Detection and quantification of cyclosporine in body fluids using an interleukin-2 reporter-gene assay

Journal of Immunological Methods 201 Ž1997. 125–135

Detection and quantification of cyclosporine in body fluids using an interleukin-2 reporter-gene assay Peter R. Wenner

a,)

, Franco Di Padova a , Paul A. Keown

b

a

b

Preclinical Research, Sandoz Pharma Ltd., CH-4002 Basel, Switzerland Immunology Laboratory, VancouÕer Hospital and HSC, VancouÕer, BC V5Z 1M9, Canada Received 25 March 1996; revised 12 September 1996; accepted 25 October 1996

Abstract Different assays are employed to monitor the concentration of immunosuppressive drugs in biological fluids. None of these methods gives direct and precise information on the actual level of immunosuppression in the patient. Here we describe the use of an interleukin-2 ŽIL-2. reporter-gene assay ŽIL-2 RGA. to monitor the concentrations of immunosuppressants in body fluids. This assay is based on a chimeric gene construct in which the human IL-2 promoter drives the expression of a reporter gene. Upon mitogenic stimulation the reporter gene is expressed and can be easily quantified. The assay is very sensitive and selective for immunosuppressive compounds inhibiting IL-2 gene expression such as cyclosporine ŽCsA. and FK506, their active metabolites and derivatives, but not for others such as rapamycin. High reproducibility, fast performance time, and high capacity are additional characteristics of the assay. The assay was developed to monitor immunosuppressive drug levels in human volunteers or in animals receiving CsA analogues as the only immunosuppressive drugs. This assay is sensitive to CsA or ascomycinrFK506 analogues and metabolites, for which there are presently no specific monoclonal antibodies available. The IL-2 reporter-gene assay may be more suitable than other in vitro systems such as MLR or mitogen stimulated PBMC which were previously used to study the immunosuppressive activity of drugs in body fluids. Keywords: Immunosuppressive; Therapeutic monitoring; Reporter-gene assay

1. Introduction Due to its immunosuppressive properties, CsA ŽBorel et al., 1976. has been successfully used to

Abbreviations: CsA, cyclosporine; IC 50 , 50% inhibiting concentration; MeLeu, methyl-leucine; MeVal, methyl-valine; RGA, reporter-gene assay; TDx, fluorescence polarization immunoassay. ) Corresponding author. At: Preclinical Research, 386r356, Sandoz Pharma Ltd., 4002 Basel, Switzerland. Tel.: 0041-61-3246876; Fax: 0041-61-324-2990.

prevent allograft rejection and to treat several autoimmune diseases ŽFeutren, 1992.. More recently FK506 ŽKino et al., 1987. has been introduced for the same purposes. Both drugs exhibit side effects which necessitate the constant monitoring of their blood levels. The risk of over- or underdosing may be reduced in the case of CsA with the introduction of Sandimmune Neoral, due to its improved oral bioavailability, dose linearity ŽMueller et al., 1994., and reduced inter- and intraindividual variability in CsA pharmacokinetics ŽKovarik et al., 1994.. But even with this formulation it is difficult to predict the

0022-1759r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 2 2 - 1 7 5 9 Ž 9 6 . 0 0 2 1 9 - 0

126

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135

blood level of CsA on the basis of dose alone ŽHolt and Johnston, 1994.. Several highly specific analytical methods have been developed for the quantification of CsA and their metabolites in whole blood. These include high-performance liquid chromatography ŽHPLC., radioimmunoassay ŽRIA., fluorescence polarization immunoassay ŽFPIA. and enzyme multiplied immunoassay technique ŽEMIT. ŽNapoli and Kahan, 1991.. The specific HPLC method is based on the physicochemical properties of the drug and provides information about the concentrations of the parent compound and each metabolite, but requires expensive equipment, time consuming extraction procedures and considerable technologic expertise. RIA, FPIA and EMIT are all based on the use of different monoclonal antibodies ŽmAbs. or antisera. mAbs offer the advantage of higher reproducibility and have replaced polyclonal antisera. In this setting these tests are fast and require minimal amounts of the biological sample. In addition the equipment required is readily available in most clinical laboratories. However these methods are not suitable for determining the biological activity of parent compounds nor can they distinguish between pharmacologically active and inactive metabolites, which are often difficult to quantify. For these purposes, a pharmacodynamic assay is required which provides a reliable measurement of the biological effect of the immunosuppressive drugs employed. The use of previous biological assays such as the mixed lymphocyte reaction ŽMLR. or mitogen-stimulated peripheral blood mononuclear cells, has been limited due to long performance times and the poor reproducibility of the results due to variability in responsiveness of cells obtained from different donors ŽSpiers and Beck, 1991.. The goal of this study was to set up and characterize a sensitive and reproducible biological test system which is able to measure the overall immunosuppressive activity of the parent compound and its metabolites in ex vivo blood samples. Since CsA and FK506 are known to inhibit IL-2 gene expression at the level of mRNA transcription ŽKroencke et al., 1984; Schreiber and Crabtree, 1992. IL-2 RGAs have been employed to investigate the effects of both substances ŽMattila et al., 1990.. The IL-2 RGA presented here is based on a chimeric

