Mechanisms of deficient interferon-γ production in atopic diseases

Mechanisms of deficient interferon-g production in atopicdiseases

T. JUNG, R. MOESSNER, K. DIECKHOFF, S. HEIDRICH and C. NEUMANN

Department of Dermatology, Go¨ttingen University, Go¨ttingen, Germany

Summary

Background The mechanisms responsible for an imbalanced cytokine response in atopicdiseases are still not understood. While impaired interferon-gamma (IFN-g) production maybe the result of a pathological T-cell/antigen-presenting cell (APC) interaction, evidencewas provided that the T cell itself may have an intrinsic defect to produce IFN-g.Objective To clarify whether impaired IFN-g production by T cells from patients withatopic dermatitis (AD) represents an intrinsic defect in producing IFN-g.Methods Effector T cells were generated from CD4þ CD45RAþ-naive precursors frompatients with AD and healthy control individuals by activation with anti-CD3þ anti-CD28MoAbs. Following restimulation, IFN-g production was measured by ELISA and flowcytometry.Results IFN-g production by atopic T cells was decreased compared with healthy T cells.IL-12 present at priming or high doses of IL-2 during the culture period, even in the absenceof IL-12, completely restored IFN-g production. Conversion of naive CD45RAþ toCD45R0þ effector cells did not differ between atopic and healthy donors’ T cells.Conclusion Impaired IFN-g production by T cells from atopic individuals is not the resultof an intrinsic, genetically fixed, defect to produce sufficient amounts of IFN-g. The dataprovides evidence that correction of an impaired TH1 response in AD may be successful atthe precursor T cell level.

Keywords: cytokines, human, inteferon-g, naive CD4þ T lymphocytes

Clinical and Experimental Allergy, Vol. 29, pp. 912–919. Submitted 17 September 1998;revised 10 November 1998; accepted 16 December 1998.

Introduction

A number of studies have clearly demonstrated thatimpaired interferon-gamma (IFN-g) associated withnormal or enhanced interleukin (IL)-4 production by acti-vated T cells represents a hallmark of an imbalancedimmune response in patients with atopic disorders, espe-cially atopic dermatitis (AD) [1]. To define new therapeuticstrategies, the cellular and molecular mechanisms of thisaltered immune response must be clarified. Particularly, it isof interest to discriminate whether the deficient IFN-g

response is due to an intrinsic defect of T cells or is theresult of an altered T-cell/antigen-presenting cell (APC)interaction.

APCs producing IL-12, IL-10 and prostaglandin E2

(PGE2) are able to modulate or even redirect T-cellresponses [2–5]. PGE2, able to decrease IFN-g productionby T cells, is strongly expressed in monocytes obtained frompatients with AD [6–8] and therefore may be a candidatemolecule for APC-mediated impairment of IFN-g produc-tion. On the other hand, the T cell itself may have anintrinsic inability to produce sufficient amounts of IFN-g

independent of costimulatory signals provided by APCs [9–12]. Since IFN-g production by T cells is mainly restrictedto committed CD45R0þ memory/effector T cells, it still isopen whether diminished IFN-g production develops duringdifferentiation of naive T cells to memory/effector T cells orwhether the naive T cell itself is altered in its IFN-g

production.To explore these alternatives, we established anin vitro

model to generate T effector cells from naive CD4þ

CD45RAþ cells in the absence of APCs. Following

Clinical and Experimental Allergy,1999, Volume 29, pages 912–919

912 q 1999 Blackwell Science Ltd

Correspondence: T. Jung, Novartis Research Institute, Molecular andCellular Biology Unit, Brunner Str. 59, A-1235 Vienna, Austria.

restimulation, IFN-g production was determined. The datashows that T effector cells from patients with AD producedlower levels of IFN-g compared with healthy control cells.IL-12 present at priming as well as high doses of IL-2completely restored IFN-g production. Thus, naive T cellsfrom atopic patients are not intrinsically deficient to producenormal levels of IFN-g.

Materials and methods

Patients and controls

A total of 13 healthy volunteers (mean age 26.1 years, range19–38) and 13 patients having severe atopic dermatitis(AD) (mean age 24.8, range 14–54) were enrolled in thisstudy. AD was diagnosed according to the criteria describedby Hanifin and Rajka [13]. Patients suffering from severeAD had a personal and family history of atopy, mean IgE inserum was 2237 kU/L (range 234–8720), mean IgE levelsfor the healthy control group were 45 kU/L (range 2–87).All atopic individuals had a positive Phadiatop test forinhalant allergens and detectable specific IgE (RAST class3 or 4) for house dust mite antigen. The mean SCORADindex [14] was 45.2 (range 15–71). None of the patientswere treated with immunosuppressive drugs.

