Published online before print November 26, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1,3
* Institute of Medical Immunology and
Department of Dermatology, University Hospital Charité, Berlin Humboldt University, Berlin, Germany
2 Correspondence at current address: Bayer Schering Pharma AG, Target Discovery, D-13342 Berlin, Germany. E-mail: wolf-dietrich.doecke{at}bayerhealthcare.com
ABSTRACT
Cytokine and anticytokine treatments represent promising approaches for therapy of immune-mediated diseases. In humans, however, regulatory consequences of interference with the cytokine network are only partially understood. Biomarker analysis in clinical studies may help to overcome this complexity and provide novel information about the in vivo relevance of individual cytokines. We report systemic immunological effects of IL-10 therapy in 10 psoriasis patients during a 7-week treatment period followed by a 7-week observation period. IL-10 was given s.c. at 8 µg/kg/day or 20 µg/kg/3x/week, and a broad range of immunological biomarkers was analyzed in an extended kinetics (17 time-points) before, during, and after IL-10 therapy. Besides the expected anti-inflammatory effects (e.g., inhibition of LPS-induced cytokine secretion), we found unexpected effects, such as activation of NK cells and an increase in parameters indicating proinflammatory activity (C-reactive protein and soluble IL-2R). Furthermore, cumulative effects (IgE and IgA), loss of effect (IL-1R antagonist and IFN-
secretion), or counter-regulation during and rebound after IL-10 therapy (TNF-
and IL-12/IL-23 p40) were found. Remarkably, some alterations were retained long after the 7-week treatment period (IL-4 secretion, monocytic CD86, and TGF-β1). In summary, we found manifold effects of IL-10 far beyond the immediate anti-inflammatory activity considered initially. These findings may explain the rather disappointing clinical effects of IL-10 therapy in exacerbated inflammation but also hint to its role for sustained immunological reshaping. They further exemplify the importance of analyzing an extended kinetics of an entire panel of biomarkers for understanding the effects of therapeutic interference with the cytokine network.
Key Words: human • Th1/Th2 • inflammation • autoimmunity
INTRODUCTION
In recent years, therapeutic interventions with biologics have come more and more into medical focus [1 2 3 4 ]. Especially for biologics interfering with the cytokine network, it is important to rely on well-characterized biomarkers, as in vivo, there are problems such as pleiotropy and redundancy as well as counter-regulation and rebound, which may not play a role in the restricted in vitro settings.
Since its discovery in 1989, IL-10 gained increasing interest as a result of its great impact on immunoregulation [5
]. Many in vitro and animal experiments suggested the anti-inflammatory, immunosuppressive potential of this cytokine. IL-10 regulates nonadaptive and adaptive immune mechanisms amongst others by inhibiting the production of proinflammatory cytokines by monocytes, macrophages, and neutrophils; inhibiting antigen presentation by monocytes/macrophages and dendritic cells; inhibiting IFN- production by T lymphocytes and NK cells, thus inhibiting TH1 and promoting TH2 responses; and in mice, inhibiting TH17 responses [5
6
7
8
9
]. Its immunoregulatory role is demonstrated impressively by IL-10-deficient mice, which develop chronic inflammatory bowel disease [10
]. However, immunostimulatory properties such as the induction of acute-phase proteins in hepatocytic cells and activation of NK and cytotoxic T cells have also been described [11
12
13
], and IL-10 is a potent B cell activator [5
]. The first single-dose administration of recombinant human (rh)IL-10 in humans was performed in healthy volunteers in 1995, followed by other studies about healthy people [14
15
16
17
18
]. Thereby, i.v. injection of IL-10 was shown to inhibit proinflammatory and type 1 cytokine production [14
, 17
]. Consequently, beneficial effects of rhIL-10 application were suggested for immune diseases characterized by type 1 cytokine overexpression. Clinical trials with IL-10 application were performed with variable success in patients with inflammatory bowel disease, rheumatoid arthritis, transplant rejection, and psoriasis [19
20
21
22
23
24
]. In the latter, one open-label and two placebo-controlled trials have been performed in patients with ongoing disease, of which, two showed a significant reduction of the psoriasis area and severity index (PASI) score [22
23
24
].