gene construct which is stably integrated into a human leukemic T cell line ŽJurkat.. In this construct the human IL-2 promoter fragment extending from position y583 to q40 Žwith respect to the start site of transcription. upon mitogenic stimulation drives the expression of the reporter-gene luciferase. It has been shown by transient expression experiments that this promoter fragment is sufficient to confer activation and CsA-mediated repression of a fused reporter gene in mitogen-stimulated T cells ŽBaumann et al., 1991.. The assay is very sensitive to immunosuppressive substances affecting IL-2 gene transcription, and detects both the parent compound and the metabolites with immunosuppressive activity. The stable cell line used in this assay can be easily cultured in any laboratory with common cell culture facilities, and the test can be performed within a working day. Here we describe the use of this assay for the evaluation of human whole blood and serum samples.

2. Material and methods 2.1. Reagents The immunosuppressive compounds CsA and its main metabolites AM1, AM9, and AM4N Žnomenclature see Consensus Document, 1990., as well as CsG ŽNva2-cyclosporine., SDZ IMM 125 Žhydroxyethyl derivate of D-serine 8-cyclosporine., FK506 and rapamycin, and the non-immunosuppressive compounds CsH ŽD-MeVal11 -cyclosporine. and SDZ 220–384 ŽMeVal 4-cyclosporine. were used as reference compounds. The anti-CsA mAb 45-45-11 used was described by Quesniaux et al. Ž1987.. All reagents were obtained from Sandoz Pharma. 2.2. Cyclophilin binding assay The solid phase enzyme immunoassay ŽELISA. was performed as described previously ŽQuesniaux et al., 1990.. Briefly, a D-Lys 8-CsA derivative was coupled to BSA and coated on to a polyvinyl microtiter plate Ž1 m grml in PBS. for 2 h at 378C. After saturation of the plate with 2% BSA in PBS Ž1 h at 378C. and washing once with 0.05% Tween 20 containing PBS and three times with PBS, human

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135

recombinant cyclophilin, expressed in Escherichia coli, was added in different concentrations Ž25–400 ngrml. in 1% BSA, 5 mmolrl 2-mercaptoethanolcontaining PBS, and incubated overnight at 48C. Bound cyclophilin was detected with anti-cyclophilin rabbit antiserum Ždiluted 25 000-fold in 1% BSA-PBS and incubated for 2 h at 378C. followed, after washing, by anti-rabbit globulin goat IgG coupled to alkaline phosphatase ŽSigma, 1r1000 in 1% BSAPBS, 2 h at 378C.. The absorbance at 405 nm was measured after hydrolysis of p-nitrophenyl phosphate Ž1 mgrml in diethanolamine buffer, pH 9.6, for 1–2 h at 378C.. 2.3. Cell lines, DNA constructs, transfection All standard DNA manipulations in the construction of the luciferase ŽAmerican firefly, Photinus pyralis . plasmid were carried out as described ŽSambrook et al., 1989.. Briefly, the RsaI restriction fragment from position y583 to q40 of the human IL-2 promotor was derived by HindIII digestion of placZH-IL-2 ŽBaumann et al., 1992.. The fragment was gel-purified and subsequently cloned into the HindIII site of pGL2-basic ŽPromega, aE1614.. In addition the neomycin resistance cartridge from pMC1neoPolyA ŽStratagen, a213201. containing the Herpes simplex thymidine kinase promotor was introduced for mediating G418 resistance of stably transfected cells. Prior to transfection, the pIL2rLUC plasmid was linearized using the ScaI site at position 6106. Jurkat cells Ž10 7 ., a human leukemic T cell line Žsubclone K16., were electroporated in 1 ml serum free medium ŽRPMI 1640. with 20 m g linearized plasmid DNA using a BioRad Gene Pulser Ž260 V, 950 m F.. Immediately after electroporation cells were suspended in RPMI-Glutamax I ŽGibco. q 10% FCS medium and incubated for 24 h Ž378C, 5% CO 2 . in a 75 cm2 vented culture flask ŽCostar.. The next day, the cells were counted and adjusted to 5 = 10 4rml, 10 4rml and 5 = 10 3rml. Cells were cultivated in 48-well Costar plates Ž1 mlrwell. adding 0.8 mgrml G418 Žhalf dose. as a selection marker. After 2 days, 250 m l of selection medium containing 1.6 mgrml G418 Žfull dose. were added. After 7–10 days the G418 dose was reduced to 0.8 mgrml. Growing clones were screened for high responsiveness to mitogen-induced expression ŽPMA,

