Prevention of autoimmune diabetes in NOD mice by troglitazoneis associated with modulation of ICAM-1 expression on pancreatic

islet cells and IFN-c expression in splenic T cellsq

Petra Augstein,a,* Annemarie Dunger,b Peter Heinke,a Gerhild Wachlin,c Sabine Berg,a

Bernd Hehmke,a and Eckhard Salzsiedera

a Institute of Diabetes ‘‘Gerhardt Katsch’’ Karlsburg e.V., Greifswalder Str.11e, Karlsburg 17495, Germanyb Fachkrankenhaus fur Orthopadie und Rheumatologie, Gommern-Vogelsang, Germany

c Department of Pathophysiology, Ernst-Moritz-Arndt University of Greifswald, Greifswald, Germany

Received 21 March 2003

Abstract

Thiazolidinediones acting as PPAR-c agonists are a new generation of oral antidiabetics addressing insulin resistance as a main

feature of type-2 diabetes. In accordance to our results, pre-clinical studies have demonstrated that the thiazolinedione troglitazone

prevents the development of insulin-dependent autoimmune type-1 diabetes. To investigate whether TGZ acts by affecting the

ICAM-1/LFA-1 pathway and/or the Th1/Th2 cytokine balance in NOD mice, we analysed the IL-1b-induced ICAM-1 expression

on islet-cells and the LFA-1, CD25, IL-2, IFN-c, IL-4, and IL-10 expression on splenocytes. After 200 days of oral TGZ admin-

istration, islet cells from TGZ-treated NOD mice showed a reduced ICAM-1 expression in response to the pro-inflammatory cy-

tokine IL-1b. The expression of the ligand LFA-1 on CD4þ and CD8þ T-cells was comparable to that of placebo- and untreated

controls. Also, the expression of Th1/Th2 cytokines was comparable in groups receiving TGZ or Placebo. Nevertheless, the in-

vestigated NOD mice segregated into IFN-c low- and IFN-c high producers as revealed by cluster analysis. Interestingly, the

majority of TGZ-treated mice belonged to the cluster of IFN-c low producers. Thus, the prevention of autoimmune diabetes in

NOD mice by TGZ seems to be associated with suppression of IL-1b-induced ICAM-1 expression leading to a reduced vulnerability

of pancreatic b-cells during the effector stage of b-cell destruction. In addition, IFN-c production was modulated, implicating that

alteration of the Th1/Th2 cytokine balance might have contributed to diabetes prevention. The findings of this study suggest that

TGZ exerts its effects by influencing both the b-cells as the target of autoimmune b-cell destruction and the T-cells as major effectors

of the autoimmune process.

� 2003 Elsevier Science (USA). All rights reserved.

Keywords: Type-1 diabetes; NOD mice; Troglitazone; ICAM-1; Pancreatic islet cells; Cytokine expression; Splenocytes

Thiazolidinediones (TDZs) which share a commonthiazolidine-2-4-dione structure [1] are a new generation

of oral antidiabetics addressing insulin resistance [2–4]

by enhancement of insulin responsiveness of skeletal

muscle, adipose tissue, and liver [5–7]. TDZs exert theirinsulin sensitising effects by binding to nuclear recep-

tors—peroxisome proliferator-activated receptors

(PPARs) [5,8,9] which in turn bind to another nuclear

receptor before gene transcription becomes activated

[1,7,10–12]. TDZs improve glycemic control in people

with type-2 diabetes through their action at PPAR-c1and PPAR-c2, and alter lipid metabolism by affecting

PPAR-a [13], all belonging to the nuclear hormone re-ceptor superfamily. The mediated effects include the

reduction of plasma glucose, insulin, triglycerides, and

fatty acids [13]. Beside this, TDZs have been shown to

Biochemical and Biophysical Research Communications 304 (2003) 378–384

www.elsevier.com/locate/ybbrc

BBRC

qAbbreviations: FCS, fetal calf serum; HE, hematoxylin; ICAM-1,

intercellular adhesion molecule 1; IL-1b, interleukin 1b; IL-4, inter-leukin 4; IL-10, interleukin 10; IFN-c, interferon c; LFA-1, leucocyte

function-associated antigen 1; MFI, relative fluorescence intensity;

NF-jB, nuclear factor jB; TDZs, thiazolidinediones; TGZ, troglitaz-

one.* Corresponding author. Fax: +49-38355-68-444.

E-mail address: augstein@mail.uni-greifswald.de (P. Augstein).

