Jeremy CohenRoyal Brisbane & Women’s Hospital

Queensland, Australia

Raging Hormones

in the Critically Ill

“An Expert is someone who has made all the mistakes that can be made, in a narrow field.”

–Niels Bohr

“Buck up, it’s probably just a virus”

–An Expert

Severe Stress

Moderate Stress

Mild Stress

RAI

CIRCI

Cor

tisol

nm

ol/l

0

1000

2000

3000

4000

Rothwell Annane Schein Sibbald Bollaert Sam Bouachour Schroeder Beishuizen Aygen Jarek De Jong Dimopoulou Oppert

Survivors Non Survivors

Cortisol values in Sepsis

“There is no greater tragedy in Science than to witness the slaying of a beautiful hypothesis by an ugly fact”

–Thomas Huxley

Glucocorticoid Mechanism of Action

GR-ALPHA

GR-BETA

GR-GAMMA,A,P

second antibody (antirabbit IgG) was substituted with NDS.Controls were always negative. Fixed, but unstained, cellsshowed only minimal autofluorescence. There were no de-tectable differences among the various treatment conditionsin the morphology of the cells or in the area of the entire cellor nucleus.

Effect of IL-1! on cytosolic GR binding

We evaluated [3H]Dex-binding sites using an in vitro ex-change assay of cytoplasmic homogenates. When most of thereceptor is in the unactivated state, cytosolic GR binding isa measure of the amount of receptor in the cells, and there-fore, its evaluation may be useful to investigate expression ofGR protein. L929 cells were treated with vehicle or IL-1!(1000 U/ml) for 24 h. As shown in Fig. 3A, IL-1! induced anincrease in cytosolic GR binding (up-regulation). The in-crease in cytosolic binding (!23%) is consistent with theincrease in the cytoplasmic fluorescent signal (!23%), thussuggesting that both of these techniques reveal increasedlevels of GR protein induced by treatment with IL-1! for 24 h.Protein concentrations were examined in every sample, andno effect of IL-1! treatment on protein content was detected[vehicle, 0.676 (se, 0.157) mg/ml; IL-1!, 0.607 (se, 0.122)mg/ml; P " NS].

Effects of IL-1 on GR-mediated gene transcription

To determine the functional correlates of IL-1!-inducedchanges in nucleocytoplasmic traffic and protein expression,we examined GR-mediated gene transcription in L929 cellsstably transfected with a CAT enzyme reporter gene that isunder hormonal control by virtue of several HREs residingin the upstream MMTV promoter (LMCAT cells). Accordingto our data on GR nucleocytoplasmic traffic and GR binding,IL-1! alone induced GR up-regulation in the absence of

steroids, and cells treated with IL-1! and Dex showed lessGR translocation than cells treated with the same dose of Dexalone. Therefore, to determine whether receptor up-regula-tion was accompanied by increased functional activity andwhether blockade of Dex-induced GR translocation was as-sociated with a relative decrease in functional activity, theimpact of the same treatment protocols on GR-mediated genetranscription in the presence or absence of various doses ofDex was determined. LMCAT cells were grown in steroid-free medium and treated with vehicle, IL-1! (100 and 1000U/ml) for 24 h, Dex (10 nm to 1 "m) for 1.5 h, or IL-1!(10–1000 U/ml) for 24 h followed by incubation of IL-1! (10and 1000 U/ml) plus Dex (10 nm to 1 "m) for 1.5 h. Resultsare presented in Figs. 3B and 4 and are expressed as relativeCAT activity (percentage of vehicle-treated control samples).Treatment with IL-1! (1000 U/ml) alone for 24 h induced anincrease in GR-mediated gene transcription (!24%). Thisincrease (!24%) was consistent with the !23% up-regulationinduced by the same dose of IL-1! described previously inthe immunostaining and binding data (Figs. 1, 2, and 3A). Incontrast, at every concentration of Dex, pretreatment withIL-1! (1000 U/ml) caused a decrease in Dex-induced GR-mediated gene transcription compared with that in cellstreated with the same dose of Dex alone (Fig. 4). Dex-inducedGR-mediated gene transcription was reduced by 4% at the10-nm dose of Dex and about 20% at the 100-nm and 1-"mDex doses (Fig. 4). As under physiological conditions, hor-mone and cytokine would occur together rather than in tan-dem as represented in the previous experiment, cells werecoincubated with IL-1! and Dex for 24 h and compared withcells treated with Dex alone for 24 h. Previous experiments

FIG. 2. Quantitation of GR fluorescence in cytoplasm and nucleus ofcells treated with IL-1! in the presence or absence of Dex. Cells weretreated as indicated in Fig. 1. Quantitative analysis of fluorescencewas performed in blindly selected regions from the cytoplasm andnucleus using digital image analysis. Results are based on quanti-tation of approximately 460 cells from three to five independent ex-periments and are expressed as the mean # SEM percentage of thesignal in vehicle-treated cells. *, Significant (P $ 0.05) difference vs.vehicle; !, significant (P $ 0.05) difference vs. Dex (10 nM), usingStudent-Newman-Keuls post-hoc analysis.