127

40 ngrml, PHA, 2 m grml. of the reporter gene luciferase. Cells from selected wells were cloned in 96-well flat bottom plates ŽCostar.. 2.4. Human whole blood specimens Whole blood samples were obtained from healthy volunteers receiving cyclosporine ŽMueller et al., 1994. or from renal transplant patients undergoing immunosuppressive therapy ŽImmunology Laboratory, Vancouver Hospital and HSC.. All patients received combination therapy with CsA, azathioprine and prednisone. For the quantification of cyclosporine in whole blood, standard curves were established by spiking human drug-free whole blood or serum Žpooled from six individuals, using EDTA as anti-coagulant. with different concentrations of cyclosporine Žin ethanolic solution. in the range of 0.003–30 m M. Samples were diluted to 1% and 0.3% respectively. 2.5. Bioanalytical methods Concentrations of CsA in whole blood from volunteers were assayed using the commercially available Sandimmune radioimmunoassay ŽRIA. for CsA ŽSandoz, Basel., which is based on the use of a monoclonal antibody specific for the parent drug ŽBall et al., 1988.. Concentrations of CsA in coded spiked or patient whole blood or serum samples were determined using a fluorescence polarization immunoassay ŽTDx assay. ŽTDx monoclonal whole blood, Abbott Laboratories, Chicago, IL, USA.. 2.6. IL-2 reporter-gene assay (RGA) procedure Serum or whole blood samples ŽEDTA 0.1% as anticoagulant. were kept at y208C until they were investigated. Immediately after thawing aliquots of 100 m l were heat inactivated at 568C for 30 min. Samples were subsequently diluted to 10% or 3% with IMDM-ATL medium ŽSchreier and Tees, 1981.. Tests were performed in 96-well plates with a total volume of 200 m lrwell. 20 m l of diluted whole blood or serum were added to each well to a final concentration of 1% or 0.3% respectively. 5 = 10 4 cellsrwell in a total volume of 200 m l were stimulated with PMA Ž40 ngrml, final dilution. and iono-

128

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135

mycin Ž2 m M. and incubated either over a period of 5 h or overnight Ž16 h., at 378C in a mixture of air and 5% CO 2 . After stimulation plates were centrifuged for 10 min Ž700 = g . and supernatants removed. Cells were lysed in lysis buffer containing 25 mM Tris-phosphate ŽpH 7.8., 2 mM DTT, 2 mM 1,2-diaminocyclohexane-N, N, N X , N-tetraacetic acid, 10% Žvrv. glycerol and 1% Žvrv. Triton X-100. 20 m l was added to each well and plates were shaken for 10 min on a Titertek shaker. The luciferase activity was detected, due to its ability to catalyze the oxidative decarboxylation of D-luciferin resulting in the production of light. The amount of light produced was quantified in a Luminoskan ŽLabsystems. and expressed as relative light units. In this instrument 50 m l of luciferase assay reaction buffer is automatically added to a final concentration of 20 m M T ric in e , 1 .0 7 m M ŽMgCO 3 .4 )MgŽOH. 2 )5H 2 O, 2.67 mM MgSO4 , 0.1 mM EDTA, 33.3 mM DTT, 270 m M coenzyme A, 470 m M Luciferin ŽChemie Brunswig., 530 m M ATP. 2.7. Calculations and statistical analysis All statistical analysis was done using the Origin software ŽMicroCal Software, Northampton, MA, USA.. Blood samples to be investigated were mea-

sured with the IL-2 RGA on two occasions in quadruplicates. Inhibition values obtained by the IL-2 RGA using the CsA standard samples were fitted using a 4-parameter logistic function. The weighting factor was reciprocal to the observed Y values. Inhibition values obtained by the IL-2 RGA testing the coded samples were converted to concentration values using the standard curve equations. CsA concentrations in whole blood and serum as determined by the TDx assay and the IL-2 RGA were correlated using linear regression. With this procedure the regression line parameters, slope and intercept, were estimated.

3. Results 3.1. Characteristics of the IL-2 RGA The stably transfected human leukemic T cell line ŽJurkat, clone 41-19. is sensitive to the effect of CsA ŽIC 50 f 3.5 nM. and IMM 125 ŽIC 50 f 4.0 nM., which both exhibit comparable immunosuppressive activity in vitro ŽBaumann et al., 1992. and in vivo ŽMueller et al., 1994.. It is also susceptible to the effect of CsG ŽIC 50 f 5.0 nM. and FK506 ŽIC 50 0.3 nM. as shown in Fig. 1. The synthetic main CsA

Fig. 1. Effect of different substances on the IL-2 RGA. Serial dilutions of all substances ŽCsA, I; SDZ IMM 125, y; CsG, `; CsH, =; AM1, q; AM9, D; AM4N, ); FK506, =; and rapamycin, e. were prepared in IMDM-ATL medium. The values for the maximum signal and the background in the absence of compounds were 180.41 Ž"12.2. and 0.32 Ž"0.06. respectively. These values were consistent in all experiments.

