Pharmacodynamics and Pharmacogenomics ofMethylprednisolone during 7-Day Infusions in Rats
ROHINI RAMAKRISHNAN, DEBRA C. DUBOIS, RICHARD R. ALMON, NANCY A. PYSZCZYNSKI, andWILLIAM J. JUSKO
Department of Pharmaceutical Sciences, School of Pharmacy and Pharmaceutical Sciences, State University of New York at Buffalo, Buffalo,New York
Received June 14, 2001; accepted August 24, 2001 This paper is available online at http://jpet.aspetjournals.org
ABSTRACTAn array of adverse steroid effects was examined on a wholebody, tissue, and molecular level. Groups of male adrenalec-tomized Wistar rats were subcutaneously implanted with Alzetmini-pumps giving zero-order release rates of 0, 0.1, and 0.3mg/kg/h methylprednisolone for 7 days. The rats were sacri-ficed at various times during the 7-day infusion period. A two-compartment model with a zero order input could adequatelydescribe the kinetics of methylprednisolone upon infusion.Blood lymphocyte counts dropped to a minimum by 6 h andwere well characterized by the cell trafficking model. The timecourse of changes in body and organ (liver, spleen, thymus,gastrocnemius muscle, and lungs) weights was described us-ing indirect response models. Markers of gene-mediated ste-roid effects included hepatic cytosolic free receptor density,receptor mRNA, tyrosine aminotransferase (TAT) mRNA, and
TAT levels. Our fifth-generation model of acute corticosteroidpharmacodynamics was used to predict the time course ofreceptor/gene-mediated effects. An excellent agreement be-tween the expected and observed receptor dynamics sug-gested that receptor events and mRNA autoregulation are notaltered upon 7-day methylprednisolone dosing. However, themodel indicated a decoupling between the receptor and TATdynamics with this infusion. The strong tolerance seen in TATmRNA induction could be partly accounted for by receptordown-regulation. An amplification of translation of TAT mRNAto TAT and/or a reduction in the enzyme degradation rate couldaccount for the observed exaggerated TAT activity. Our resultsexemplify the importance of biological signal transduction vari-ables in controlling receptor/gene-mediated steroid responsesduring chronic dosing.
The clinical use of steroids is extensive and frequentlychronic. Corticosteroids currently are among the most impor-tant drugs used for the treatment of a variety of immunolog-ical conditions such as lupus erythematosus, rheumatoidarthritis (Canvin and el-Gabalawy, 1999), organ transplan-tation, bronchial asthma (Boushey, 1998), and inflammatorybowel disease (Selby, 1993), to name a few. Short-term treat-ment in acute or transient illness is generally not associatedwith major side effects. However, the multiple and potentmetabolic effects of steroids become prominent upon chronicdosing, leading to an increased risk of toxicity thus limitingtheir usefulness. The undesirable metabolic effects of corti-costeroids cannot be completely separated from their favor-able anti-inflammatory effects because most actions aremanifested using the same glucocorticoid receptor. The inte-grated effects result in hyperglycemia, negative nitrogen bal-ance, and fat redistribution leading to complications, includ-ing diabetes, muscle wasting, hypertension, cataracts, and
peptic ulcers (David et al., 1970; Baxter and Forsham, 1972;Swartz and Dluhy, 1978). Another important clinical conse-quence of long-term use of steroid is adrenal insufficiencyafter withdrawal of therapy (Swartz and Dluhy, 1978).
Because the principal undesirable effects of steroids aremanifested only upon long-term dosing, it can be expectedthat there might be additional factors contributing to thedynamics of steroid responses under these dosing conditions.Results (Ramakrishnan, 2001) from steady-state studies con-ducted previously in our laboratory suggested that a decou-pling between receptor and enzyme dynamics occurs uponlong-term continuous dosing of steroid in male adrenalecto-mized rats. It is possible that nuclear molecular events (re-ceptor translocation, chromatin binding) are altered uponchronic dosing. On the other hand, signal transduction pro-cesses involved in generation of the response can be ampli-fied/diminished due to global systemic effects of the steroid.Corticosteroid responses may be countered by other hor-mones whose circulating levels are affected by prolongedsteroid exposure (Baxter and Forsham, 1972). For instance,the release of insulin is stimulated in response to hypergly-
This study was supported by Grant GM24211 from the National Institutesof Health.
cemia, thus partly reversing steroid effects on gluconeogen-esis. An understanding of the in vivo receptor regulationmachinery and components of the transduction processesupon long-term steroid treatment is vital for successful ste-roid hormone therapy. Hence, it is of interest to investigateand model the response of the system upon administration oflengthy steroid dosage regimens. Our current model of gene-mediated corticosteroid effects provides an ideal tool to detectthe components of the receptor/signal tranduction systemresponsible for any decoupling of acute versus chronic steroideffects.
Our 1-week methylprednisolone infusion studies reportedpreviously were designed to characterize steady-state re-sponses with respect to body weight loss, changes in organweights, and receptor/tyrosine aminotransferase (TAT) lev-els. Although we noted that the steady-state receptor dynam-ics occurred as expected, the ultimate TAT response seemedto be exaggerated. This was a single 7-day time point studyand we did not obtain any information on the time course ofchanges in the various dynamic measures. This made it dif-ficult for us to make extrapolations regarding any differencesbetween acute versus long-term steroid effects. In this report,we have therefore extended our studies to uncover the entireprofile of temporal changes in the various steroid effects.Rats were administered two infusion regimens for 7 days,which allowed steady-state conditions to be achieved. Byusing s.c. infusion regimens instead of multiple i.v. doses, wecould collect a rich data set with relatively few animals. Also,the use of two different infusion rates enabled us to assessthe role of dose size, rate of drug input, and duration ofexposure on the dynamics. Various measures of toxicity andimmunosuppression were measured as well during differenttimes over the infusion period.
The undesirable metabolic effects of steroids were quanti-tated at three levels as follows: 1) whole body (body weightloss); 2) tissue (changes in liver, spleen, thymus, lungs, mus-cle, heart, and kidney weights); and 3) molecular (down-regulation of receptor mRNA, free receptor density, and en-hancement of TAT mRNA and TAT enzyme activity).Changes in blood lymphocyte counts were used as a markerof the rapid immunosuppressive effects of the steroid.
Simulations were performed using the fifth-generationmodel for corticosteroid receptor/gene-mediated effects to ob-tain the expected time course of receptor and TAT dynamics.Comparisons were made between the expected and observeddata patterns. The TAT dynamics was fitted to the model toobtain parameters specific for the long-term infusion effects.The most prominent differences in the parameters were eval-uated and used to make judgments as to which model com-ponents could have been possibly altered upon long-termdosing.