Purification of naive T cells and generation of effector cells

T cells were purified by sheep red blood cell rosetting ofperipheral blood mononuclear cells (PBMCs). Naive pre-cursor T cells were purified by MACS as described [15]using a cocktail of MoAbs directed against CD8, CD45R0and HLA-DR (Becton Dickinson, Heidelberg, Germany),the purity of the depleted cells was>97% for CD45RAþ,> 90% for CD4þ, <2% for CD14, CD19 and HLA-DR.Contaminating cells were CD45RAþ CD56þ NK cells (5–8%). Cells were cultured in Iscove’s medium supplementedwith 5% FCS (Biochrom, Berlin, Germany) and werestimulated with immobilized anti-CD3 MoAbs (OKT-3,20mg/mL) in 96-well flat-bottom plates plus anti-CD28MoAbs (soluble 100 ng/mL, Becton Dickinson) in the pre-sence of IL-2 (20 U/mL or 100 U/mL) (Stratman Biotec,Hannover, Germany). Recombinant IL-4 and IL-12, pur-chased from R & D Systems (Wiesbaden, Germany), andneutralizing anti-IL-4 (clone 8F12, kindly provided by Ch.Heusser, Novartis, Basel) were added to the cultures at10 ng/mL and 10mg/mL, respectively, for the first 5 days.Thereafter, cells were washed and kept in medium contain-ing IL-2 (20 U/mL) until day 7. Cells were washed andrestimulated with immobilized anti-CD3 plus soluble anti-CD28 MoAbs in the presence of IL-2 20 U/mL (low dose) or100 U/mL (high dose), but in the absence of any othercytokines. During the IL-2 high-dose protocol only, cells

were harvested at day 9, washed and split 1:2 every secondday in medium containing 100 U/mL IL-2 until day 14.Cells were counted and restimulated either with anti-CD3þ

anti-CD28þ IL-2 or phorbol-ester (PMA, 10 ng/mL) andionomycin (1mmol/L) (Sigma, Deisenhofen, Germany).Supernatants were harvested after 48 h.

ELISA

IFN-g production was measured as described earlier usingthe capture MoAb 43-11 and biotinylated detection MoAb45–15, both kindly provided by Ch. Heusser [16]. Theoptical density given by the reaction with ABTS (2,20-azino-bis3-ethylbenzthiazoline-6-sulphonic acid) (Sigma)was read at 405 nm (Dynatech, Stuttgart, Germany). Thedetection limit was 20–40 pg/mL. Recombinant IFN-g waspurchased from Stratman Biotec.

Flow cytometry

CD45R isoform expression was measured at day 14 beforerestimulation using anti-CD45RA-FITC and anti-CD45R0-PE MoAbs (Becton Dickinson). Intracellular detection ofIFN-g production was performed as described [15,17]. Incase of PMA/ionomycin restimulation, brefeldin A(10mmol/L) was added and cells were harvested after 6 h.TCR-activated T cells were also harvested after 6 h, brefel-din A was added to the cultures after the first 2 h to ensurethat it did not interfere with early signalling. Cells werefixed with 4% paraformaldehyde, permeabilized with 0.1%saponin and stained with FITC-labelled anti-IFN-g antibo-dies (clone 45–15). Samples were measured on a PAS IIIflow cytometer (Partec, Mu¨nster, Germany) equipped with a50 mW argon laser and analysed by WinList 3.0 software(Verity Software House, Topsham, ME, USA).

Statistics

Mann–WhitneyU-test (Prism 2.01, GraphPad Software,San Diego, CA, USA) was used to compare the resultsobtained from the study groups for significant differences.Correlation was calculated using the Spearman test.