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135

metabolites AM1 ŽEberle and Nuninger, 1992., AM9, and AM4N ŽWenger et al., 1992. showed immunosuppressive activities with an IC 50 of f 28 nM, f 76 nM, and 209 nM respectively. No response was observed with the non-immunosuppressive CsH ŽIC 50 ) 100 nM. or with rapamycin ŽIC 50 ) 100 nM.. With this cell line the signal noise ratio Žratio between the maximum- and the background-signal intensity measured in relative light units. was around 550 in the absence of inhibiting substances.

129

3.2. Specificity of the assay To demonstrate the specificity of the test system, different concentrations of a CsA-specific monoclonal antibody were added to rising concentrations of CsA thus preventing the penetration of CsA into the cell. A dose-dependent inhibitory effect was observed ŽFig. 2a.. These data were confirmed with a nonimmunosuppressive CsA analogue SDZ 220– 384 in which the amino acid MeLeu in position 4 is exchanged for MeVal. SDZ 220–384 binds to cy-

Fig. 2. a: blocking effect of a CsA antibody in the IL-2 RGA. Rising concentrations Ž0.03 m g, `; 0.1 m g, D; 0.3 m g, =; 1 m g, e; and 3 m g, q. of a CsA antibody ŽmAb 45-45-11. reverses the inhibition of luciferase expression caused by a serial dilution of CsA Žsubstance alone, I.. b: competition by SDZ 220–384 in the IL-2 RGA. Addition of increasing concentrations of SDZ 220–384 Žsubstance alone, =; 100 nM, `; 200 nM, D; 400 nM, =; 800 nM, e; 1600 nM, q. to serial dilutions of CsA Žsubstance alone, I. leads to a parallel shift of the CsA dose response curve, indicating competitive antagonism.

130

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135

clophilin with an affinity similar to that of CsA but has no inhibitory effect even at a concentration of 1000 nM; CsA already causes a 100% inhibition of reporter-gene expression at a concentration of 30 nM. As shown in Fig. 2b, rising concentrations of the antagonist SDZ 220–384 can reverse the inhibitory effect of CsA in a dose dependent fashion. 3.3. Effect of whole blood or plasma on the IL-2 RGA

Table 1 Effect of heat inactivation on the ability of cyclophilin to bind to CsA tested in an ELISA Concentration a Žngrml.

Cyclophilin A not heated

Cyclophilin A 568C, 30 min

0 25 50 100 200 400

0.14 1.47 2.11 2.49 2.67 2.61

0.14 0.14 0.13 0.13 0.12 0.12

a

Before studying the effect of immunosuppressive compounds in biological fluids we analyzed the direct effect of whole blood and serum. As mentioned in Section 2 all specimens were kept frozen at y208C until they were investigated. In the case of whole blood freezing and thawing causes lysis of erythrocytes resulting in the release of intracellular cyclophilin, the receptor which mediates the biological activity of CsA. To investigate whether heat inactivation is sufficient to prevent binding of CsA to this extracellular cyclophilin we used an ELISA ŽQuesniaux et al., 1990. in which cyclophilin can bind to a CsA derivative conjugated to BSA and coated on to a solid phase. As shown in Table 1 the interaction between cyclophilin and the CsA conjugate was inhibited after heat inactivation of cyclophilin for 30 min at 568C. Despite heat inactivation, whole blood or serum samples from normal donors at a final dilution of 3%

Amount of cyclophilin A added to a CsA derivative coupled to BSA coated on to a solid phase. The reaction was revealed with anti-cyclophilin antiserum followed by anti-rabbit globulin enzymatic conjugate. Results are expressed in absorbance units Ž405 nm..

and 10% causes almost complete inhibition of the reporter-gene activity Ždata not shown.. At a final dilution of 0.3% and 1% the inhibitory effect of whole blood is reduced to a point where immunosuppressive activity can be easily detected. Moreover at these sample dilutions the inhibitory effect of the anticoagulant EDTA is abolished. The Ca2q chelating agent EDTA inhibits the reporter-gene expression with an ED50 of 0.045%, which is well above the final concentration Ž0.001%. present in a 1r100 dilution of blood.

Fig. 3. Effect of whole blood samples on the IL-2 RGA. Whole blood samples from a healthy male volunteer were taken at different time points after the oral administration of 200 mg CsA. The inhibitory effect of all whole blood samples on the IL-2 RGA ŽI. is compared to the pharmacokinetic data Ž`. as determined by RIA measurements.

