteroid dehydrogenase type 1 inhibitorcarbenoxolone, Diabetes 54 (2005) 872–879.

[25] D.J. Wake, N.Z. M Homer, R. Andrew, B.R. Walker, Acute in vivo regulation of11b-hedroxysteroid dehydrogenase type 1 activity by insulin and intralipidinfusions in humans, J. Clin. Endocrinol. Metab. 91 (2006) 4682–4688.

[26] K. Lanckmans, S. Sarre, I. Smolders, Y. Michotte, Use of a structural analogue versusa stable isotope labeled internal standard for the quantification of angiotensin IV inrat brain dialysates using nano-liquid chromatography/tandem massspectrometry, Rapid Commun. Mass Spectrom. 21 (2007) 1187–1195.

[27] S.N. Lin, J.M. Slopis, I.J. Butler, R.M. Caprioli, In vivo microdialysis and gaschromatography/mass spectrometry for studies on release of N-acetylaspartlyglutamate and N-acetylaspartate in rat brain hypothalamus, J.Neurosci. Methods 62 (1995) 199–205.

[28] K. Lanckmans, S. Sarre, I. Smolders, Y. Michotte, Quantitative liquidchromatography/mass spectrometry for the analysis of microdialysates,Talanta 74 (2008) 458–469.

[29] H. Shibasaki, T. Furuta, Y. Kasuya, Simultaneous determination of cortisol andcortisone in human plasma by stable-isotope dilution mass spectrometry, J.Chromatogr. 576 (1992) 47–52.

[30] L. Sun, J.A. Stenken, A.Y. Yang, J.J. Zhao, D.G. Musson, An in vitro microdialysismethodology to study 11b-hydroxysteroid dehydrogenase type 1 enzymeactivity in liver microsomes, Anal. Biochem. 370 (2007) 26–37.

[31] P.M. Bungay, P.F. Morrison, R.L. Dedrick, Steady-state theory for quantitativemicrodialysis of solutes and water in vivo and in vitro, Life Sci. 46 (1990) 105–119.

[32] C.C.D. Moore, S.H. Mellon, J. Murai, P.K. Siiteri, W.L. Miller, Structure andfunction of the hepatic form of 11b-hydroxysteroid dehydrogenase in thesquirrel monkey, an animal model of glucocorticoid resistance, Endocrinology133 (1993) 368–375.

[33] D.K. Hansen, M.I. Davies, S.M. Lunte, C.E. Lunte, Pharmacokinetic andmetabolism studies using microdialysis sampling, J. Pharm. Sci. 88 (1999)14–27.


[34] J.A. Stenken, Methods and issues in microdialysis calibration, Anal. Chim. Acta379 (1999) 337–358.

[35] M. Muller, R. Schmid, A. Georgopoulos, A. Buxbaum, C. Wasicek, H.G. Eichler,Application of microdialysis to clinical pharmacokinetics in humans, Clin.Pharmacol. Ther. 57 (1995) 371–380.

[36] J.A. Stenken, L. Stahle, C.E. Lunte, M.Z. Southard, Monitoring in situ livermetabolism in rats using microdialysis. Comparison of microdialysis mass-transport model predictions to experimental metabolite generation data, J.Pharm. Sci. 87 (1998) 311–320.

[37] J.A. Stenken, D.M. Holunga, S.A. Decker, L. Sun, Experimental and theoreticalmicrodialysis studies of in situ metabolism, Anal. Biochem. 290 (2001) 314–323.

[38] M. Hult, H. Jornvall, U.C.T. Oppermann, Selective inhibition of human type 111b-hydroxysteroid dehydrogenase by synthetic steroids and xenobiotics,FEBS Lett. 441 (1998) 25–28.

[39] M.L. Ricketts, K.J. Shoesmith, M. Hewison, A. Strain, M.C. Eggo, P.M. Stewart,Regulation of 11b-hydroxysteroid dehydrogenase type 1 in primary cultures ofrat and human hepatocytes, J. Endocrinol. 156 (1998) 159–168.

[40] N. Shafqat, B. Elleby, S. Svensson, J. Shafquat, H. Jornvall, L. Abrahmsen, U.Oppermann, Comparative enzymology of 11b-hydroxysteroid dehydrogenasetype 1 from glucocorticoid resistant (Guinea pig) versus sensitive (Human)species, J. Biol. Chem. 278 (2003) 2030–2035.

[41] Y. Kasuya, A. Yokokawa, S. Takayama, H. Shibasaki, T. Furuta, Evaluation of11b-HSD activities in vivo following oral administration of cortisol-13C4,2H1to a human subject, Steroids 68 (2003) 167–176.

[42] Y. Kasuya, A. Yokokawa, H. Shibasaki, T. Furuta, Measurement of cortisolsecretion rate by stable isotope methodology following oral administration of13 C-labeled cortisol to a human subject, Steroids 67 (2002) 783–788.

[43] J.W. Holaday, J.L. Meyerhoff, B.H. Natelson, Cortisol secretion and clearance inthe rhesus monkey, Endocrinology 100 (1977) 1178–1185.

an in vivo microdialysis coupled with liquid chromatography/tandem mass spectrometry study of...

Documents