Results

Development of IFN-g deficient T effector cells from naiveprecursors in patients with atopic dermatitis

CD4þ CD45RAþ HLA-DR-purified T cells from patientswith AD and healthy controls were activated with anti-CD3þ anti-CD28 MoAbs in the presence of low-doseexogenous IL-2. Cells were restimulated twice, at day 7with anti-CD3þ anti-CD28 MoAbs (TCR stimulus) and atday 14 either with anti-CD3þ anti-CD28 MoAbs or PMA/

IFN-g production 913

q 1999 Blackwell Science Ltd,Clinical and Experimental Allergy, 29, 912–919

ionoymcin. An < sevenfold expansion of proliferating Tcells was observed under these conditions. Supernatantswere harvested after 48 h and IFN-g production was mea-sured by ELISA. Figure 1(a) shows that IFN-g productionwas significantly (P< 0.05) reduced in patients’ T cells

compared with healthy control T cells when restimulatedvia TCR. Restimulation with PMA/ionomycin inducedhigher levels of IFN-g similarly in both healthy and atopicT cells, the difference, however, between the groups was notstatistically significant (P¼ 0.09). An inverse correlationbetween IgE levels and IFN-g production followingTCR (r ¼ ¹0.8571,P¼ 0.0107), but not PMA/ionomycinstimulation (r ¼ 0.3, not significant) was observed. Figure 1(b)

914 T. Junget al.


(b)

% IF

N-

+γ

28

23

18

13

88

4

0

P < 0.01

Anti-TCR Anti-TCR Pl Pl

P < 0.05

(a)

IFN

- (

ng

/mL)

γ

175

125

75

2515

10

5

0

HealthyAtopic

P = 0.09


P < 0.05

Fig. 1. IFN-g production by T cells from patients with AD isimpaired. T effector cells were generated from CD4þ CD45RAþ-naive T cells from healthy and atopic (AD) individuals by culturingwith anti-CD3þ anti-CD28þ IL-2 (20 U/mL; low dose). Cells wererestimulated with anti-CD3þ anti-CD28 MoAbs (anti-TCR)(Healthy and AD, eachn¼ 8) or PMA/ionomycin (PI) (healthyand AD, eachn¼ 9). (a) Supernatants were harvested 48 h later andanalysed for IFN-g production by ELISA. Data are presented asmean (SEM). (b) Cells were harvested 6 h after restimulation andassessed for intracellular IFN-g production. Results show portionof IFN-g-producing cells as mean (SEM).

% IF

N-

+γ

P = 0.16

(b) 605040302010

8

4

0Anti-TCR Anti-TCR Pl Pl

(a)

IFN

- (

ng

/mL)

γ

350

250

150

5050403020100

HealthyAtopic


Fig. 2. Priming with IL-12 restores IFN-g production in atopic Tcells. T effector cells were generated from CD4þ CD45RAþ-naiveT cells from healthy and atopic (AD) individuals by culturing withanti-CD3þ anti-CD28þ IL-2 (20 U/mL; low dose) in the presenceof IL-12 (10 ng/mL) during the first 5 days of culture. Cells wererestimulated at day 14 with anti-CD3þ anti-CD28 MoAbs (anti-TCR) (healthy and AD eachn¼ 8) or PMA/ionomycin (PI)(healthy and AD, eachn¼ 9). (a) Supernatants were harvestedand analysed for IFN-g production by ELISA after 48 h. Data arepresented as mean (SEM). (b) Cells were harvested 6 h afterrestimulation and assessed for intracellular IFN-g production.Results show portion of IFN-g-producing cells as mean (SEM).

shows the corresponding frequencies of IFN-g-producingcells confirming a significant reduction of IFN-g productionby atopic T cells.

Priming with IL-12 nearly restores IFN-g production inatopic T cells

When naive CD4þ T-cells were cultured in the presence of

IL-12 during the first 5 days, IFN-g production of TCR-restimulated T cells increased in both healthy (2.3-fold) andatopic (10-fold) donors (Fig. 2a). Consequently, IFN-g

levels of atopic T cells were nearly normalized. Also,following PMA/ionomycin restimulation, a nearly completerestoration of IFN-g production was observed (Fig. 2a). Asexpected, the frequencies of IFN-g-producing cellsincreased compared with stimulation in the absence ofIL-12 (Fig. 2b) and no difference was seen between groupsof healthy and atopic subjects.

Priming in the presence of high-dose IL-2 corrects IFN-g

deficiency

To investigate the role of IL-2 for restoration of IFN-g

production, naive CD4þ T cells were cultured with highdoses of IL-2 (100 U/mL) and by washing and splitting thecells during the second week every second day. Thisprocedure expanded T cells up to 20-fold after 14 days.When these T effector cells were restimulated, IFN-g levelsbetween healthy and atopic T cells did not differ (Fig. 3a).This finding was confirmed at the single cell level (Fig. 3b)showing that the frequency of IFN-g-producing cells hadpredominantly increased in the group of AD patients.Additional priming with IL-12 induced a further twofoldup-regulation of IFN-g production in T cells from bothpatients and healthy subjects (data not shown). Thus, IL-2 athigh doses and the ability of T cells to expand optimally,completely restores IFN-g production in atopic T cells.