0006-291X/03/$ - see front matter � 2003 Elsevier Science (USA). All rights reserved.

doi:10.1016/S0006-291X(03)00590-4

possess anti-inflammatory properties [14] in several hu-man diseases [15–18]. The observed immune modulatory

effects [19] have been linked to a variety of mechanisms

including impaired cytokine production of activated

human lymphocytes [20], suppression of pro-inflamma-

tory mediators [18], and down-regulation of adhesion

molecules [21–25].

In animal models for type-1 diabetes, TDZs are able

to reduce T-cell-mediated inflammatory processes asso-ciated with autoimmune b-cell destruction, thereby

preventing and delaying the onset of type-1 diabetes [26–

28]. The aim of this study was to examine whether TGZ

prevents autoimmune diabetes by modulation of ICAM-

1 expression in pancreatic islet cells and/or by affecting

the ICAM-1/LFA-1 pathway in association with an in-

fluence on the T-cell cytokine profile. Both pathways

have been shown to be effected by TGZ in other diseases[22–25,29,30] and are also thought to be involved in the

progression of autoimmune islet inflammation (insulitis)

to overt type-1 diabetes [20,31–36].

Materials and methods

Animals and isolation of pancreatic islets and splenocytes. NOD

mice were purchased from the MØllegard-Breeding-Centre (M&B, A/

S, Denmark) and housed at the animal facilities of the Institute of

Diabetes ‘‘Gerhardt Katsch’’ Karlsburg. TGZ (500mg� kg�1 � d�1)

was administered orally to 28-days-old female mice (n ¼ 20) by ad-

dition of the drug to the drinking water. Equally, the control group

(n ¼ 20) received placebo containing the tablet core ingredients. Ten

mice remained untreated. Diabetes development was monitored by

biweekly measurements of urinary glucose. Hyperglycemia was con-

firmed by blood glucose determination and diabetes diagnosed if

blood glucose levels exceeded P12mmol/L. After treatment for 30

weeks the mice were sacrificed by general anesthesia. Spleens were

removed and splenocytes isolated by Percoll denisity gradient

(f ¼ 1:086) centrifugation (Amersham Pharmacia Biotech, Freiburg,

Germany). After taking biopsies for diagnosis of b-cell destruction by

insulitis grading and insulin staining [37], remaining tissue was ap-

plied to islet isolation procedure. Islets from NOD mice and BB/OK

rats (8–12 days old; Department of Laboratory Animal Science,

Medical Faculty, University of Greifswald) were isolated by means of

fractionated collagenase digestion (Serva, Heidelberg, Germany;

8mg/15ml Hanks balanced salt solution [38]). BB/OK rat islets were

purified by density gradient centrifugation [39]. Islets were hand-

picked and pre-cultured (37 �C, 5% CO2) in RPMI 1640 (Bio-

Whittaker, Verviers, Belgium) supplemented with 10% FCS and

antibiotics.

In vitro induction and analysis of ICAM-1 expression. Islets from

BB/OK rats were cultured for 24 h in the presence of IL-1b (10U/ml),

TNF-a (500U/ml), and IFN-c (500U/ml) purchased from Biosource,

Camarillo, California. Controls were kept untreated. Islets of TGZ-,

placebo-treated and untreated mice were exposed to 20U/ml IL-1b(Dunn, Asbach, Germany), TGZ (10lmol/L), the combination of both

or remained untreated as controls. After 24 h of cytokine exposure,

islets were disintegrated into single islet cells by fractionated dispase

digestion [38] and analysed for ICAM-1 expression using FITC-la-

belled anti-mouse CD54 antibody (BD Pharmingen, Heidelberg,

Germany). TGZ-mediated inhibition (%) of IL-1b-induced ICAM-1

expression was calculated by setting the ICAM-1 expression induced

by IL-1b alone as 100%.

Staining for surface markers and cytokines in T-cell subsets.

Splenocytes were subjected to detection of LFA-1 (integrin aL chain,

clone 2D7) and CD25 surface expression (IL-2 receptor a chain; clone

PC 61) according to standard protocols or were stimulated in vitro for

intracellular cytokine staining as described before in detail [40]. Briefly,

1:5� 106 splenocytes were restimulated with phorbol 12-myristate 13-

acetate (10 ng/ml; Sigma, Taufkirchen, Germany) and ionomycin (1lg/ml, Sigma) in the presence of brefeldin A (10lg/ml, Sigma) for 4 h