FIG. 3. Increased cytosolic GR binding and GR-mediated gene tran-scription after treatment with IL-1! alone. A, L929 cells were treatedwith vehicle (!) or IL-1! (1000 U/ml; f) for 24 h. Cell lysates werecentrifuged, and the supernatant-cytosol was added to incubationsolutions containing radiolabeled (3H) Dex for the exchange assay.Values are presented as the mean # SEM GR binding in vehicle- andIL-1!-treated cells from five independent experiments. *, Significant(P $ 0.05) difference vs. vehicle, using Student’s t test [t(23) " 3.27;P " 0.003]. B, LMCAT cells were grown in steroid-free medium andtreated with vehicle (!) or IL-1! (1000 U/ml; f) for 24 h. Cells werefractionated, and lysates were analyzed for relative CAT enzymeactivity (presented as a percentage of vehicle activity). The results areshown as the mean # SEM from seven independent experiments. *,Significant (P $ 0.05) difference vs. vehicle using Student’s t test[t(12) " 2.35; P " 0.036].

4362 IL-1! INHIBITS GR TRANSLOCATION AND FUNCTION Endo • 1999Vol 140 • No 9

by on February 2, 2009 endo.endojournals.orgDownloaded from

have demonstrated that under these conditions (Dex treat-ment for 24 h), GR down-regulation does not occur andtherefore would not confound interpretation of the results(35). Coincubation of cells with Dex plus IL-1! (100 and 1000U/ml) led to marked reduction (42%) in relative CAT activitycompared with treatment with Dex alone (Fig. 5A). In ad-dition, to confirm the specificity of the effects of IL-1! onGR-mediated gene transcription, IL-1ra was added to thecombination of IL-1! plus Dex (10 nm). To optimize thestochiometry of the antagonist treatments in these experi-ments, the lower dose of IL-1! (100 U/ml) was used incombination with a 1000-fold excess of IL-1ra calculated ona mg per mg basis. As shown in Fig. 5B, IL-1ra completelyeliminated the inhibitory effects of IL-1! on Dex-inducedGR-mediated gene transcription.

DiscussionIL-1! inhibits GR translocation and function and inducesGR up-regulation

The present study was conducted to investigate the effectsof the proinflammatory cytokine, IL-1!, on GR translocationand function. We used an in vitro experimental system em-ploying the mouse fibroblast cell line L929, a cell line that hasbeen previously described by our group and others to besensitive to the effects of GR agonists, antagonists, and non-steroid GR modulators, including antidepressants (31), heatshock (30), and cytokines (36). We also used a panel of quan-titative assays to investigate GR nucleocytoplasmic traffic,GR expression, and GR-mediated gene transcription. Ourresults demonstrate that 24-h treatment with IL-1! inhibitsDex-induced translocation of the GR from cytoplasm to nu-cleus. Moreover, 24-h pretreatment with IL-1! or 24-h coin-cubation with IL-1! led to significant reduction in Dex-

induced GR-mediated gene transcription, an effect that wasreversed by IL-1ra. Finally, evaluation of both immunocy-tochemical staining of the GR and cytosolic GR binding incells treated with IL-1! for 24 h provided evidence of GRup-regulation.

These results contribute to the identification of a possiblemechanism by which cytokines can induce glucocorticoidresistance. As previously noted, several studies have shownthat in vivo and in vitro treatment with proinflammatorycytokines, including IL-1, decreases GR function (GR resis-tance). For example, in vivo treatment of mice or in vitrotreatment of a hepatoma cell line with IL-1! and IL-1" hasbeen found to decrease glucocorticoid-stimulated inductionof the gluconeogenesis enzyme, phosphenolpyruvate car-boxykinase (15, 17, 18). Of note is that the degree of maxi-mum inhibition of enzyme induction in these previousstudies (!25%) is comparable to the percent inhibition ofGR-mediated gene transcription found in the presentstudy (20–40% inhibition). Other studies have reported sim-ilar results. For example, in vitro treatment with the IL-1

FIG. 4. Inhibition of Dex-induced GR-mediated gene transcriptionafter pretreatment with IL-1!. LMCAT cells were grown in steroid-free medium and treated with vehicle plus Dex (10 nM to 1 #M) for 1.5 h(!), or IL-1! (1000 U/ml) for 24 h followed by coincubation of IL-1!(1000 U/ml) and Dex (10 nM to 1 #M) for 1.5 h (f). Cell lysates wereanalyzed for relative CAT enzyme activity (presented as a percentageof vehicle activity). The results are shown as the mean " SEM from fiveindependent experiments. There was a significant [F(2,1) # 32.1; P $0.0001] main effect of IL-1! plus Dex treatment vs. Dex alone usingtwo-way ANOVA. *, Significant (P $ 0.05) difference using Student’st test [t(3) # 5.3; P # 0.013].