% IF

N-

+γ

(b) 28

23

18

13

88

4

0Anti-TCR Anti-TCR Pl Pl

(a)

IFN

- (

ng

/mL)

γ

150

100

50

0

HealthyAtopic


Fig. 3. High doses of IL-2 restore IFN-g production by T cellsfrom patients with AD. T effector cells were generated from CD4þ

CD45RAþ-naive T cells from healthy (n¼ 6) and atopic (AD,n¼ 6) individuals by culturing with anti-CD3þ anti-CD28þ IL-2(100 U/mL; high dose). Cells were restimulated at day 14 with anti-CD3þ anti-CD28 MoAbs (anti-TCR) or PMA/ionomycin (PI) and(a) supernatants were harvested and analysed for IFN-g productionby ELISA after 48 h. Data are presented as mean (SEM). (b) Cellswere harvested 6 h after restimulation and assessed for intracellularIFN-g production. Results show portion of IFN-g-producing cellsas mean (SEM).

CD45RA+ CD45RO–

CD45RA+ CD45RO+

CD45RA– CD45RO+

100

75

50

25

0

% g

ated

cel

ls

AD_1 Healthy_1 AD_2 Healthy_2

Fig. 4. Conversion of CD45RA to CD45R0 is incomplete afterculturing with low-dose IL-2. Phenotypic analysis was performedafter 14 days of culture either with low-dose IL-2 (20 U/mL)(AD_1 (n¼ 6) and healthy_1 (n¼ 6)) or with high dose IL-2(100 U/mL) (AD_2 (n¼ 5) and healthy_2 (n¼ 6)). The proportionof CD45RAþ CD45R0¹, CD45RAþ CD45R0þ and CD45RA¹

CD45R0þT cells was determined by flow cytometry and isexpressed as mean (SEM).

Conversion of CD45RAþ naive to CD45R0þ effector T cellsis equally efficient in healthy and atopic T cells and isprimarily dependent on IL-2

The development of effector T cells was determined byCD45R isoform expression. Figure 4 shows that the major-ity of CD45RAþ CD45R0¹ T cells from both healthy andatopic subjects converted to CD45R0þ CD45RA¹ cellsusing the IL-2 high-dose protocol, whereas only< 30% ofthe cells converted during the IL-2 low-dose protocol.Although priming with IL-12 restored IFN-g productionby atopic T cells, it did not further promote CD45Rconversion (data not shown).

IL-4 at priming inhibits IFN-g production in both atopic andhealthy CD4þ effector T cells

IL-4 is known to inhibit IFN-g production by activated Tcells [18]. Its production is thought to be enhanced under avariety of in vitro conditions in atopic T cells [19–22] andthe susceptibility to IL-4 was reported to be increased due toa mutation in the IL-4Ra-chain found in a subset of atopicpatients [23]. Therefore, the role of IL-4 for IFN-g produc-tion during the differentiation of naive T cells was explored.Under both IL-2 low (Fig. 5a) and high (Fig. 5b) conditions,IL-4 present at priming during the first 5 days of cultureinhibited IFN-g production following restimulation.Healthy and atopic T cells were similarly susceptible forIL-4-driven down-regulation of IFN-g production. How-ever, while IFN-g production was significantly (P<0.05)different between healthy and atopic T cells following theIL-2 low-dose protocol, theP-value was< 0.01 followingIL-4 treatment, indicating that IFN-g production by atopic Tcells was moderately more inhibited. The inhibition ofendogenously produced IL-4 by culturing in the presenceof a neutralizing antibody did not change IFN-g levelsduring the IL-2 high-dose protocol (Fig. 5b). Under IL-2low-dose conditions, a further down-regulation of IFN-g

production was observed in the healthy individuals, whileIFN-g production by atopic T cells did not alter (Fig. 5a).