(37 �C, 5% CO2). Afterwards the cells were fixed and permeabilised

using Permeafix (Ortho Diagnostics Systems, Neckargem€uund, Ger-

many) as recommended by the supplier. After washing with PBS

containing 5% FCS and EDTA (0.2mmol/L), cells were double stained

for CD4 (clone L3T4) or CD8 (clone Ly-2) surface markers with

FITC-labelled antibodies and for one intracellular cytokine (IL-2,

clone S4B6; IFN-c, clone XMG1.2; IL-4, clone 11B11; and IL-10,

JES5-16E3) with PE-labelled monoclonal antibodies. Each test in-

cluded primary control antibodies with the same immunoglobulin

subclass as the corresponding cytokine antibodies. To evaluate the

cytokine staining procedure, the percentage of positive stained cells

obtained with cytokine antibodies was compared to that obtained with

isotype-matched control antibodies by t test (CD4/IL-2, p < 0:01;

CD4/IFN-c, p < 0:01; CD4/IL-4, p < 0:01; CD4/IL-10, p < 0:01;

CD8/IL-2, p < 0:01; CD8/IFN-c, p < 0:01; CD8/IL-4, ns; and CD8/

IL-10, p < 0:01). All antibodies used for flow cytometry were pur-

chased from BD Pharmingen (Heidelberg, Germany) and applied at

immunoglobulin concentrations according to the recommendations

given by the supplier. The samples were analysed on an EPICS/XL

flow cytometer (Coulter, Krefeld, Germany) equipped with the System

II Software, version 3.0. Cytokine expression was evaluated by

quantifying the percentage cytokine and T-cell marker double-positive

cells, the cytokine-positivity (%) of the two T-cell subsets and mean

relative fluorescent intensity (MFI in log units).

Statistical analysis. Kaplan–Meier survival analyses were used to

estimate the time for diabetes development. Survival curves were

compared using the log rank test. To analyse intracellular cytokine

expression, a cluster analysis with squared Euclidean distances of all

mice was performed to separate the animals in distinct groups. Each

cluster included mice with similarities for selected parameters.

Results are expressed as means�SD. Statistical significance was

assessed using unpaired Student�s t test. Differences were considered to

be statistically significant at a level of p < 0:05.

Results

TGZ reduces the incidence of diabetes in NOD mice and

ICAM-1 expression on islet cells in response to IL-1b

At the end-point of the study the TGZ-treated group

had a reduced diabetes incidence of 5.0% compared to

31.1% (p < 0:05) and 43.8% (p < 0:05) in placebo and

untreated animals, respectively (Figs. 1 and 2).

To investigate effects of TGZ-treatment on ICAM-1

expression, pancreatic islets of TGZ- and placebo-trea-ted and control mice were isolated and exposed in vitro

to IL-1b as strong inducer of ICAM-1 (Figs. 3A and B)

in the presence or absence of TGZ. In general, TGZ re-

duced the IL-1b-induced ICAM-1 expression of NOD-

mouse islets (p < 0:01; Fig. 3). However, TGZ exerted

the strongest inhibitory effects on islets from TGZ-trea-

ted mice. Calculation of the TGZ-mediated inhibition

(%) demonstrated that TGZ suppressed the IL-1b-in-duced ICAM-1 expression to 57:6� 9:6% (n ¼ 9;

P. Augstein et al. / Biochemical and Biophysical Research Communications 304 (2003) 378–384 379

p < 0:01) in islets derived from TGZ-treated NOD mice

compared to islets from placebo-treated mice and con-

trol mice which showed an inhibition of 32:2� 8:0%(n ¼ 11) and 35:0� 7:0% (n ¼ 4), respectively.

TGZ modulates intracellular IFN-c production of CD4þ

and CD8þ T-cells

Surface expression of LFA-1 and CD25 as well asintracellular expression of IL-2, IFN-c, IL-4, and IL-10

were determined in CD4þ and CD8þ splenic T cells. Asdemonstrated in Table 1, there were no differences in the

percentage binding of tested T-cell markers between the

NOD mouse groups.

Cluster analysis including TGZ- and placebo-treated

mice was performed to distinguish NOD mice on the

basis of intracellular cytokine expression (Fig. 4). As

shown in Fig. 5, NOD mice segregated into two clusters

characterised by different amounts of intracellular IFN-c production in CD4þ and CD8þ T-cells as indicated by

differences of the MFI values. The IFN-c-low producers

were grouped in cluster A, whereas cluster B included

the IFN-c-high producers.