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135

3.4. Detection of immunosuppressiÕe actiÕity in whole blood An important clinical application of the IL-2 RGA could be the direct measurement of immunosuppressive activity in blood samples. To examine this, we used the IL-2 RGA for the detection of immunosuppressive activity in whole blood samples from a tolerance study ŽMueller et al., 1994. where a single oral dose of CsA Ž200 mg. was administered to healthy male volunteers. Whole blood samples were diluted to 0.3%. Results were expressed as percent inhibition Ž%. and were compared with the pharmacokinetic profile as determined by RIA. Data obtained from samples of a single representative volunteer are shown in Fig. 3. Similar results have been obtained with other samples from a tolerance study where a single dose of SDZ IMM 125 was administered either i.v. or orally to healthy male volunteers Ždata not shown.. As seen in Fig. 3, both pharmacokinetic and pharmacodynamic profiles overlap. Both curves peak simultaneously 2 h after CsA administration ŽRIA s 1291 nM, IL-2 RGA s 33% inhibition..

131

3.5. Quantification of CsA in whole blood To determine the concentration of CsA, whole blood from six healthy donors was pooled and spiked with different concentrations of the compound. The samples were diluted to 0.3% and 1% respectively. As expected the CsA concentration range is affected by the whole blood dilution factor. As shown in Fig. 4 the part of the 1% and the 0.3% whole blood standard curve which can be used to calculate the sample values range from 35% to 90% and from 50% to 90% respectively and cover a concentration range between 75 nM and 2000 nM and between 25 nM and 750 nM respectively in the original sample. Both reference inhibition curves were fitted using a 4-parameter logistic function. In this study at the dilution of 0.3% and 1% whole blood alone inhibits the reporter-gene expression by 27% and 48% respectively. Higher dilution factors diminish the level of interference by biological fluids, but the assay becomes rather insensitive Žsee below.. The characteristic of both curves remains the same. The slopes Ž0.3% s 1.20 Ž"0.13. and 1% s 1.30 Ž"0.11. are almost identical.

Fig. 4. CsA whole blood standard curve. Whole blood from six volunteers was pooled and spiked with different concentrations of CsA. Standard samples were diluted to 1 % ŽI. and 0.3% Ž`. respectively and measured on three different occasions in quadruplicates. The equation for both curves was Y s Ž27.19 y 100.rŽ1 q Ž Xr1.66.1.23 . q 100 Ž0.3% curve. and Y s Ž48.34 y 100.rŽ1 q Ž Xr1.65.1.37 . q 100 Ž1% curve., respectively.

132

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135

Coded whole blood samples investigated in this study included standard samples spiked with different concentrations of CsA, sensitivity samples which contained no drug, as well as samples from transplanted patients undergoing immunosuppressive treatment. Linearity samples were produced by repeatedly diluting a patient sample 1:1. The concentrations determined with the IL-2 RGA for the standard samples were significantly correlated with the results obtained by the TDx assay Ž0.3% whole blood Ž r s 0.978; P - 0.005., 1% whole blood, Ž r s 0.998; P - 0.005.. In contrast to previously tested standard samples, which were still inhibitory even in the absence of inhibitory concentrations of CsA, all sensitivity samples failed to inhibit the IL-2 RGA. This might be due to different storage conditions or differing concentrations of preservatives. Measurements of CsA concentrations in whole blood samples from patients Ž n s 20. by the IL-2 RGA and by the TDx assay were linearly correlated Ž P - 0.05.. Fig. 5 ŽA q B. displays the regression lines for both the 0.3% and the 1% dilution group.

The IL-2 RGArTDx assay ratios for all patient samples in both dilution groups were closely scattered around the line of unity ŽFig. 5 C q D. with a slight positive bias by the IL-2 RGA at CsA concentrations below 300 nM. This was mainly observed in the 0.3% group. 3.6. Quantification of CsA in serum To prove that a matrix other than whole blood can be employed in the IL-2 RGA, we tested coded serum samples which included spiked as well as patient specimens. Reference curves were prepared as described above for whole blood. Regression analysis using data obtained with patient serum samples gave the following equations for the 0.3% ŽA. and 1% ŽB. dilution group respectively. A: Y s y71.318 q 1.347) X Ž r s 0.883; n s 14; P - 0.005. and B: Y s y48.964 q 0.991) X Ž r s 0.972; n s 23; P 0.005.. CsA concentrations as measured by the TDx assay were, on average, lower in serum samples than they were in whole blood samples Ždata not shown..

Fig. 5. Relationship between concentrations of CsA in blood as measured by the TDx assay and the IL-2 RGA in patient samples that were diluted to 0.3% Ž A. and 1% Ž C . The equations for both curves were Y s y46.874 q 0.843) X Ž r s 0.957. and Y s 9.034 q 0.9346) X Ž r s 0.954., respectively. Assay ratios over the whole concentration range are plotted for the 0.3% Ž B . and the 1% Ž D . dilution group. The line of unity is drawn for better comparison.