Discussion

The current study shows that IFN-g production by poly-clonally activated atopic T effector cells generated fromCD4þ CD45RAþ naive precursors is impaired. IFN-g

production could be completely restored by priming withhigh doses of IL-2 or IL-12. Circumventing accessorysignals from APCs, this model strongly indicates thatimpaired IFN-g production in atopy is not due to an intrinsicdefect of T cells to produce IFN-g and that naive T cellsmay develop into potent IFN-g-producing cells underappropriate conditions.

Diminished IFN-g production by polyclonally activatedT cells from patients with atopic disorders has been reportedby a number of groups [7,9,10,19,21,24]. CD45RAþ-naivecord blood cells derived from new-borns with high risk todevelop atopy produce low levels of IFN-g. This argues fora diminished TH1-response as a very early, possibly pri-mary component of atopy [25,26]. This assumption wassupported by the observation that reduced IFN-g productionin cord blood cells was associated with the development ofAD [25,26]. In addition, short-term activated naiveCD45RAþ, as well as effector CD45R0þ T cells fromatopic children produced low amounts of IFN-g and IL-2[10], suggesting a deficiency at both the precursor, as well asthe effector level of T cells in AD. This phenomenon may be

916 T. Junget al.


(b)

IFN

- (

ng

/mL)

γ

75

50

25

0Medium IL-4 Anti-IL-4

(a) P < 0.05

IFN

- (

ng

/mL)

γ

15

10

5

0Medium IL-4 Anti-IL-4

HealthyAtopic

P < 0.01

Fig. 5. IL-4 inhibits IFN-g production by both atopic and healthyCD4þ effector T cells. CD4þ¹ effector cells were generated fromCD4þ CD45RAþ naive T cells from healthy and atopic individualsby culturing with anti-CD3þ anti-CD28þ low-dose IL-2 (a,n¼ 7)or high-dose IL-2 (b,n¼ 5) in the absence or presence of IL-4(10 ng/mL) or anti-IL-4 MoAb (10mg/mL) during the first 5 days ofculture. Cells were restimulated at day 14 with anti-CD3þ anti-CD28 MoAbs and supernatants were harvested and analysed forIFN-g production by ELISA after 48 h. Data are presented as mean(SEM).

due to the genetic background of the T cell and, as describedin an infectious disease model in mice [27], can beindependent of accessory signals from APCs.

To clarify this problem we established anin vitro modelto generate effector T cells from purified CD4þ CD45RAþ

CD45R0-HLA-DR-naive T cells from patients with AD inthe absence of accessory signals from APCs. Cells wereactivated by engagement of the TCR and CD28 in thepresence of IL-2 and restimulated after 7 and 14 days.After the second restimulation, IFN-g production wasdetermined. Effector T cells from atopic patients producedlow levels of IFN-g associated with low frequencies of IFN-g-producing cells. Since low IFN-g levels were negativelycorrelated with serum IgE levels, this observation confirmsthe hypothesis that atopic T cells are intrinsically deficientto produce sufficient amounts of IFN-g resulting in high IgElevels. In contrast, when naive T cells were primed in thepresence of IL-12 to promote IFN-g-producing cells [2],IFN-g production was restored to nearly normal levels. Thisstrongly indicates that impaired IFN-g production is notexplainable by an intrinsic, genetically fixed defect. On theother hand, IL-12 did not restore IFN-g production byactivated PBMCs [28] or purified, unfractioned T cellsfrom AD individuals [12], suggesting that naive T cells,but not effector T cells from atopic subjects, are susceptibleenough for IL-12 to up-regulate IFN-g production to normallevels.

When effector T cells were generated in the presence ofhigh doses of IL-2, IFN-g production was completelyrestored, even in the absence of IL-12. Since IL-2 produc-tion has been reported to be impaired in atopic T cells[1,10,29], low endogenous IL-2 production may play amajor role for the insufficient development of IFN-g-producing T cells. Therefore, one may argue that impairedIFN-g production by T cells from atopic subjects is in partdue to low endogenous IL-2 production.