Discussion

Thiazolidinediones (TDZs) originally designed for

ameliorating insulin resistance represent a new class of

therapeutic agents with increasing importance in the

treatment of type-2 (non-insulin-dependent) diabetes

[4,7,9]. Since their introduction in the late 1990s of the

last century, the most important drugs pioglitazone,

troglitazone (now withdrawn), and rosiglitazone have

been subjected to intensive pre-clinical research. Recentstudies reported that TDZs have cardioprotective

[16,41], antihypertensive [42], anti-inflammatory [18,24],

and immune modulatory properties [20] as well and

point to a broad range of therapeutic effects of these

Fig. 2. Islets from non-diabetic TGZ-treated mice were stained with HE (A,B) and insulin antibody (C,D) for diagnosis of b-cell destruction. TGZ

treatment resulted in a heterogeneous level of insulitis reaching from nearly unaffected (A,C) up to severely infiltrated islets (B,D).

Fig. 1. Development of autoimmune diabetes is reduced by TGZ in

NOD mice (p < 0:05 vs. TGZ). TGZ (j) and placebo (s) were ad-

ministered orally to 28-days-old female NOD mice and diabetes de-

velopment was monitored up to >30 weeks of age compared to

untreated controls (�).

380 P. Augstein et al. / Biochemical and Biophysical Research Communications 304 (2003) 378–384

substances [19]. Moreover, we observed the preventionof autoimmune type-1 diabetes by TGZ similar to

studies performed before using TGZ [27], pioglitazone

[28], and rosiglitazone [27]. To gain further insight into

the mechanisms of action accounting for diabetes pre-

vention, we examined ICAM-1 expression at the level of

pancreatic islet cells and LFA-1-, CD25-, and Th1/Th2-

cytokine expression at the level of splenic T cells.

After long-term treatment with TGZ, islet cells fromNOD mice showed a reduced ICAM-1 expression in

response to the pro-inflammatory cytokine IL-1b, whichis considered to be an important mediator of b-cell cy-totoxicity [32,43]. Consequently, the present investiga-

tions confirm the results of a preceding in vitro study onrat islets where a reduction of ICAM-1 expression on b-cells after IL-1b treatment in the presence of TGZ was

observed [44]. In accordance to our results IL-1b [45],

TNF-a, and IFN-c [46,47], all released in the microen-

vironment of infiltrated islets [32], have been reported to

induce the expression of ICAM-1. Generally, ICAM-1 is

thought to initiate the first contact between effector T

cells and target cells by antigen non-specific interactionswith its ligand LFA-1. In the course of antigen recog-

nition by the T cell, the formation of the immunological

synapse [48,49] is thought to finalise responses to the

target cell including killing by effector cells. In autoim-

mune diabetes, ICAM-1 seems to affect the generation

and/or expansion of islet-specific T-cells [50] and, ap-

parently, plays a key role during the effector phase of

b-cell destruction [50], especially after T-cell co-stimu-lation through B7-1 [51]. Taking the strong evidence for

autoimmune b-cell destruction by T-cell-mediated

mechanisms into account [36,52], it might be argued that

the TGZ-mediated reduction of cytokine-induced

ICAM-1 expression on rodent islet cells [44] prevents the

high affinity attachment of autoreactive T cells to the b-cell surface, thereby enabling b-cells to survive. Impor-

tantly, the blockade of the ICAM-1/LFA-1 pathway hasbeen shown to be a successful strategy to prevent au-

toimmune diabetes in various animal models [33,35,53].

Besides the observed effects, it might be reasonable to

assume that TGZ affected ICAM-1 expression also at

earlier stages of the insulitis process. ICAM-1 expression

appears early in the course of disease development

[54,55] and has been shown to contribute importantly to

the trafficking of T cells from the blood into the islets[56,57]. Both PPAR-c activators [23,24] and TGZ [22]

inhibit ICAM-1 expression on activated endothelial cells

which leads to a reduced homing of monocytes and

macrophages at the site of the disease [24]. Furthermore,

additional effects observed with TGZ and other ligands

of PPAR-c and PPAR-a such as reduction of myocar-

dial infarct size have been linked to the reduction of

ICAM-1 expression [21,58]. While PPAR-c and nuclearfactor jB (NF-jB) are known to contribute to the reg-

ulation of ICAM-1 expression [41,59], both have been

shown to be affected by TGZ and other glitazones.

However, it cannot be excluded that other mechanisms

thought to be involved in the pathogenesis of type-1

diabetes are affected by TGZ. Thus, we observed in a

former study a diminished generation of nitric oxide

after cytokine treatment of rat islets [60]. In this context,it is interesting that rosiglitazone and pioglitazone exert

their anti-inflammatory effects in adjuvant-induced ar-

thritis via inhibition of the NF-jB pathway [25].