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135

4. Discussion In this study we used a stably transfected cell line carrying a plasmid where the reporter-gene luciferase is driven by the human IL-2 promoter. Mitogenic stimulation leads to the expression of luciferase whose enzymatic activity can be measured due to its ability to catalyze the oxidative decarboxylation of D-luciferin and which results in the production of light. Testing different cyclosporines, we confirmed the results of Baumann et al. Ž1992. that CsA and SDZ IMM 125 equipotently inhibit the reporter-gene activity in a dose dependent manner. The main metabolites of CsA ŽAM1, AM9, and AM4N. showed reduced activities compared to the parent compound but were readily detectable. The IC 50 determined with the IL-2 RGA correlates well with the results of Freed et al. Ž1991. who determined the immunosuppressive activity of the metabolites AM1 ŽIC 50 f 42 nM. and AM9 ŽIC 50 f 83 nM. by measuring the inhibition of IL-2 production in Jurkat cells. The immunosuppressive activity of CsG is reduced by a factor 2 compared to CsA, due to a modification in position 2 of the CsA molecule which decreases the affinity of CsG to human cyclophilin ŽSchneider et al., 1994.. CsH ŽD-MeVal11 . does not bind to cyclophilin, since the modification in position 11 affects its binding domain, and consequently shows no inhibitory effect on the RGA. FK506, which inhibits the expression of early genes ŽSiekierka and Sigal, 1992., shows immunosuppressive activity in the IL-2 RGA that is 100-fold more potent than the latter drug. These findings confirm the observations by Bolton Ž1992. who compared the immunosuppressive activity of FK506 and CsA in an in vitro model. Although rapamycin and its structural analog FK506 bind to the same immunophilin ŽFKBP., rapamycin acts at a later stage in T cell cycle progression by blocking cytokinemediated signal transduction pathways ŽSehgal, 1995. and does not inhibit reporter-gene expression. For accurate quantification of immunosuppressive substances in body fluids it is important that the observed inhibition of the reporter gene is due to the specific activity of the immunosuppressant and not related to nonspecific activities of other interfering substances. Therefore we added different concentrations of an anti-CsA monoclonal antibody ŽmAb

133

45-45-11. to a serial dilution of CsA. Increasing concentrations of the antibody were shown to reverse the inhibition of the reporter-gene activity mediated by CsA due to the trapping of CsA outside the cell thus preventing the formation of the cyclophilinandror calcineurin-substrate complex. An additional method to demonstrate the specificity of the test system was the use of the CsA analogue SDZ 220– 384. This molecule shows the same affinity for cyclophilin as CsA but when complexed with cyclophilin fails to bind to calcineurin due to the modification in position 4 of the molecule and is consequently devoid of immunosuppressive activity. Titration of increasing concentrations of SDZ 220– 384 to serial dilutions of CsA shifts the inhibition curves to the right without affecting the slopes. This behavior indicates classical competitive antagonism, where rising concentrations of the competitor ŽSDZ 220–384. prevent binding of CsA to cyclophilin ŽZenke et al., 1993.. Whole blood has been recommended by different authors as the matrix of choice for the determination of CsA blood levels ŽShaw, 1989; Consensus Document, 1990., because the distribution of CsA between the plasma Žor serum. and blood cells is temperature- and time-dependent and can thus result in variations in plasma and serum concentrations ŽAnnesley et al., 1986.. Furthermore, higher concentrations in whole blood results in higher signal : noise ratios in the drug measurements ŽVine and Bowers, 1987.. Repeated freezing and thawing of whole blood samples results in the lysis of erythrocytes, with the release of intracellular stored cyclophilin, which can be inactivated by heating the samples to 568C for 30 min as shown by a solid phase enzyme immunoassay ŽQuesniaux et al., 1990.. The latter point is of particular importance, since the introduction of active cyclophilin into the test system could interfere with the outcome of the assay by competing with intracellular cyclophilin for binding of the immunosuppressive drug. As shown in this study, even after heat inactivation final dilutions of whole blood in the order of 3–10% cause an almost complete suppression of the IL-2 RGA. The factors responsible for this effect are presently unknown. Concentrations of whole blood in the order of 0.3–1% permit a proper use of the IL-2 RGA, although at the expense of a