Besides IL-2, other soluble mediators produced duringthe priming phase may contribute to an insufficient induc-tion of IFN-g-producing cellsin vivo. Studies demonstratingelevated levels of IL-4 by T cells from atopic patients [19–21,24] suggest that TH2 cytokines like IL-4 or IL-10 mayinhibit IFN-g production. However, levels of IL-4 in short-term activated CD45R0þ T cells [10] and in supernatants ofpurified activated T cells were normal [12]. In addition,IL-10 production by atopic T cells was not enhanced [12]. Inthe present study, neutralization of endogenously producedIL-4 by anti-IL-4 MoAbs did not change levels of IFN-g

during restimulation. This observation confirms previousstudies showing that IL-4 production by naive T cells is not[10,30] or only hardly detectable [31] and suggests that IL-4production by naive T cells from atopic patients is notenhanced. Interestingly, neutralization of endogenous IL-4inhibited IFN-g production by T cells from healthy

individuals during the IL-2 low-dose protocol. This phe-nomenon may be due to the fact that neutralizing antibodiesmay not only antagonize, but also agonize a cytokineactivity, as shown in a murine model for IL-4, IL-3 andIL-7 [32]. Taken together, T-cell derived IL-4 does not seemto play a major role for impaired IFN-g production in AD.Invivo, however, other IL-4-producing cells like mast cells[33] or basophils [34] may contribute to a local environmentenriched for IL-4. When naive T cells were primed with IL-4, IFN-g production decreased similarly in atopic andhealthy T cells. Therefore, the susceptibility of naive Tcells to IL-4 from healthy and atopic subjects does not seemto be different, which is in contrast with a recent reportdemonstrating a mutation in thea-chain of the IL-4 receptorwhich has been found to be associated with higher suscept-ibility to IL-4 in a subset of patients [23]. On the other hand,soluble mediators like IL-12 or PGE2 produced by APCsmay contribute to IFN-g production. Since IL-12 productionby atopic monocytes was reported to be decreased [8,35] ornormal [28], its contribution to impaired IFN-g productionremains presently unclear. PGE2 may down-regulate IFN-gproduction by primed naive T cells [36], decreases IL-12production by dendritic cells [5] and was found to beoverexpressed in monocytes from patients with AD [7,37].Therefore, it seems likely thatin vivo the interactionbetween APCs and naive T cells is altered. This mayresult in an incomplete activation of T cells associatedwith low IL-2 production, which additionally inhibits thedevelopment of IFN-g-producing cells.

To determine the degree of conversion of naive toeffector T cells, CD45R isoform expression was determinedbased on the observation that CD4þ CD45RAþ-naive Tcells differentiate towards CD4þ CD45R0þ effector T cellsduring long-term culturing [38]. Although humanCD45R0þ memory/effector T cells may reconvert to thenaive phenotypein vitro and in vivo [39], the dissection ofnaive and effector T cells based on CD45R isoform expres-sion still is a valid model since i) immature low CD45RAand low CD45R0 cells from human cord blood differentiateinto CD45RAþ T cells [40], and (ii) CD45RAþ T cells donot respond to recall-antigens, but to neo-antigens in thepresence of professional APCs [41]. Since CD45R0þ T cellsare more potent to produce IFN-g compared withCD45RAþ T cells, an incomplete conversion by atopicT cells might be causal for impaired IFN-g production inthe IL-2 low-dose protocol. However, frequencies ofCD45R T cells were not different between groups of healthyand atopic individuals, suggesting a similar differentiationin vitro. Therefore, the expression of CD45R0 may reflectconversion of naive to effector T cells, but is not necessarilycorrelated with the ability to produce IFN-g and, therefore,not a predictor of IFN-g production.

In summary, the present study shows that IFN-g



production by CD4þ effector T cells generatedin vitro fromnaive precursors from patients with AD is completelyrestored in the presence of either high doses of IL-2 or IL-12 during priming. This finding argues that naive T cells arenot genetically deficient to produce IFN-g and that redirec-tion of an impaired TH1 response in atopy seems to besuccessful in naive, but not in committed effector T cells.This may be important to develop therapeutic strategies toprevent atopic diseases.

Acknowledgements

This work was supported by Deutsche Forschungsge-meinschaft grant Ju 351/1–1. We would like to thankJoseph Eamer for proof reading the text.

References

1 Leung DY. Atopic dermatitis: the skin as a window into thepathogenesis of chronic allergic diseases. J Allergy ClinImmunol 1995; 96:302–18.

2 Trinchieri G. Interleukin-12: a proinflammatory cytokine withimmunoregulatory functions that bridge innate resistance andantigen-specific adaptive immunity. Annu Rev Immunol 1995;13:251–76.

3 de Waal Malefyt R, Yssel H, Roncarolo MG, Spits H, de VriesJE. Interleukin-10. Curr Opin Immunol 1992; 4:314–20.