LFA-1, the ligand of ICAM-1, is the most important

integrine expressed by leukocytes that regulate the ini-

tiation of an immune response and transendothelial

leukocyte migration [61]. At least at the present

Fig. 3. Capacity of cytokines to induce ICAM-1 on pancreatic islet

cells from BB/OK rats detected as percentage binding (A) and antigen

density (MFI, B). IL-1b and the combination of TNF-a plus IFN-c-induced ICAM-1 expression. The effect mediated by IL-1b was not

further enhanced by TNF-a and IFN-c (A,B). TGZ (C) suppressed the

up-regulation of ICAM-1 by IL-1b on pancreatic islet cells from NOD

mice after TGZ-treatment (n ¼ 9) for >30 weeks, in comparison to

placebo-treated (n ¼ 11) or untreated controls (n ¼ 4). After in vitro

exposure to IL-1b alone or in combination with TGZ for 24 h in

comparison to untreated � or TGZ-treated controls, islets were

disintegrated and ICAM-1 surface expression analysed by flow cy-

tometry. Data are given as mean binding (%)� SD (**p < 0:01 vs. IL-

1b-exposed islets from placebo-treated NOD mice).

P. Augstein et al. / Biochemical and Biophysical Research Communications 304 (2003) 378–384 381

Table 1

CD4 and CD8 positive splenocytes from non-diabetic TGZ- and placebo-treated NODmice were analysed for expression of surface markers (LFA-1,

CD25) and intracellular cytokines (IL-2, IFN-c, IL-4, and IL-10)

T-cell subset Marker TGZ Placebo

Double positive splenocytes Positivity Double positive splenocytes Positivity

(%) MFI (%) (%) MFI (%)

CD4þ LFA-1 26:2� 8:9 36:3� 13:9 99:5� 0:7 22:8� 3:6 29:3� 13:2 99:1� 0:8

CD25 4:6� 1:1 25:4� 15:6 15:9� 2:6 4:2� 0:6 14:7� 5:5 17:8� 3:5

IL-2 5:7� 1:2 12:8� 2:6 23:1� 6:5 5:1� 1:8 11:7� 2:6 27:2� 9:8

IFN 4:1� 1:4 36:8� 10:1 15:4� 5:3 3:9� 2:0 37:2� 11:0 20:4� 7:7

IL-4 0:4� 0:4 5:6� 2:8 2:0� 3:3 0:3� 0:2 4:6� 1:0 1:9� 1:4

IL-10 0:4� 0:3 6:2� 6:0 1:5� 1:4 0:4� 0:3 10:5� 16:9 2:0� 2:1

CD8þ LFA-1 9:5� 6:0 33:1� 13:8� 99:3� 1:1 9:3� 6:3 17:6� 10:2 98:2� 1:7

CD25 0:1� 0:1 76:3� 103:8 1:5� 1:4 0:2� 0:3 17:3� 11:7 1:6� 1:8

IL-2 0:5� 0:3 7:2� 8:4 7:2� 5:7 0:5� 0:4 5:5� 1:8 10:3� 3:4

IFN 1:5� 0:4 31:7� 10:3 18:4� 6:2 1:4� 0:6 36:4� 14:8 17:5� 8:6

IL-4 0:2� 0:2 4:2� 2:4 2:7� 5:3 0:1� 0:1 3:7� 1:2 0:3� 0:5

IL-10 0:1� 0:1 5:0� 5:7 1:0� 2:1 0:1� 0:1 5:6� 2:4 1:4� 1:2

Results are shown as percentage of double-positive splenocytes and the corresponding mean fluorescent intensity (MFI) as well as percentage

positivity in the CD4þ and CD8þ T-cell subset. Data are given as means�SD (*p < 0:05 vs. placebo).

Fig. 4. Representative flow cytometric histograms demonstrating the intracellular expression of IL-2 and IFN-c in CD4þ and CD8þ T-cells of TGZ-

treated NOD mice (A, CD8þ/INF-cþ; B, CD4þ/INF-cþ; C, CD8þ/IL-2; D, CD4þ/IL-2).