Materials and MethodsAnimals
Adrenalectomized male Wistar rats with body weights of 339 � 28(S.D.) g were used in the study. All animals were housed in ourUniversity Laboratory Animal Facility maintained under constanttemperature (22°C) and humidity with a controlled 12-h light/darkcycle. A time period of at least 2 weeks was allowed before they wereprepared for surgery. Rats had free access to rat chow and 0.9% NaCldrinking water. This research adheres to Principles of Laboratory
Animal Care (National Institutes of Health publication 85-23, re-vised 1985) and was approved by the Institutional Animal Care andUse Committee of the State University of New York at Buffalo.
Experimental
Rats were divided into four groups. Two treatment groups con-taining 36 rats each were administered 0.1 and 0.3 mg/kg/h infusionsof methylprednisolone sodium succinate (Solu-Medrol; The UpjohnCompany, Kalamazoo, MI) reconstituted in supplied diluent. Theinfusions were given using Alzet osmotic pumps (model 2001, flowrate 1 �l/h; Alza, Palo Alto, CA). The pump drug solutions wereprepared for each rat based on its predose body weight. On the dayof implantation, rats were anesthetized using 60 to 80 mg/kg ket-amine and 8 to 10 mg/kg xylazine i.m. Pumps were subcutaneouslyimplanted between the shoulder blades on the back. Rats weresacrificed at various times up to 7 days, the time points includedbeing 6, 10, 13, 18, 24, 36, 48, 72, and 96 h. The third group (eightrats) was administered an i.v. bolus dose of 50 mg/kg. There werefour rats sacrificed after 5 h and the remaining four at the end of 6 h.The control group of eight animals was implanted with a saline-filledpump and sacrificed at various times throughout the 7-day studyperiod. Before pump implantation, the body weight of each rat wasmeasured and a blood sample was withdrawn from the tail vein toobtain the predose blood lymphocyte counts. The body weight of eachrat was recorded upon sacrifice and the sacrifice blood was used todetermine the plasma methylprednisolone concentrations and thelymphocyte counts. Various organs, including the heart, kidney,gastrocnemius muscle, lungs, spleen, thymus, and liver were excisedand weighed. One gram of liver tissue was immediately processed forTAT enzymatic activity measurements and the remaining liver tis-sue was flash frozen for cytosolic receptor mRNA, free receptordensity, TAT mRNA, and cAMP measurements. Thus, from each rat,we obtained one pharmacokinetic measurement (sacrifice methyl-prednisolone plasma concentrations) and three sets of pharmacody-namic measurements, including the body weight/organ weightchanges, blood lymphocytes, and gene-mediated effects (hepatic re-ceptor regulation and TAT dynamics).
Assays
Normal phase high-performance liquid chromatography with alimit of quantitation of 10 ng/ml was used to measure plasma meth-ylprednisolone concentrations (Ebling et al., 1985; Sun et al., 1998b).A previously established radiolabeled ligand binding assay (Boudi-not et al., 1986; Sun et al., 1998a) was used to quantitate the freereceptor density in rat liver cytosol. The cytosolic receptor density(Bmax) was estimated by solving the following equations simulta-neously:
DT � DNS �Bmax � Df
KD � Dfand DNS � K � Df (1)
where KD and K are the equilibrium dissociation constants for spe-cific and nonspecific binding. The mRNA for the receptor and theTAT mRNA were assayed using quantitative Northern hybridization(DuBois et al., 1993, 1995). Spectrophotometric determination ofhepatic TAT activity was performed using the Diamondstone color-imetric assay (Diamondstone, 1966). The protein content of the sam-ples, as determined by the Lowry assay (Lowry et al., 1951), wasused to normalize the free receptor density and TAT activity. Bloodlymphocyte counts were measured using the automated CELL-DYN1700 system (Abbott Diagnostics, Abbott Park, IL). For quantitativedetermination of cAMP, a commercially available competitive en-zyme immunoassay kit was used (Correlate-EIA Direct cAMP kit;Assay Designs, Ann Arbor, MI). In this assay, the cAMP in thesample competes with an alkaline phosphatase molecule covalentlyattached to cAMP, to bind to a cAMP polyclonal antibody. Theenzyme reaction generates a yellow color, which is read at 405 nm ina microplate reader. The optical intensity is inversely proportional to
the concentration of cAMP in the samples or standards. For samplepreparation, 0.5 g of frozen liver tissue was weighed and homoge-nized in 5 ml of ice-cold 0.1 N HCl. The homogenates were subject toa low-speed centrifugation at 16,000 rpm for 30 min followed by asecond centrifugation at 40,000 rpm for 1 h at 4°C. The supernatantwas frozen at �80°C until assayed. The assay was performed induplicate as per the kit instructions for the nonacetylated version ofthe assay, which has a sensitivity of 0.39 pmol/ml.
Pharmacokinetic/Pharmacodynamic Model
Pharmacokinetics. The pharmacokinetics of methylpred-nisolone for the rats administered the infusion regimens were de-scribed by a two-compartment model with a zero-order input k0 intothe central plasma compartment as follows:
dAp
dt� k0 � k21 � At � k12 � Ap � �CL
Vp� � Ap (2)
dAt
dt� k12 � Ap � k21 � At where Ap � Cp � Vp (3)
where k12, k21 and Vp are the distribution rate constants and centralvolume of distribution. These parameters were fixed based on pre-vious literature (Sun et al., 1998b) values, whereas the clearance(CL) was estimated from our data. The Ap and At are the amounts ofdrug in the plasma and tissue compartments. For the i.v. boluskinetics, the same equations were used with the exception of thezero-order input function. Parameters obtained from our previousstudy (Sun et al., 1998b) were used for the simulations. All kineticparameters were fixed (Table 1) and used as a driving force for thedynamics.
Pharmacodynamics. Fittings for all doses were performed si-multaneously and the data from all individual rats were used for theanalysis.
Body Weight. The catabolic effects of the steroid on the body weremodeled using indirect response model IV (Dayneka et al., 1993)with the stimulation function applied to the degradation rate asfollows:
dRdt
� ksyn � kdeg � R � �1 �Smax � Cp
SC50 � Cp� (4)
The sacrifice body weight of each rat was expressed as a percentageof the predose body weight. Adrenalectomized animals do not haveany steroid in circulation, hence the baseline for eq. 4 is ksyn � kdeg �
R0, where ksyn and kdgr are the zero- and first order production anddegradation rate constants, and R0 is the baseline response (100%).The Smax and SC50 are the steroid-specific parameters representingthe maximal possible increase in degradation rate and the steroidconcentrations required for half-maximal stimulation.