134

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135

lower sensitivity. At these dilutions the concentration of endogenous and exogenous inhibitors such as EDTA is lowered to the point where they do not influence the test system. We have shown that the IL-2 RGA measurements mirror the pharmacokinetic data obtained by RIA. As we did not detect any hysteresis we can exclude the possibility that detectable amounts of immunosuppressive metabolites are either not produced after a single oral administration or that the time point after which metabolites can be detected in blood is beyond the period of observation Žhere 48 h.. Similarly we have used the IL-2 RGA to demonstrate the appearance of immunosuppressive blood levels in rats treated with different CsA analogues Ždata not shown.. Moreover, in testing coded samples spiked with CsA or coded samples from patients under CsA therapy, we found a good correlation between results obtained by TDx assay and the IL-2 RGA. The concentration range covered by the IL-2 RGA Ž25– 2000 nM. is valid for measurement of CsA concentrations in clinical samples since the median blood CsA concentration can range from - 100 nM to almost 1000 nM ŽHolt and Johnston, 1992.. All concentrations determined with the IL-2 RGA in whole blood patient samples correlated well with the concentrations obtained by the TDx assay, while the correlation coefficient was always above 0.9. This was true also for the results obtained using serum samples from patients, although the correlation coefficient for the 0.3% dilution group was slightly lower than 0.9 Ž0.883.. These results are an important improvement over previous studies comparing the concentrations of CsA and its metabolites determined by a human MLR and the TDx assay, which found no satisfactory correlation between these pharmacodynamic and pharmacokinetic assays using either a specific monoclonal antibody to CsA Ž r s 0.612. or a monoclonal with broader specificity Ž r s 0.673. ŽRussell et al., 1991..

Acknowledgements The authors are grateful to Jane Glenn ŽVancouver General Hospital. for excellent technical assistance, Dr. Edgar Muller for providing the volunteer whole ¨

blood samples, to Dr. Valerie Quesniaux for ELISA testing, and to Dr. Gotz ¨ Baumann Žall Sandoz Pharma Ltd.. for the contribution of the reporter-gene constructs and for helpful discussion.

References Annesley, T.M., Giacherio, D.A. and Feldkamp, C.S. Ž1986. The concentratio-related distribution of cyclosporine in blood. J. Clin. Immunoassay 9, 53. Ball, P.E., Munzer, H., Keller, H.P., Abisch, E. and Rosenthaler, J. Ž1988. Specific 3 H radioimmunoassay with a monoclonal antibody for monitoring cyclosporine in blood. Clin. Chem. 34, 257. Baumann, G., Geisse, S. and Sullivan, M. Ž1991. Cyclosporin A and FK-506 both affect DNA binding of regulatory nuclear proteins to the human interleukin-2 promoter. New Biol. 3, 270. Baumann, G., Andersen, E., Quesniaux, V. and Eberle, M.K. Ž1992. Cyclosporine and its analogue SDZ IMM 125 mediate very similar effects on T-cell activation – a comparative analysis in vitro. Transplant. Proc. 24, 43. Bolton, C. Ž1992. The efficacy of cyclosporin A, FK-506 and prednisolone to modify the adoptive transfer of experimental allergic encephalomyelitis ŽEAE.. Agents Actions 35, 79. Borel, J.F., Feurer, C., Gubler, H.U. and Stahelin, H. Ž1976. ¨ Biological effect of cyclosporin A: a new antilymphocytic agent. Agents Actions 6, 468. Consensus Document: Ž1990. Hawk’s Cay meeting on therapeutic drug monitoring of cyclosporine. Transplant. Proc. 22, 1357. Eberle, M.K. and Nuninger, F. Ž1992. Synthesis of the main metabolite ŽOL-17. of cyclosporin A. J. Org. Chem. 57, 2689. Feutren, G. Ž1992. The optimal use of cyclosporin A in autoimmune diseases werratum appeared in J. Autoimmun. Ž1992. 5, 545x. J. Autoimmun. 5 Žsuppl. A., 183. Freed, B.M., Stevens, C., Brooks, C., Cramer, S., Lempert, N. and Rosano, T.G. Ž1991. Assessment of the biological activity of cyclosporine metabolites using the human Jurkat cell line. Transplant. Proc. 23, 980. Holt, D.W. and Johnston, A. Ž1992. Cyclosporin monitoring: its role in autoimmune indications. J. Autoimmun. 5, 177. Holt, D.W. and Johnston, A. Ž1994. Pharmacokinetics and monitoring of cyclosporin A. In: A.W. Thomson and T.E. Starzl ŽEds.., Immunosuppressive Drugs–Developments in Anti-Rejection Therapy. Edward Arnold, London, p. 37. Kino, T., Hatanaka, H., Hashimoto, M., Nishiyama, M., Goto, T., Okuhara, M., Kohsaka, M., Aoki, M. and Imanaka, H. Ž1987. FK-506, a novel immunosuppressant isolated from a streptomyces. I. Fermentation, isolation, physico-chemical characteristics. J. Antibiot. 40, 1249. Kovarik, J.M., Mueller, E.A., van Bree, J.B., Tetzloff, W. and Kutz, K. Ž1994. Reduced inter- and intraindividual variability in cyclosporine pharmacokinetics from a microemulsion formulation. J. Pharm. Sci. 83, 444.