382 P. Augstein et al. / Biochemical and Biophysical Research Communications 304 (2003) 378–384

time-points and by investigation of the peripheral lym-

phocyte compartment, we found no evidence that TGZ

modulates LFA-1 expression on peripheral T cells. In

addition to the ICAM-1/LFA-1 pathway, we have ex-amined whether TGZ alters the CD25 surface marker

and/or changes the Th1/Th2 cytokine balance which is

thought to be important for the extent of autoimmune

b-cell destruction [62–65]. Suppressor T cells, such as the

CD4þ=CD25þ subset, are reduced in type-1 diabetes [63]

and have been shown to have important immunoregu-

latory functions [64]. In spite of these earlier reports, we

observed no alteration of CD25 on CD4þ and CD8þ

splenocytes and detected no differences in the expression

of Th1/Th2 cytokines between the TGZ- and placebo

groups. However, by cluster analysis of cytokine ex-

pression, the investigated NOD mice segregated into

two clusters consisting of IFN-c low and high produc-

ers. The majority of TGZ-treated mice belonged to the

group of IFN-c low producers. In support of a report

about reduced production of IFN-c by human T-cells invitro [20], our present findings indicate that TGZ mod-

ulates the level of IFN-c production. These results

provide a possible explanation for the observed pre-

vention of autoimmune diabetes by TGZ. It is con-

ceivable that TGZ might arrest the mononuclear cell

infiltration of the islets at the stage of the Th2-mediated

benign insulitis. The progression of benign insulitis into

destructive insulitis is critically dependent on enhance-ment of IFN-c production [31,32,36] and according to

the present findings, might be abrogated by TGZ

treatment. Likewise, other immune intervention thera-

pies interfering with progression and diabetes develop-ment were shown to induce a Th1-to-Th2 shift both in

islet-invading and splenic T cells [66].

Taken together, the prevention of autoimmune type-1

diabetes by TGZ seems to be mediated by at least two

distinct mechanisms. The first mechanism refers to the

reduction of IL-1b-induced ICAM-1 expression on

pancreatic islet cells and strongly suggests that ICAM-1

is a target molecule of TGZ-affected gene expression inautoimmune type-1 diabetes. The second mechanism

comprises diminution of IFN-c production in peripheral

T cells and points to a TGZ-induced modulation of the

Th1/Th2 cytokine balance. Nevertheless, additional

studies of the cytokine production capacity of islet-in-

vading T cells and of ICAM-1 expression in earlier

stages of periinsular islet infiltration are needed to fur-

ther elucidate the proposed mechanisms of action ofTGZ.

Acknowledgments

This work was supported by a grant from the Ministerium f€uur

Bildung, Wissenschaft und Kultur Mecklenburg-Vorpommern IDK 97

007 80/SOM and IDK 97 007 80/HSP III. The authors thank the

Glaxo Wellcome Company for funding the present investigation.

References

[1] T.M. Willson, J.M. Lehmann, S.A. Kliewer, Ann. N.Y. Acad. Sci.

804 (1996) 276–283.

[2] R.A. DeFronzo, Neth. J. Med. 50 (1997) 191–197.

[3] S. Mudaliar, R.R. Henry, Annu. Rev. Med. 52 (2001) 239–257.

[4] A.R. Saltiel, J.M. Olefsky, Diabetes 45 (1996) 1661–1669.

[5] F. Picard, J. Auwerx, Annu. Rev. Nutr. 22 (2002) 167–197.

[6] C. Otto, M. Lehrke, B. Goke, Pharmacogenomics 3 (2002) 99–

116.

[7] C. Furnsinn, W. Waldhausl, Diabetologia 45 (2002) 1211–1223.

[8] B.M. Spiegelman, Diabetes 47 (1998) 507–514.

[9] M. Stumvoll, H.U. Haring, Ann. Med. 34 (2002) 217–224.

[10] J.M. Lehmann, L.B. Moore, T.A. Smith Oliver, W.O. Wilkison,

T.M. Willson, S.A. Kliewer, J. Biol. Chem. 270 (1995) 12953–

12956.

[11] J. Berger, D.E. Moller, Annu. Rev. Med. 53 (2002) 409–435.

[12] T.M. Willson, M.H. Lambert, S.A. Kliewer, Annu. Rev. Biochem.

70 (2001) 341–367.

[13] U. Smith, S. Gogg, A. Johansson, T. Olausson, V. Rotter, B.

Svalstedt, FASEB. J. 15 (2001) 215–220.

[14] P. Delerive, J.C. Fruchart, B. Staels, J. Endocrinol. 169 (2001)

453–459.

[15] R.F. Grimble, Curr. Opin. Clin. Nutr. Metab. Care 5 (2002) 551–

559.

[16] K.W. Buchan, D.G. Hassall, Med. Res. Rev. 20 (2000) 350–366.

[17] B. Farrow, B.M. Evers, Surg. Oncol. 10 (2002) 153–169.

[18] A. Chawla, Y. Barak, L. Nagy, D. Liao, P. Tontonoz, R.M.

Evans, Nat. Med. 7 (2001) 48–52.