4 van der Pouw Kraan TC, Boeije LC, Smeenk RJ, Wijdenes J,Aarden LA. Prostaglandin E2 is a potent inhibitor of humaninterleukin 12 production. J Exp Med 1995; 181:775–9.

5 Kalinski P, Hilkens CM, Snijders A, Snijdewint FG, Kapsen-berg ML. IL-12-deficient dendritic cells. Generated presenceprostaglandin E2 promote type 2 cytokine production inmaturing human naive T helper cells. J Immunol 1997;159:28–35.

6 Hanifin JM, Chan SC. Monocyte phosphodiesterase abnorm-alities and dysregulation of lymphocyte function in atopicdermatitis. J Invest Dermatol 1995; 105:84S–8S.

7 Chan S, Henderson WR Jr, Li SH, Hanifin JM. ProstaglandinE2 control of T cell cytokine production is functionally relatedto the reduced lymphocyte proliferation in atopic dermatitis. JAllergy Clin Immunol 1996; 97:85–94.

8 Snijders A, van der Pouw Kraan TC, Engel M et al. Enhancedprostaglandin E2 production by monocytes in atopic dermatitis(AD) is not accompanied by enhanced production of IL-6,IL-10 or IL-12. Clin Exp Immunol 1998; 111:472–6.

9 Reinhold U, Wehrmann W, Kukel S, Kreysel HW. Evidencethat defective interferon-gamma production in atopic dermati-tis patients is due to intrinsic abnormalities. Clin Exp Immunol1990; 79:374–9.

10 Jung T, Lack G, Schauer U et al. Decreased frequency ofinterferon-g- and interleukin-2-producing cells in patients withatopic diseases measured at the single cell level. J Allergy ClinImmunol 1995; 96:515–27.

11 Nakagawa S, Aiba S, Tagami H. Decreased frequency of

interferon-g-producing CD4þ cells in the peripheral blood ofpatients with atopic dermatitis. Exp Dermatol 1998; 7:112–8.

12 Jung T, Witzak K, Dieckhoff K et al. Mechanisms of deficientinterferon-gamma production in atopic diseases. IFN-g pro-duction is only partially restored by costimulation of peripheralT cells with IL-12, IL-2, IL-15, IL-18 or engagement of CD28.Clin Exp Allergy 1999; 29:207–16.

13 Hanifin JM, Rajka G. Diagnostic features of atopic dermatitis.Acta Derm Venereol (Suppl.) Stockh 1980; 92:44–7.

14 Severity scoring of atopic dermatitis. The SCORAD index.Consensus Report of the European Task Force on AtopicDermatitis. Dermatol 1993; 186:23–31.

15 Jung T, Schauer U, Heusser C, Neumann C, Rieger C. Detec-tion of intracellular cytokines by flow cytometry. J ImmunolMeth 1993; 159:197–207.

16 Bauer H, Jung T, Tsikas D, Stichtenoth DO, Fro¨lich JC,Neumann C. Nitric oxide inhibits the secretion of T-helper 1and T-helper 2 associated cytokines in activated human T-cells. Immunol 1997; 90:205–11.

17 Jung T, Wijdenes J, Neumann C, de Vries JE, Yssel H.Interleukin-13 is produced by activated human CD 45RAþ

and CD45R0þ T cells. Modulation by interleukin-4 and inter-leukin-12. Eur J Immunol 1996; 26:571–7.

18 Peleman R, Fargeas C, Wu JP, Delespesse G. IL-4 suppressesthe production of IFN-g by human lymphocytes. J Exp Med1989; 170:1751–61.

19 Rousset F, Robert J, Andary M et al. Shifts in interleukin-4 andinterferon-gamma production by T cells of patients withelevated serum IgE levels and the modulatory effects ofthese lymphokines on spontaneous IgE synthesis. J AllergyClin Immunol 1991; 87:58–69.

20 Renz H, Jujo K, Bradley KL, Domenico J, Gelfand EW, LeungDY. Enhanced IL-4 production and IL-4 receptor expression inatopic dermatitis and their modulation by interferon-gamma. JInvest Dermatol 1992; 99:403–8.

21 Jujo K, Renz H, Abe J, Gelfand EW, Leung DY. Decreasedinterferon gamma and increased interleukin-4 production inatopic dermatitis promotes IgE synthesis. J Allergy ClinImmunol 1992; 90:323–31.