Organ Weight. Organ weight ratios based on the control organweights and predose body weights of the treated rats were calcu-lated. The organ weights of the control animals were normalized by
the corresponding body weights. The mean of these was then calcu-lated to obtain the ideal organ weight ratio. A predose organ weightwas calculated for each treated animal based on its predose bodyweight and the ideal organ weight ratio. The organ weight ratio wassimply a ratio of the measured organ weight at the time of sacrificeand the estimated predose organ weight. The changes in organweight ratio were modeled using indirect response models (Daynekaet al., 1993). The hypertrophy of the liver was modeled using astimulation function S(t) on the production rate, whereas the stim-ulation function was applied to the degradation rate to capture thenet catabolic effects on the lymphoid tissues, lungs, and muscle. Thedifferential equations used included the following:
dRdt
� ksyn � �1 � S�t�� � kdeg � R (5)
for the liver and
dRdt
� ksyn � kdeg � R � �1 � S�t�� (6)
for the other tissues where S(t ) � (Smax � Cp)/SC50 � Cp).Gene-Mediated Effects. An intracellular model as depicted in
Fig. 1 was used to describe the receptor-gene-mediated corticosteroideffects of methylprednisolone in terms of changes in receptor dynam-ics and TAT induction in the rat liver cytosol. This is the mostcurrent model (Ramakrishnan, 2001), which was developed by ourlaboratory based on acute dosing of MPL in rats. The major compo-nents of the model included the following.
Down-Regulation of Receptor mRNA as Controlled by Ac-tivated Steroid Receptor Complex. An inhibition of receptor genetranscription ksyn_Rm by the activated drug-receptor complex in thenucleus DR(N) was assumed to be the major mechanism of receptormRNA (Rm) down-regulation as follows:
dRm
dt� ksyn_Rm � �1 �
DR(N)IC50_Rm � DR(N)�� kdgr_Rm � Rm (7)
where the degradation rate kdgr_Rm/Rm0, Rm0 being the baselinereceptor mRNA levels obtained from the control animals.
Receptor Dynamics. Free receptor density in liver cytosol wasmodeled by taking into account the kinetics of drug-receptor binding(kon), receptor recycling (kre/Rf), translation of receptor mRNA(ksyn_R), and turnover rate of the receptor (kdgr_R) as follows:
dRdt
� ksyn_R � Rm � Rf � kre � DR�N� � kon � D � R � kdgr_R � R (8)
where D is the total molar concentration of the steroid in circulation.The baseline was defined as ksyn_R � (R0/Rm0) � kdgr_R where R0 is thebaseline receptor levels.
The differential equations for the drug-receptor complex in thecytoplasm DR and that bound to the GRE in the nucleus DR(N) weregiven as follows:
TABLE 1Pharmacokinetic and toxicity parameters upon long-term steroid infusion in male ADX rats
Induction of TAT mRNA. The enhancement in the zero ordersynthesis (ksyn_tm) of TAT mRNA was dependent on the amount ofactivated steroid-receptor complex bound to the GRE in the nucleus,as controlled by the linear efficiency constant of gene transcription(S) and the basal degradation rate (kdgr_tm) as follows:
The basal transcription rate of TAT mRNA was defined as ksyn_tm �kdgr_tm � TATm0, where TATm0 is the baseline message levels obtainedfrom the control rats.
Induction of TAT. The sequential processes in the model culmi-nate with the enhanced expression of TAT as governed by the trans-lation rate of mRNA (EF) and its first order degradation (kdgr, TAT).The differential equation describing TAT dynamics is as follows:
dTATdt
� EF � �TATm)� � kdgr_t � TAT (12)
where � is an amplification factor indicating that on average, a singlemRNA transcript can be used to translate multiple copies of theenzyme. The baseline control TAT levels (TAT0) were used to definethe efficiency of translation (EF) as follows:
EF � � TAT0
(TATm0)�� � kdgr_t (13)
The parameters reported earlier and as listed in Table 2 were usedto simulate the receptor/gene-mediated effects for the IV and infu-sion regimens. The increase in liver weight ratio (R) with time wasfitted to the following equation:
dRdt
� ksyn � �1 � Smax � DR�N�� � kdeg � R (14)
The drug-receptor complex in the nucleus DR(N) was assumed tocause a linear stimulation (Smax) of the zero order production (ksyn) ofthe response. The first order degradation rate kdgr controlled thebaseline response R0 as follows: ksyn � kdeg � R0.
The above-mentioned equation along with the estimated parame-ters was used to predict the time course of organ weight ratiosthroughout the 7-day treatment period for the two infusion groups.The predicted message levels for receptor and TAT were divided bythe corresponding predicted liver weight ratios to account for thedilution in the mRNA levels that were expressed on a per gram basis.The parameters for the receptor dynamics were fixed, whereas theTAT mRNA and TAT data from our infusion study were fitted to themodel to obtain parameters specific for long-term dosing.
Lymphocyte Trafficking. A cell trafficking model was used todescribe the change in blood lymphocyte counts. According to thismodel, lymphocytes from the tissues enter into blood at a constantzero order rate kin and return from the blood to these extra vascularsites is controlled by a first order rate kout. It is assumed that steroidsinstantaneously cause a change in the affinity of these extracellulartissues, thus inhibiting the egress of lymphocytes from the blood totissues. Hence, an inhibition function was applied on the zero orderentry rate kin as follows:
dRdt
� kin � �1 �Cp
IC50 � Cp�� kout � R (15)
where R represents the lymphocytes in blood and IC50 is the drugconcentration that inhibits kin by 50%. The lymphocyte counts wereexpressed as a percentage of the predose value (defined as 100%) foreach animal and this was defined as the response R. The meanpercentage of predose lymphocyte counts were used for the fittings.The control animals had an average 20% drop in blood lymphocytecounts. Hence, the baseline was modeled as kin � kout � 0.8 � R0, whereR0 is 100%. The kout was fixed to 0.643 h�1 based on literatureestimates (Ferron et al., 1999).
All data analyses were performed using the ADAPT II software(D’Argenio and Schumitzky, 1997) by using the maximum likelihoodmethod. The extended least-squares variance model was specified asV(�,�,ti) � �1
2 � Y(�,ti)�2 where V(�,�,ti) is the variance for the ith
Fig. 1. Fifth-generation model for long-term corticosteroid receptor/gene-mediated effects. The dotted lines leading to the open and closed rectan-gles indicate stimulation and inhibition of the first order synthesis rate ofthe response variable. The model is described in the text from eqs. 7 to 14.
TABLE 2Parameters for acute and long-term receptor/gene-mediated steroid effects
Receptor Dynamics (Fixed) Value TAT DynamicsAcute Long-Term
point, � represents the structural parameters and �1 and �2 are thevariance parameters that were fitted. Different variance parameterswere estimated for each data set that was obtained by a differentassay methodology.