[19] X. Zhang, H.A. Young, Int. Immunopharmacol. 2 (2002) 1029–

1044.

[20] A.E. Giorgini, P.E. Beales, A. Mire Sluis, D. Scott, R. Liddi, P.

Pozzilli, Horm. Metab. Res. 31 (1999) 1–4.

Fig. 5. Cluster analysis of intracellular IFN-c expression in splenocytes

from non-diabetic NOD mice treated with TGZ (j) and placebo (s)

for >30 weeks. According to the relative amount of IFN-c produced

by CD4þ and CD8þ T-cells (MFI), the mice segregated into two dis-

tinct clusters. Cluster A contained the low IFN-c producers and cluster

B (n ¼ 15) contained the high IFN-c producers.

P. Augstein et al. / Biochemical and Biophysical Research Communications 304 (2003) 378–384 383

[21] N.S. Wayman, Y. Hattori, M.C. McDonald, H. Mota Filipe, S.

Cuzzocrea, B. Pisano, P.K. Chatterjee, C. Thiemermann, FASEB.

J. 16 (2002) 1027–1040.

[22] N.G. Chen, X. Han, Biochem. Biophys. Res. Commun. 282 (2001)

717–722.

[23] L. Cominacini, U. Garbin, A.F. Pasini, A. Davoli, M. Campagn-

ola, A. Rigoni, L. Tosetti, V. Lo-Cascio, Cell Adhes. Commun. 7

(1999) 223–231.

[24] V. Pasceri, H.D. Wu, J.T. Willerson, E.T. Yeh, Circulation 101

(2000) 235–238.

[25] T. Shiojiri, K. Wada, A. Nakajima, K. Katayama, A. Shibuya, C.

Kudo, T. Kadowaki, T. Mayumi, Y. Yura, Y. Kamisaki, Eur. J.

Pharmacol. 448 (2002) 231–238.

[26] P.E. Beales, R. Liddi, A.E. Giorgini, A. Signore, E. Procaccini, K.

Batchelor, P. Pozzilli, Eur. J. Pharmacol. 357 (1998) 221–225.

[27] P.E. Beales, P. Pozzilli, Diabetes Metab. Res. Rev. 18 (2002) 114–

117.

[28] T. Takamura, H. Ando, Y. Nagai, H. Yamashita, E. Nohara, K.

Kobayashi, Diabetes Res. Clin. Pract. 44 (1999) 107–114.

[29] L.J. Saubermann, A. Nakajima, K. Wada, S. Zhao, Y. Terauchi,

T. Kadowaki, H. Aburatani, N. Matsuhashi, R. Nagai, R.S.

Blumberg, Inflamm. Bowel Dis. 8 (2002) 330–339.

[30] P. Gosset, A.S. Charbonnier, P. Delerive, J. Fontaine, B. Staels, J.

Pestel, A.B. Tonnel, F. Trottein, Eur. J. Immunol. 31 (2001) 2857–

2865.

[31] H. Kolb, Diabetes Metab. Rev. 13 (1997) 139–146.

[32] A. Rabinovitch, Diabetes Metab. Rev. 14 (1998) 129–151.

[33] S.M. Katz, F. Bennett, K. Stecker, J.H. Clark, T. Pham, M.E.

Wang, B.D. Kahan, S.M. Stepkowski, Cell Transplant. 9 (2000)

817–828.

[34] J. Mysliwiec, A. Kretowski, M. Kinalski, I. Kinalska, Immunol.

Lett. 70 (1999) 69–72.

[35] Y. Hasegawa, K. Yokono, T. Taki, K. Amano, Y. Tominaga, R.

Yoneda, N. Yagi, S. Maeda, H. Yagita, K. Okumura, et al., Int.

Immunol. 6 (1994) 831–838.

[36] I. Andre-Schmutz, C. Hindelang, C. Benoist, D. Mathis, Eur. J.

Immunol. 29 (1999) 245–255.

[37] P. Augstein, A.G. Elefanty, J. Allison, L.C. Harrison, Diabeto-

logia 41 (1998) 1381–1388.

[38] S. Witt, B. Ziegler, M. Ziegler, Acta. Histochem. 86 (1989) 111–

115.

[39] B. Hehmke, K.D. Kohnert, R. Odselius, Diabetes Res. 3 (1986)

13–16.

[40] B. Hehmke, E. Salzsieder, G.B. Matic, R.E. Winkler, M. Tiess, W.

Ramlow, Ther. Apher. 4 (2000) 296–302.