P.R. Wenner et al.r Journal of Immunological Methods 201 (1997) 125–135 Kroencke, M., Leonard, W.J., Depper, M.J., Arya, S.K., WongStaal, F., Gallo, R.C., Waldmann, T.A. and Greene, T.C. Ž1984. Cyclosporin A inhibits T-cell growth factor gene expression at the level of mRNA transcription. Proc. Natl. Acad. Sci. USA 81, 5214. Mattila, P.S., Ullman, K.S., Fiering, S., Emmel, E.A., McCutcheon, M., Crabtree, G.R. and Herzenberg, L.A. Ž1990. The actions of cyclosporin A and FK506 suggest a novel step in the activation of T lymphocytes. EMBO J. 9, 4425. Mueller, E.A., Kovarik, J.M., van Bree, J.B., Tetzloff, W., Grevel, J. and Kutz, K. Ž1994. Improved dose linearity of cyclosporine pharmacokinetics from a microemulsion formulation. Pharm. Res. 11, 301. Napoli, K.L. and Kahan, B.D. Ž1991. Considerations for monitoring cyclosporine in transplant recipients. Clin. Lab. Med. 11, 671. Quesniaux, V.F.J., Schreier, M.H., Wenger, R.M., Hiestand, P.C., Harding, M.W. and van Regenmortel, M.H.V. Ž1987. Cyclophilin binds to the region of cyclosporine involved in its immunosuppressive activity. Eur. J. Immunol. 17, 1359. Quesniaux, V.F., Schmitter, D., Schreier, M.H. and Van Regenmortel, M.H. Ž1990. Monoclonal antibodies to cyclosporin are representative of the major antibody populations present in antisera of immunized mice. Mol. Immunol. 27, 227. Russell, R.L., Donnelly, J.G., Palaszynski, E.W., Chan, M.M. and Soldin, S.J. Ž1991. A preliminary study to evaluate an in vitro assay for determining patient whole blood immunosuppressive cyclosporine A and metabolite activity: comparison with cytosolic binding assays using cyclophilin or a 50-kilodalton binding protein, and the Abbott TDx cyclosporine A parent, and parent and metabolites assays. Ther. Drug Monit. 13, 32. Sambrook, J., Fritsch, E.F. and Maniatis, T. Ž1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Schneider, H., Charara, N., Schmitz, R., Wehrli, S., Mikol, V., Zurini, M.G., Quesniaux, V.F. and Movva, N.R. Ž1994. Hu-

135

man cyclophilin C: primary structure, tissue distribution, and determination of binding specificity for cyclosporins. Biochemistry 33, 8218. Schreiber, S.L. and Crabtree, G.R. Ž1992. The mechanism of action of cyclosporin A and FK506. Immunol. Today 13, 136. Schreier, M.H. and Tees, R. Ž1981. Long-Term Culture and Cloning of Specific Helper T Cells. In: I. Lefkovits and B. Pernis ŽEds.., Immunological Methods, Vol. 2. Academic Press, New York, p. 263. Sehgal, S.N. Ž1995. Rapamune ŽSirolimus, rapamycin.: an overview and mechanism of action. Ther. Drug Monit. 17, 660. Shaw, L.M. Ž1989. Advances in cyclosporine pharmacology, measurement, and therapeutic monitoring. Clin. Chem. 35, 1299. Siekierka, J.J. and Sigal, N.H. Ž1992. FK-506 and cyclosporin A: immunosuppressive mechanism of action and beyond. Curr. Opin. Immunol. 4, 548. Spiers, E.M. and Beck, J.S. Ž1991. Inter- and intra-subject variation of the suppression of mitogen-induced proliferation of human lymphocytes by cyclosporin-A: reduction of response with delayed addition may be relevant to timing of therapy. Int. J. Immunopharmacol. 13, 245. Vine, W. and Bowers, L.D. Ž1987. Cyclosporine: structure, pharmacokinetics and therapeutic drug monitoring wReviewx. Crit. Rev. Clin. Lab. Sci. 25, 275. Wenger, R.M., Martin, K., Timbers, C. and Tromelin, A. Ž1992. Structure of cyclosporine and its metabolites: total synthesis of cyclosporine metabolites formed by oxidation at positions 4 and 9 of cyclosporine. Preparation of leucine-4-cyclosporine, Žg-hydroxy.-N-methyl-leucine-9-cyclosporine and leucine-4Žg-hydroxy.-N-methyl-leucine-9-cyclosporine. Chimia 46, 314. Zenke, G., Baumann, G., Wenger, R., Hiestand, P., Quesniaux, V., Andersen, E. and Schreier, M.H. Ž1993. Molecular mechanisms of immunosuppression by cyclosporins. Ann. N.Y. Acad. Sci. 685, 330.