ResultsThe results reported here are the latest from a series of
“giant rat” studies conducted in our laboratory that involvessacrificing animals to obtain serial blood and tissue samples.The data generated is such that each point represents themeasurement from one separate rat and in effect, the mea-surements from all these different rats are pooled to obtain atime course as though it was obtained from one giant rat. Anaive pooled data analysis approach was therefore used forall model fittings.
Pharmacokinetics
The plasma methylprednisolone concentrations-time pro-files after the two infusion regimens are shown in Fig. 2.Methylprednisolone is known to undergo nonlinear intercon-version and oxidative elimination processes in rats (Kongand Jusko, 1991). Plasma protein binding is constant (77%)with concentration (Haughey and Jusko, 1992). Steady-stateconcentrations in the infusion study were 100-fold lower thanthat in the previous bolus studies (Sun et al., 1998a,b).Hence, CL was estimated from the infusion data. The ascend-ing part of the curve, which gives us information regardingthe absorption of the drug from the subcutaneous site, wasunavailable and hence the tissue distribution constants andthe volume of distribution were fixed based on the previousbolus estimates. Furthermore, it was assumed that the bio-availability is complete and the absorption rate is muchfaster than the rate of drug release from the pump. Theclearance increased from 4 to 5.6 l/h/kg, which suggests thathigher concentrations are associated with saturation of drug-metabolizing enzymes. Table 1 lists the pharmacokinetic pa-rameters describing the data for the two infusion regimens.
Pharmacodynamics
Body Weights. The control animals showed constant bodyweights. The two infusion groups showed dose-dependentlosses in body weights, which continued throughout the
treatment period (Fig. 3). The low- and high-dose groups fellto 89 and 82% of control by 96 h. The model could satisfac-torily capture the loss in body weight for both groups. Aslisted in Table 1, the kdeg, Smax and SC50 values are compa-rable to those estimated in our previous study (Ramakrish-nan, 2001). These parameters were used to perform simula-tions for the expected loss in body weight upon acute dosing.The high-dose group given an infusion rate of 0.3 mg/kg/hreceived a total dose of 50.4 mg/kg in 7 days. Figure 4 showsa simulation for the changes in body weight for a 50.4-mg/kgi.v. bolus dose as predicted by our model. The simulationsshow that when such a high dose of steroid is administered asa single i.v. bolus, hardly any loss in body weight is expected,whereas the same total dose given in the form of 7-dayinfusion could be expected to cause substantial losses in bodyweight. The dose- and duration-dependent steroid effects onbody weight can be explained by considering the pharmaco-kinetics of methylprednisolone. The steroid is almost com-pletely cleared from the circulation in 6 h after an i.v. bolusdose, whereas methylprednisolone concentrations are main-tained above the SC50 for 7 days after the infusion regimen,leading to a continued body weight loss in these rats.
Organ Weights. Fig. 5 shows the time course of change inorgan weights for the spleen, thymus, lungs, muscle, liver,kidney, and heart as well as the fittings of our model for thevarious tissues. The 7-day organ weights from our previousstudy (Ramakrishnan, 2001) were included. The organweight ratios changed in a dose-related manner. The ana-bolic effects of the steroid on the liver tissue were observed ashypertrophy of the liver, whereas the lymphoid organs, mus-cle, and lungs were subject to the catabolic actions of thesteroid. The heart and kidney weight ratios remained fairlyconstant with time in both treated groups. Table 1 lists theparameters estimated for the different organs. The IC50 var-ied from tissue to tissue, suggesting that the sensitivity of thevarious organs to steroid treatment is different. The muscleweights dropped linearly over the period of 7 days, which
Fig. 2. Pharmacokinetics of methylprednisolone upon administration of0.1 (F) and 0.3 (E) mg/kg/h infusions for 7 days. Solid and broken lines areresults of a simultaneous fitting by using eqs. 2 and 3. The pharmacoki-netic parameters are listed in Table 1.
Fig. 3. Indirect response model for effects of methylprednisolone on bodyweight (top) and the time course of changes in body weights for the0.1-mg/kg/h (F) and 0.3 mg/kg/h (E)-infusion groups (bottom). The solidand broken lines are the simultaneous fittings for the two dose levels byusing eq. 4.
Pharmacodynamics and Pharmacogenomics of Methylprednisolone 249
makes it difficult to obtain a reliable estimate of Smax fromthe data, as reflected by the abnormally high value esti-mated. Figure 6 shows the area between the baseline andeffect curve from 0 to 168 h (ABEC0–168 h), which is anindicator of the net cumulative response of the tissue tosteroid treatment. The ABEC values for the thymus andspleen were the highest, whereas the lungs and muscle fol-lowed. The hypertrophy of the liver was comparable to theinvolution effect on the lungs. In general, the cumulativeeffect on all the tissues was concentration-dependent.
Gene-Mediated Enzyme Induction. Typical phospho-rimages of Northerns for the GCR mRNA and TAT mRNAare provided in Fig. 7. The receptor mRNA, free receptordensity, TAT mRNA, and TAT activity data at 5 and 6 h aftera 50-mg/kg i.v. bolus dose are shown in Fig. 8. Simulationsusing the fifth-generation model and its parameters couldsatisfactorily describe the data patterns for this group of rats.Hence, these parameters were used to perform simulationsfor receptor and TAT dynamics during the two infusion reg-imens of 0.1 and 0.3 mg/kg/h (Fig. 9). The half-life of meth-ylprednisolone is very short (�30 min), which results insteady-state concentrations being achieved within few hours.However, the simulations show that it takes up to 24 h for thesteady state in dynamic measures to be attained. The recep-tor mRNA profiles fall early and reach steady-state levelsthat are 50 to 55% of control values. Free receptor levels dosedependently fall to a new steady state within 1 day. An earlyrise followed by a dramatic decrease stabilizing at valuesclose to that of controls reflect the remarkable tolerance inthe TAT mRNA and TAT profiles. Differences among the two
doses are prominent early on but they vanish at steady statedue to occurrence of down-regulation/tolerance. The simula-tions indicate that responses governed by receptor/gene-me-diated events may show tolerance upon long-term continuousdosing of steroid due to receptor down-regulation. Figure 10shows the receptor and TAT results from our infusion study.The 7-day receptor and TAT levels from our previous steady-state study (Ramakrishnan, 2001) were included in the fig-ures. There was a drop in the receptor mRNA levels indicat-ing down-regulation in the message expression. The cytosolicfree receptor density fell as a result of receptor binding andnuclear translocation as well as down-regulation to reachdose-related steady-state levels within 24 h. The TAT mes-sage levels rose to a maximum at 6 h after which the levelsfell to values close to that of the control. The TAT mRNA forthe lower dosing group was variable and stayed close to
Fig. 4. Simulations for the expected change in body weight for a 7-day0.3-mg/kg/h infusion (broken line) and 50.4-mg/kg i.v. bolus (solid line)dose (top). Simulated pharmacokinetic profiles after administration of a7-day infusion at a rate of 0.3 mg/kg/h (broken line) and a 50-mg/kg i.v.bolus dose (solid line) (bottom). Dotted line is the estimated SC50 (19.56ng/ml) for methylprednisolone effect on body weight.