FIG. 5. Inhibition of GR-mediated gene transcription after 24-h co-incubation of Dex and IL-1!; reversal by IL-1 receptor antagonist.LMCAT cells were grown in steroid-free medium and treated withvehicle (!), Dex (10 nM) for 24 h (f), Dex (10 nM) plus IL-1! (100–1000U/ml) for 24 h ( ), or Dex (10 nM), IL-1! (100 U/ml), and IL-1ra for24 h ( ); in panel B only). Cells were fractionated, and lysates wereanalyzed for relative CAT enzyme activity (presented as a percentageof vehicle activity). A, Results depicted are from a representativeexperiment in which the mean " SEM of five or six separate deter-minations per treatment group are shown. There was a significantmain effect of treatment condition on GR-mediated gene activity us-ing one-way ANOVA [F(2,16) # 10.41; P $ 0.002]. These results werereplicated in two (1000 U/ml) to six (100 U/ml) independent experi-ments. *, Significant (P $ 0.05) difference from the indicated groupvs. the vehicle plus Dex group using the Student-Newman-Keuls test.B, Results depicted are from a representative experiment in which themean " SEM of three to six separate determinations per treatmentgroup are shown. There was a significant main effect of treatmentcondition on GR-mediated gene activity using one-way ANOVA[F(2,11) # 31.40; P $ 0.001]. These results were replicated in threeindependent experiments. *, Significant (P $ 0.05) difference fromvehicle plus Dex using the Student-Newman-Keuls test.

IL-1! INHIBITS GR TRANSLOCATION AND FUNCTION 4363

by on February 2, 2009 endo.endojournals.orgDownloaded from

Inhibition of GR translocation and function by IL-1

Pariante et al 1999

Webster et al, PNAS 2000

11-b

HSD

-1 A

ctiv

ity

0

7.5

15

22.5

30

Liver Adipose

Sepsis Control

11-Beta HSD-1 Activity in Rats with Experimentally Induced Sepsis

p < 0.03

p < 0.02

Boonen et al, NEJM 2013

“I think you’ll find its a bit more complicated than that”

–An Expert

LPS

LPSDexamethasone

Cytokine Release

Cytokine Release

Cardinal et al , JCEM 2010

Cohen et al , unpublished

data

TNFa

0.0

0.2

0.4

0.6

0.8

1.0

Ra&o

((Dex./Max.s&m

ulated)

Pa&ents Controls

IL#6

0.0

0.2

0.4

0.6

0.8

1.0

Ra'o

)(Dex./Max.s'm

ulated)

Pa'ents Controls

Cohen et al , unpublished data

Relative mRNA expression in patients and controls

Cohen et al , unpublished data

Queensland Princess Alexandra Royal Brisbane & Womens Ipswich Logan Mater Health-Public Mater Private HospitalNambour RedcliffeToowoombaWesley

NSWBlacktownCalvary Mater – NewcastleJohn HunterLiverpoolNepeanPrince of WalesRoyal North shoreRoyal Prince AlfredSt GeorgeSt Vincent’s SydneyTamworthWoollongong

VictoriaAustinBendigoDandenongGeelongMonashNorthernRoyal MelbourneSt Vincent’s MelbourneWestern

South AustraliaLyell McEwinQueen Elizabeth

Western AustraliaFremantleRoyal Perth

Northern TerritoryDarwin

New ZealandAuckland DCCMAuckland CVICUChristchurchMiddlemoreNorth ShoreTaurangaWellingtonWaikato

ADRENAL 42 SITES IN AUSTRALIA & NEW ZEALAND

Republic of IrelandCork University Hospital

United KingdomSt Georges NHS + 7 sites

DenmarkRighospitalet

IndiaChristian Medical College Manipal HospitalColumbia Asia

Saudia ArabiaKing Abdul Aziz Medical CityKing Fahad Medical CityKing Faisal Medical City

ADRENAL ETHICS APPROVAL INTERNATIONALLY

“The time you perform a Google Image search for “Raging Hormones”on your work computer will be the same time your Director enters unannounced”

–An Expert