22 Chan SC, Li SH, Hanifin JM. Increased interleukin-4 produc-tion by atopic mononuclear leukocytes correlates withincreased cyclic adenosine monophosphate-phosphodiesteraseactivity and is reversible by phosphodiesterase inhibition. JInvest Dermatol 1993; 100:681–4.

23 Khurana Hershey GK, Friedrich MF, Esswein LA, ThomasML, Chatila TA. The association of atopy with a gain-of-function mutation in thea subunit of the interleukin-4 receptor.N Engl J Med 1997; 337:1720–5.

24 Nakazawa M, Sugi N, Kawaguchi H, Ishii N, Nakajima H,Minami M. Predominance of type 2 cytokine-producing CD4þ

and CD8þ cells in patients with atopic dermatitis. J AllergyClin Immnol 1997; 99:673–82.

25 Tang ML, Kemp AS, Thorburn J, Hill DJ. Reduced interferon-gamma secretion in neonates and subsequent atopy. Lancet1994; 344:983–5.

26 Warner JA, Miles EA, Jones AC, Quint DJ, Colwell BM,Warner JO. Is deficiency of interferon gamma production by

918 T. Junget al.


allergen triggered cord blood cells a predictor of atopiceczema? Clin Exp Allergy 1994; 24:423–30.

27 Hsieh CS, Macatonia SE, O’Garra A, Murphy KM. T cellgenetic background determines default T helper phenotypedevelopment in vitro. J Exp Med 1995; 181:713–21.

28 Lester MR, Hofer MF, Gately M, Trumble A, Leung DY.Down-regulating effects of IL-4 and IL-10 on the IFN-gammaresponse in atopic dermatitis. J Immunol 1995; 154:6174–81.

29 Kapp A. Atopic dermatitis — the skin manifestation of atopy.Clin Exp Allergy 1995; 25:210–9.

30 Ehlers S, Smith KA. Differentiation of T cell lymphokine geneexpression: the in vitro acquisition of T cell memory. J ExpMed 1991; 173:25–36.

31 Demeure CE, Yang LP, Byun DG, Ishihara H, Vezzio N,Delespesse G. Human naive CD4 T cells produce interleu-kin-4 at priming and acquire a Th2 phenotype upon repetitivestimulations in neutral conditions. Eur J Immunol 1995;25:2722–5.

32 Finkelman FD, Madden KB, Morris SC et al. Anti-cytokineantibodies as carrier proteins. Prolongation of in vivo effects ofexogenous cytokines by injection of cytokine-anti-cytokineantibody complexes. J Immunol 1993; 151:1235–44.

33 Bradding P, Feather IH, Howarth PH et al. Interleukin 4 islocalized to and released by human mast cells. J Exp Med1992; 176:1381–93.

34 Brunner T, Heusser CH, Dahinden CA. Human peripheral

blood basophils primed by IL-3 produce IL-4 in response toIgE receptor stimulation. J Exp Med 1993; 177:605–14.

35 van der Pouw Kraan TC, Boeije LC, de Groot ER et al.Reduced production of IL-12 and IL-12-dependent IFN-gamma release in patients with allergic asthma. J Immunol1997; 158:5560–5.

36 Katamura K, Shintaku N, Yamauchi Y et al. Prostaglandin E2at priming of naive CD4þ T cells inhibits acquisition of abilityto produce IFN-gamma and IL-2, but not IL-4 and IL-5. JImmunol 1995; 155:4604–12.

37 Chan SC, Kim JW, Henderson WR Jr, Hanifin JM. Alteredprostaglandin E2 regulation of cytokine production in atopicdermatitis. J Immunol 1993; 151:3345–52.

38 Beverley PC. Functional analysis of human T cell subsetsdefined by CD45 isoform expression. Semin Immunol 1992;4:35–41.

39 Bell EB, Sparshott SM, Bunce C. CD4þ T-cell memory.CD45R subsets and the persistence of antigen — a unifyingconcept. Immunol Today 1998; 19:60–4.

40 Bofill M, Akbar AN, Salmon M, Robinson M, Burford G,Janossy G. Immature CD45RA(low)RO(low) T cells in thehuman cord blood. I. Antecedents of CD45RAþ unprimed Tcells. J Immunol 1994; 152:5613–23.

41 Mehta Damani A, Markowicz S, Engleman EG. Generation ofantigen-specific CD4þ T cell lines from naive precursors. Eur JImmunol 1995; 25:1206–11.



Documents

Mechanisms of deficient interferon-γ production in atopic diseases