Fig. 5. Effect of a 0.1- (F) and 0.3 (E)-mg/kg/h 7-day infusion on weightratios of liver, lymphoid organs (spleen and thymus), lungs, gastronemiusmuscle, heart, and kidney. The solid and broken lines for the liver dataare the simultaneous fittings for the low- and high-dose levels by usingeq. 5 derived from indirect response model III (top right). The solid andbroken lines for the spleen, thymus, lungs, and muscle are the simulta-neous fittings for the low- and high-dose levels by using eq. 6 derived fromindirect response model IV (top right). The solid and broken lines for thekidney and heart data connect the mean at the different time points forthe low and high doses.
control values throughout the infusion duration. MaximumTAT activity was observed by 10 h followed by a steep dropclose to control levels at 24 h for both the treatment groups.Thus, severe tolerance was observed at steady state that wasmaintained for the entire 7-day infusion duration. The riseand fall in TAT activity as well as its message levels weredose-dependent during the early time frame but the differ-ences almost disappeared when steady state was attained at24 h. The temporal data patterns observed seemed to be wellpredicted by our model. The higher dose was associated witha greater decrease in free receptor levels corresponding to agreater extent of TAT induction. Table 2 includes the param-eters obtained from the simultaneous fitting of the indirect
response model (eq. 15) to the liver weight ratios for the twoinfusion groups. The fitted curves were comparable to thoseobtained using the indirect response modeling approach (Fig.5). Simulations for the receptor/gene-mediated effects ac-counting for this change in liver weights were superimposedon the observed data to quantitatively compare the modelpredicted and experimentally observed results. The excellentagreement between the time course of observed and pre-dicted receptor dynamics indicates that the model parame-ters based on acute dosing could well account for the rate andextent of changes in receptor mRNA and free receptor levelsduring infusion. The predicted TAT mRNA curves consis-tently overpredicted the observed change in the levels up to24 h. On the other hand, the early rise in TAT activity wasseverely underpredicted by the model. Although our modelpredicted tolerance in TAT activity, the extent of tolerance inthe observed data was far greater than that expected basedon extrapolations from single dosing. The results point to thepossibility that there is a dissociation between receptor andTAT dynamics during long-term corticosteroid dosing.
Table 2 lists the parameters for the TAT dynamics specificto our long-term infusion study and Fig. 11 shows the fittingof the model to the data. A comparison of the parametersindicated that there was more than a 50% change in the S, �,and kdeg t parameters. This suggests that the efficiency ofTAT gene induction had been reduced during long-term dos-ing. Also, an amplification of the efficiency of translation ofTAT mRNA to TAT as well as an increase in the half-life ofthe TAT protein might have contributed to the differentialeffect on induction of TAT activity.
There is considerable evidence in the literature thatchanges in cAMP can alter the extent of TAT gene expression(Wicks, 1968; Hashimoto et al., 1984). We measured cAMP inrat liver to determine whether there was any fall in cAMPlevels that could have contributed to a reduction in the effi-ciency of TAT gene expression. As seen in Fig. 12, there wasno change in cAMP levels in the rats administered the i.v.bolus, but levels were in fact 3- to 6-fold higher than controlvalues between 18 and 72 h for the infusion groups. However,at early times up to 18 h, the levels remained close to control,which suggests that cAMP action cannot explain the discrep-ancy in the receptor and TAT dynamics during long-termdosing.
Lymphocyte Trafficking. As shown in Fig. 13, within6 h, the lymphocyte counts had plummeted to a minimum inall treatment groups. The percentage of predose lymphocytecounts dropped in a dose-dependent manner to mean steady-state values of 19 and 13%. Due to unavailability of earlytime points governing the down-curve, we did not have thepower to estimate kout, which is a physiological drug-inde-pendent parameter. Therefore, its value was fixed based on aliterature reported value of 0.643 h�1 from a study done inadrenalectomized rats that were administered a single i.v.bolus of prednisolone (Ferron et al., 1999). The IC50 esti-mated was 6.15 ng/ml. Surprisingly, the control group ani-mals showed a mean 20% drop in blood lymphocyte counts,including the first measured time point itself. These animalsbeing adrenalectomized, do not have any steroid in circula-tion, and hence should be expected to have a constant base-line blood lymphocyte count. However, it is possible thatpump implantation might have caused some tissue trauma
Fig. 6. Plot of ABEC from 0 to 168 h for the different organs. Low dose (f)and high dose (p).
Fig. 7. Top, representative phosphorimage of quantitative Northernanalysis for GCR mRNA. The seven lanes on the left represent GCRcRNA standards. The 18 lanes on the right represent six liver total RNAsamples run in triplicate. Bottom, representative phosphorimage of quan-titative Northern analysis for TAT mRNA. The top signals representhybridization to the TAT probe, and the bottom signals represent hybrid-ization to GRG external standard cRNA, which is added to the tissue foryield correction. The seven lanes on the left represent cRNA standards;the 18 lanes on the right represent six liver total RNA samples run intriplicate.
Pharmacodynamics and Pharmacogenomics of Methylprednisolone 251
associated with a local inflammatory reaction, causing achange in the baseline lymphocyte counts.
DiscussionThe 7-day infusion of MPL caused pronounced losses in
body weights of rats, which does not occur in acute dosingstudies (Sun et al., 1998b). This dose and duration-dependentdifferential effect was well captured by our model. The IC50
and Smax for the catabolic effects on the various organs var-ied substantially, indicating that these tissues show differentsensitivities and capacity to respond to steroid treatment.Assuming that all the tissues studied were well perfused, thedissociation constant for the drug-receptor binding and thefree receptor density, as controlled by recycling fraction andreceptor mRNA autoregulation would limit the overall re-sponse of a tissue to steroid treatment. Tissue-specific regu-lation of glucocorticoid receptor mRNA levels has been re-ported in normal, adrenalectomized as well as steroid-treatedrats (Kalinyak et al., 1987; Miller et al., 1998). The Kd fordexamethasone-receptor binding has been reported to be sig-nificantly higher in the spleen and thymus of adrenalecto-mized rats compared with the lung or liver (Ichii, 1981). Weobtained higher IC50 values for the spleen and thymus sug-gesting that drug-receptor binding may reflect the sensitivity
of these tissues to steroids. In adrenalectomized as well asnormal rats, the thymus has been reported to have the high-est concentration of cytosolic glucocorticoid receptor followedby the liver, spleen, and lung (Ichii, 1981; Miller et al., 1998).Except for the spleen whose Smax was higher than thatfor the liver, our results are consistent with this order,indicating that the Smax estimated is a measure of the over-all receptor density in these tissues. Furthermore, theABEC
0–168 hvalues further confirm the observation that the
net steroid effect on any organ would be proportional to thefree receptor density.