[41] N. Khandoudi, P. Delerive, I. Berrebi Bertrand, R.E. Bucking-

ham, B. Staels, A. Bril, Diabetes 51 (2002) 1507–1514.

[42] Q.N. Diep, M. El Mabrouk, J.S. Cohn, D. Endemann, F. Amiri,

A. Virdis, M.F. Neves, E.L. Schiffrin, Circulation 105 (2002)

2296–2302.

[43] T. Mandrup-Poulsen, Diabetologia 39 (1996) 1005–1029.

[44] U. Fischer, A. Dunger, Diabetologia 42 (1999) A156.

[45] A.K. Cardozo, M. Kruhoffer, R. Leeman, T. Orntoft, D.L.

Eizirik, Diabetes 50 (2001) 909–920.

[46] D. Chakrabarti, X. Huang, J. Beck, J. Henrich, N. McFarland,

R.F. James, T.A. Stewart, Diabetes 45 (1996) 1336–1343.

[47] B. Kuttler, A. Hartmann, H. Wanka, Transplantation 74 (2002)

440–445.

[48] S.K. Bromley, W.R. Burack, K.G. Johnson, K. Somersalo, T.N.

Sims, C. Sumen, M.M. Davis, A.S. Shaw, P.M. Allen, M.L.

Dustin, Annu. Rev. Immunol. 19 (2001) 375–396.

[49] T.N. Sims, M.L. Dustin, Immunol. Rev. 186 (2002) 100–117.

[50] B. Balasa, A. La Cava, K. Van Gunst, L. Mocnik, D. Balakrishna,

N. Nguyen, L. Tucker, N. Sarvetnick, J. Immunol. 165 (2000)

7330–7337.

[51] S.A. Camacho, W.R. Heath, F.R. Carbone, N. Sarvetnick, A.

LeBon, L. Karlsson, P.A. Peterson, S.R. Webb, Nat. Immunol. 2

(2001) 523–529.

[52] J.W. Yoon, H.S. Jun, Arch. Pharm. Res. 22 (1999) 437–447.

[53] N.K. van den Engel, M. an-Haack, S. Martin, H. Kolb, J. Mol.

Med. 80 (2002) 301–308.

[54] S. Martin, T. Hibino, A. Faust, R. Kleemann, H. Kolb, J.

Autoimmun. 9 (1996) 637–643.

[55] G. Papaccio, B. De Luca, F.A. Pisanti, J. Cell Biochem. 71 (1998)

479–490.

[56] C. Faveeuw, M.C. Gagnerault, F. Lepault, J. Immunol. 152

(1994) 5969–5978.

[57] R. Ludwig, M. Kretschmer, G. Caspar, J. Bojunga, A. Oldenburg,

P. Schumm-Draeger, M. Stegmuller, G. von Minckwitz, K.H.

Usadel, K. Kusterer, Immunology 98 (1999) 111–115.

[58] A. Frustaci, J. Kajstura, C. Chimenti, I. Jakoniuk, A. Leri, A.

Maseri, B. Nadal-Ginard, P. Anversa, Circ. Res. 87 (2000) 1123–

1132.

[59] K.A. Roebuck, A. Finnegan, J. Leukoc. Biol. 66 (1999) 876–

888.

[60] A. Dunger, S. Berg, U. Fischer, Diabetologia 41 (1998) A160.

[61] Y. van Kooyk, C.G. Figdor, Curr. Opin. Cell. Biol. 12 (2000) 542–

547.

[62] L.M. Bradley, V.C. Asensio, L.K. Schioetz, J. Harbertson, T.

Krahl, G. Patstone, N. Woolf, I.L. Campbell, N. Sarvetnick, J.

Immunol. 162 (1999) 2511–2520.

[63] A. Kukreja, G. Cost, J. Marker, C. Zhang, Z. Sun, K. Lin-Su, S.

Ten, M. Sanz, M. Exley, B. Wilson, S. Porcelli, N. Maclaren, J.

Clin. Invest 109 (2002) 131–140.

[64] L. Chatenoud, B. Salomon, J.A. Bluestone, Immunol. Rev. 182

(2001) 149–163.

[65] G. Papaccio, F.A. Pisanti, R.D. Montefiano, A. Graziano, M.V.

Latronico, J. Cell Biochem. 86 (2002) 651–664.

[66] H. Kolb, U. Worz-Pagenstert, R. Kleemann, H. Rothe, P.

Rowsell, S. Rastegar, F.W. Scott, Autoimmunity 26 (1997) 1–6.

384 P. Augstein et al. / Biochemical and Biophysical Research Communications 304 (2003) 378–384