Because our fifth-generation model covers the completetime course of receptor and enzyme induction events, it pro-vided us with the opportunity to perform simulations forprediction of expected results during long-term treatment.The parameters from the single-dosing studies could satis-factorily predict the temporal patterns of receptor mRNA andreceptor dynamics, suggesting that receptor regulation is notaltered upon long-term dosing. However, we noted a markeddisagreement between the model-predicted and observedTAT dynamics in our study. The remarkable pathophysiolog-ical effects of 7-day steroid treatment on these animals leadsus to postulate that other hormones might come into playunder these conditions, thus opposing or enhancing the ste-
Fig. 8. Time course of receptor mRNA (top left), free cytosolic receptor density (bottom left), TAT mRNA (top right), and TAT activity (bottom right)after a single 50-mg/kg i.v. bolus dose of methylprednisolone in male ADX Wistar rats. Solid circles are experimental data from individual rats at 5and 6 h after injection and solid lines are simulations with the fifth-generation model (eqs. 7–13). The parameters used include those for acute steroideffects as listed in Table 2.
roid effects on enzyme induction. Hyperglycemia could beexpected to cause elevated plasma insulin levels. The globaleffects of the steroid could lead to alterations in liver cAMPlevels via changes in glucagon secretion.
It is known that hormones, including insulin, glucagon,and glucocorticoids control TAT activity (Holten and Kenney,1967; Ernest and Feigelson, 1979). Genes that encode glu-coneogenic enzymes such as phosphoenolpyruvate carboxyki-nase and TAT are transcriptionally up-regulated by glucocor-ticoids as well as by glucagon via cyclic AMP (Yeung andOliver, 1968; Scherer et al., 1982; Hashimoto et al., 1984).Unlike glucocorticoids that bind to a cytosolic receptor anddirectly enhance gene transcription, glucagon binds to cellsurface receptors and activates a second messenger system(cAMP), leading to phosphorylation and activation of a vari-ety of transcription factors up-regulating TAT activity(Schmid et al., 1987). The cAMP levels in the hepatic cytosolremained close to control and hence cannot explain the lackof expected induction in TAT mRNA seen between 10 and24 h. Moreover, our cAMP data would lead us to expecthigher than predicted TAT mRNA levels between 24 and48 h. On the contrary, we found that message levels werelower implying that the expected cAMP-mediated TAT geneinduction at these times is opposed by other factors. Studiesin a rat hepatoma cell line have implicated interactions withspecific target sequences in the cAMP response element andhepatic nuclear factor-4 binding sites adjacent to the GRE in
mediating the glucocorticoid/insulin and cAMP/insulin an-tagonism of TAT gene expression (Ganss et al., 1994). It hasalso been shown that insulin alone induces TAT mRNA andTAT activity (Reel et al., 1970; Spencer et al., 1978), but inthe presence of glucocorticoids, insulin acts at a post-tran-scriptional level only increasing TAT activity (Crettaz et al.,1988). We propose that gluconeogenesis stimulated by meth-ylprednisolone in the treated rats might have led to elevatedplasma glucose levels, stimulating the release of insulin by10 h. This could have repressed the TAT gene expression andfurther antagonized the cAMP-mediated gene expression aswell. Studies using hepatoma cells suggest that whereassteroids enhance gene transcription, insulin acts on a post-transcriptional or translational step by increasing the rate atwhich existing TAT mRNA is translated to TAT (Kenney etal., 1970) as well as impairing the degradation rate of theenzyme (Crettaz et al., 1988). Our modeling results suggestthat both a change in the message translation rate as well asa decrease in the enzyme degradation rate may have contrib-uted to the amplification of TAT activity upon long-termcontinuous dosing, which further supports the role of insulin.
We incorporated the effects of tissue organ weights on themeasured molecular markers into our model. Because theliver weight changes in response to long-term steroid treat-ment, traditional normalization of the receptor/TAT mRNAlevels on a per gram liver tissue basis would lead to a netdilution effect, resulting in an underestimation of the true
Fig. 9. Simulations using the fifth-generation model (eqs. 7–14) for the time course of receptor mRNA (top left), free cytosolic receptor density (bottomleft), TAT mRNA (top right), and TAT activity (bottom right) during 0.1- (solid line) and 0.3 (dashed line)-mg/kg/h 7-day infusions of methylpred-nisolone in male ADX Wistar rats. The parameters used include those for acute steroid effects as listed in Table 1.
Pharmacodynamics and Pharmacogenomics of Methylprednisolone 253
increases in these measures. Simultaneous modeling of theliver weight and delayed gene-mediated effect data allowedus to relate the anabolic effects on the liver tissue to those ata molecular level. The steroid-receptor complex DR(N) in thenucleus can mediate the induction of a variety of metabolicgenes (TAT being one of them), leading to altered expressionof enzymes. This would contribute to the liver hypertrophyvia a net increased synthesis of glucose, RNA, and proteins aswell as fat deposition (Baxter and Forsham, 1972). It was notattempted to directly correlate the rise in TAT to increase inliver weight because TAT is only one of the numerous genesinduced by steroids and might not be the rate-limiting factorcausing liver hypertrophy.
The rapid immunosuppressive effects of various steroidson blood lymphocyte counts upon single IV bolus dosing inanimals (Ferron and Jusko, 1998; Ferron et al., 1999) as wellas humans (Fisher et al., 1992; Chow et al., 1999) has beendescribed earlier. The loss in blood lymphocyte counts uponlong-term dosing could be adequately described using the celltrafficking model in our study.
Several studies by us and other investigators have demon-strated that the rate and extent of receptor depletion corre-lates with the biological response upon acute dosing in adre-nalectomized rats. However, an understanding of the in vivoreceptor regulation and biological response upon long-term
steroid dosing is limited. Repeated stress has been shown tobe associated with decrease in cytosolic receptor numbers inliver (Alexandrova and Farkas, 1992) and brain (Sapolsky etal., 1984) of intact rats. However, stress is a complex andnonspecific stimulus and receptor down-regulation cannot bedefinitively attributed to being glucocorticoid-induced. Yo-shida et al. (1986) reported that receptor function is notdistorted upon chronic steroid treatment in adrenalecto-mized rats. The authors also proposed that the correlationbetween receptor loss and TAT induction disappears uponlong-term dosing. Our studies and quantitative analysis ex-tend their results in that we found that long-term continuoussteroid treatment does not cause alterations in receptormRNA autoregulation as well as receptor dynamics and nu-clear molecular events (receptor binding, translocation, chro-matin binding).
Steroid kinetics coupled with receptor mRNA autoregula-tion seem to be the primary factors governing the time courseof free receptor levels during both acute and long-term dos-ing. Furthermore, our results suggest that postreceptorevents might have contributed to a decoupling between re-ceptor dynamics and gene induction during long-term dosing.Although the early TAT response was exaggerated, continu-ous steroid treatment was associated with the developmentof tolerance due to receptor down-regulation. In line with our
Fig. 10. Time course of receptor mRNA (top left), free cytosolic receptor density (bottom left), TAT mRNA (top right) and TAT activity (bottom right)upon 0.1- (F) and 0.3 (E)-mg/kg/h 7-day infusions of methylprednisolone in male ADX Wistar rats. The solid and broken lines are the simulations usingthe fifth-generation model taking into account the effects on liver weights for the low and high doses (eqs. 7–14). The parameters used include thosefor acute steroid effects as listed in Table 2.
findings, down-regulation of cytoplasmic receptors has beenproposed to be one possible mechanism for the developmentof hormonal resistance during chronic therapy. Our results
suggest that the extent of tolerance in the biological effects,however, may be countered by alterations in signal transduc-tion processes by other hormones whose levels may change asa result of the global steroid effects during continuous dosing.
Fig. 11. Time course of TAT mRNA (top) and TAT activity (bottom) upon0.1- (F) and 0.3 (E)-mg/kg/h 7-day infusions of methylprednisolone inmale ADX Wistar rats. Symbols are the mean data and errors are thestandard deviations. The solid and broken lines are the fittings using thefifth-generation model for the low and high doses (eqs. 11–13). Thereceptor dynamics and liver weight ratios were fixed as listed in Table 2.The estimated parameters for long-term TAT dynamics are also indicatedin Table 2.
Fig. 12. Left, time course of cAMP concentrations in hepatic rat cytosol upon 0.1- (F) and 0.3 (E)-mg/kg/h infusions of methylprednisolone. Solid circlesare the mean data and bars are the standard deviations. The solid and broken lines connect the means. Right, cAMP concentrations in the controlsand the animals administered the 50-mg/kg i.v. bolus dose, which were sacrificed at 5 and 6 h.
Fig. 13. Cell trafficking model for redistribution of lymphocytes betweenthe blood and lymphoid organs (top) and the time course of predosenormalized blood lymphocyte count for the controls (f), the 0.1- (F) and0.3 (E)-mg/kg/h infusion groups (bottom). The kin and kout are the zero-and first order rates of transfer of lymphocytes to and from the bloodcompartment, The solid rectangle indicates inhibition of lymphocyteegress from the tissue and is defined by the inhibition function indicatedin eq. 16. The solid and broken lines are simultaneous fittings for the low-and high-dose groups by using eq. 16, whereas the dotted line representsthe mean counts (80%) in the control group.
Pharmacodynamics and Pharmacogenomics of Methylprednisolone 255
These studies assess corticosteroid kinetics and dynamicson the molecular, tissue, whole body, and mathematical mod-eling levels, and thus provide a physiological integrated ex-amination of how multiple factors interact to control the invivo responses to an important type of therapeutic agent.This also allows evaluation of the relevance of related mea-surements carried out in isolated systems such as cell cul-ture. The 7-day infusion of methylprednisolone in adrenalec-tomized rats provides organ responses that mimic adverseeffects during chronic dosing in humans; thus, this may be ahighly useful model system to further address methods ofavoiding or ameliorating such effects in therapeutic situa-tions. As the integration of gene microarray technology al-lows a more comprehensive examination of multiple genechanges, the present mRNA patterns indicate that such mea-surements will require interpretation in view of dose, dura-tion, pharmacokinetic, pharmacodynamic, and toxicity vari-ables controlling receptor/gene-mediated drug effects.
ReferencesAlexandrova M and Farkas P (1992) Stress-induced changes of glucocorticoid recep-
tor in rat liver. J Steroid Biochem Mol Biol 42:493–498.Baxter JD and Forsham PH (1972) Tissue effects of glucocorticoids. Am J Med
53:573–589.Boudinot FD, D’Ambrosio R, and Jusko WJ (1986) Receptor-mediated pharmacody-
namics of prednisolone in the rat. J Pharmacokinet Biopharm 14:469–493.Boushey HA (1998) Effects of inhaled corticosteroids on the consequences of asthma.
J Allergy Clin Immunol 102:S5–S16.Canvin JM and el-Gabalawy HS (1999) Anti-inflammatory therapy. Phys Med Re-
habil Clin N Am 10:301–317.Chow FS, Sharma A, and Jusko WJ (1999) Modeling interactions between adrenal
suppression and T-helper lymphocyte trafficking during multiple dosing of meth-ylprednisolone. J Pharmacokinet Biopharm 27:559–575.
Crettaz M, Muller-Wieland D, and Kahn CR (1988) Transcriptional and posttran-scriptional regulation of tyrosine aminotransferase by insulin in rat hepatomacells. Biochemistry 27:495–500.
Dayneka NL, Garg V, and Jusko WJ (1993) Comparison of four basic models ofindirect pharmacodynamic responses. J Pharmacokin Biopharm 21:457–478.
D’Argenio DZ and Schumitzky A (1997) ADAPT II User’s Guide: Pharmacokinetic/Pharmacodynamic Systems Analysis Software. Biomedical Simulations Resource,Los Angeles, CA.
David DS, Grieco MH, and Cushman P Jr (1970) Adrenal glucocorticoids aftertwenty years. A review of their clinically relevant consequences. J Chronic Dis22:637–711.
Diamondstone TI (1966) Assay of tyrosine aminotransferase activity by conversion ofp-hydroxyphenylpyruvate to p-hydroxybenzaldehyde. Anal Biochem 16:395–401.
DuBois DC, Almon RR, and Jusko WJ (1993) Molar quantification of specific mes-senger ribonucleic acid expression in northern hybridization using cRNA stan-dards. Anal Biochem 210:140–144.
DuBois DC, Xu ZX, McKay L, Almon RR, Pyszczynski N, and Jusko WJ (1995)Differential dynamics of receptor down-regulation and tyrosine aminotransferaseinduction following glucocorticoid treatment. J Steroid Biochem Mol Biol 54:237–243.
Ebling WF, Szefler SJ, and Jusko WJ (1985) Methylprednisolone disposition inrabbits. Analysis, prodrug conversion, reversible metabolism, and comparisonwith man. Drug Metab Dispos 13:296–304.
Ernest MJ and Feigelson P (1979) Multihormonal control of tyrosine aminotrans-ferase in isolated liver cells. Monogr Endocrinol 12:219–241.
Ferron GM and Jusko WJ (1998) Pharmacokinetic and pharmacoimmunodynamicinteractions between prednisolone and sirolimus in rabbits. Pharm Res 15:1888–1894.
Ferron GM, Pyszczynski NA, and Jusko WJ (1999) Pharmacokinetic and pharma-coimmunodynamic interactions between prednisolone and sirolimus in adrenalec-tomized rats. J Pharmacokinet Biopharm 27:1–21.
Fisher LE, Ludwig EA, and Jusko WJ (1992) Pharmacoimmunodynamics of meth-ylprednisolone: trafficking of helper T lymphocytes. J Pharmacokinet Biopharm20:319–331.
Ganss R, Weih F, and Schutz G (1994) The cyclic adenosine 3,5-monophosphate-and the glucocorticoid-dependent enhancers are targets for insulin repression oftyrosine aminotransferase gene transcription. Mol Endocrinol 8:895–903.
Haughey DB and Jusko WJ (1992) Bioavailability and nonlinear disposition ofmethylprednisolone/methylprednisone in the rat. J Pharm Sci 80:117–121.
Hashimoto S, Schmid W, and Schutz G (1984) Transcriptional activation of the ratliver tyrosine aminotransferase gene by cAMP. Proc Natl Acad Sci USA 81:6637–6641.
Holten D and Kenney FT (1967) Regulation of tyrosine alpha-ketoglutarate transam-inase in rat liver. VI. Induction by pancreatic hormones. J Biol Chem 242:4372–4377.
Ichii S (1981) Depletion and replenishment of glucocorticoid receptor in cytosols ofrat tissues after administration of various glucocorticoids. Endocrinol Jpn 28:293–304.
Kalinyak JE, Dorin RI, Hoffman AR, and Perlman AJ (1987) Tissue-specific regu-lation of glucocorticoid receptor mRNA by dexamethasone. J Biol Chem 262:10441–10444.
Kenney F, Lee KL, Reel JR, and Hoel DG (1970) Regulation of tyrosine alpha-ketoglutarate transaminase in rat liver. IX. Studies of the mechanisms of hor-monal inductions in cultured hepatoma cells. J Biol Chem 245:5806–5812.
Kong AN and Jusko WJ (1991) Disposition of methylprednisolone and its sodiumsuccinate prodrug in vivo and in perfused liver of rats: nonlinear and sequentialfirst-pass elimination. J Pharm Sci 80:409–415.
Lowry OM, Rosebrough NJ, Farr AL, and Randall RJ (1951) Protein measurementwith the Folin phenol reagent. J Biol Chem 193:265–272.
Miller AH, Spencer RL, Pearce BD, Pisell TL, Azrieli Y, Tanapat P, Moday H, RheeR, and McEwen BS (1998) Glucocorticoid receptors are differentially expressed inthe cells and tissues of the immune system. Cell Immunol 186:45–54.
Ramakrishnan R (2001) Fifth Generation Model for Corticosteroid Pharmacodynam-ics: Application to Steady-State Receptor Down-Regulation and Enzyme InductionPatterns upon Long-Term Continuous Administration of Methylprednisolone inRats, chapter 2. Ph.D. thesis, State University of New York at Buffalo, Buffalo,NY.
Reel JR, Lee KL, and Kenney FT (1970) Regulation of tyrosine alpha-ketoglutaratetransaminase in rat liver. 8. Inductions by hydrocortisone and insulin in culturedhepatoma cells. J Biol Chem 245:5800–5805.
Sapolsky RM, Krey LC, and McEwen BS (1984) Stress down-regulates corticosteronereceptors in a site-specific manner in the brain. Endocrinology 114:287–292.
Scherer G, Schmid W, Strange CM, Rowekamp W, and Schutz G (1982) Isolation ofcDNA clones coding for rat tyrosine aminotransferase. Proc Natl Acad Sci USA79:7205–7208.
Schmid E, Schmid W, Jantzen M, Mayer D, Jastorff B, and Schutz G (1987) Tran-scription activation of the tyrosine aminotransferase gene by glucocorticoids andcAMP in primary hepatocytes. Eur J Biochem 165:499–506.
Selby W (1993) Current management of inflammatory bowel disease. J GastroenterolHepatol 8:70–83.
Spencer CJ, Heaton JH, Gelehrter TD, Richardson KI, and Garwin JL (1978) Insulinselectively slows the degradation rate of tyrosine aminotransferase. J Biol Chem253:7677–7682.
Sun YN, DuBois DC, Almon RR, and Jusko WJ (1998a) Fourth-generation model forcorticosteroid pharmacodynamics: a model for methylprednisolone effects on re-ceptor/gene-mediated glucocorticoid receptor down-regulation and tyrosine amino-transferase induction in rat liver. J Pharmacokinet Biopharm 26:289–317.
Sun YN, DuBois DC, Almon RR, Pyszczynski NA, and Jusko WJ (1998b) Dose-dependence and repeated-dose studies for receptor/gene-mediated pharmacody-namics of methylprednisolone on glucocorticoid receptor down-regulation and ty-rosine aminotransferase induction in rat liver. J Pharmacokinet Biopharm 26:619–648.
Swartz SL and Dluhy RG (1978) Corticosteroids: clinical pharmacology and thera-peutic use. Drugs 16:238–255.
Wicks WD (1968) Tyrosine-alpha-ketoglutarate transaminase: induction by epineph-rine and adenosine-3,5-cyclic phosphate. Science (Wash DC) 160:997–998.
Yeung D and Oliver IT (1968) Induction of phosphopyruvate carboxylase in neonatalrat liver by adenosine 3,5-cyclic monophosphate. Biochemistry 7:3231–3239.
Yoshida A, Noguchi T, Taniguchi S., Mitani Y, Ueda M, Urabe K, Adachi T, OkamuraY, Shigemasa C, and Abe K (1986) Receptor dynamics and tyrosine aminotrans-ferase induction during the course of chronic treatment of rats with glucocorticoid.Endocrinol Jpn 33:769–775.
Address correspondence to: Dr. William J. Jusko, Department of Pharma-ceutical Sciences, 565 Hochstetter Hall, School of Pharmacy and Pharmaceu-tical Sciences, State University of New York at Buffalo, Buffalo, NY 14260.E-mail: [email protected]