In contrast to numerous reports about the immunological effects of IL-10 in vitro and in animal experiments, there is, however, only limited knowledge about the in vivo effects of IL-10 on human immune parameters, in particular, with regard to effects after long-term application [14 15 16 17 18 , 22 , 23 , 25 , 26 ].
The aim of the present study was to extensively characterize the systemic immunological effects of long-term IL-10 application in 10 psoriasis patients during a 7-week treatment period and a subsequent 7-week observation period. IL-10 was given s.c. at 8 µg/kg/day or 20 µg/kg/3x/week, and the clinical results of this study were reported by us in 1999 [22 ]. Systemic, immunological effects were investigated weekly and additionally at 6, 24, and 72 h after the first IL-10 application. We analyzed in detail concentrations of anti- and proinflammatory cytokines in plasma; blood sedimentation rates and systemic levels of acute-phase reactants, dehydroepiandrosterone sulfate (DHEA-S), and Igs; numbers of circulating leukocytes; phenotype and function of circulating monocytes; and phenotypes and functions of circulating lymphocytes.
MATERIALS AND METHODS
Patients and IL-10 application protocol
Ten otherwise healthy, adult patients with moderate-to-severe, chronic plaque psoriasis were chosen for treatment with IL-10 (for patient characterization, see Asadullah et al. [22
, 27
]). There was no systemic treatment or phototherapy during 7 weeks preceding the study and no treatment with any topical therapy for psoriasis other than with bland emollients during 3 weeks preceding the study. The Institutional Review Board of the Medical Faculty approved the study, and written, informed consent was obtained from all patients.
Over a period of 7 weeks, the psoriasis patients received s.c. nonlesional applications of rhIL-10 (Tenovil/SCH 52000; kindly provided by Schering Plough Research Institute/Essex Pharma, Munich, Germany) in a daily dose of 8 µg/kg body weight (n=5) or 20 µg/kg body weight/3x/week (n=5). Tenovil is produced by a strain of Escherichia coli, and the drug substance is a highly purified (>95% chromatographically pure), sterile, water-soluble, nonglycosylated, homodimeric protein with a calculated monomeric molecular weight of 18.7 kDa [28
]. For clinical use, Tenovil is provided as sterile powder and is reconstituted with sterile water immediately prior to injection [28
]. In CD-1 mice, s.c. administration of up to 8000 µg Tenovil/kg body weight (which would have corresponded to an 
500-fold higher dose than in our daily treated patients) did not result in toxicity (neither clinically nor laboratory-confirmed), indicating no relevant contamination with LPS or other inflammatory stimuli [28
]. Furthermore, Limulus amoebocyte lysate (LAL) testing of rhIL-10 at a concentration of 400 µg/ml revealed an LPS content of 0.162 IE/ml (Kinetic QCL test, BioWhittaker, Walkersville, MD, USA; sample dilution 1:100). According to the administration of rhIL-10, this corresponded to a maximal LPS exposition per application of 0.003 and 0.008 IE/kg body weight daily at 8 µg/kg body weight or 3x/week at 20 µg/kg body weight, respectively. Such LPS concentrations are far below the concentrations described for the induction of an inflammatory response in humans [29
, 30
].
After cessation of IL-10 therapy, patients were followed for an additional 7 weeks. There was no systemic treatment other than with IL-10 and no topical treatment other than with bland emollients during the whole treatment and observation period.
Asadullah et al. [22 ] reported the clinical results of this study. IL-10 treatment was well-tolerated, and antipsoriatic effects were found in nine out of 10 patients, resulting in a significant decrease of the PASI score by 55.3 ± 11.5% (mean±SEM; P<0.02; Wilcoxon test) [22 ]. The antipsoriatic effects were confirmed by the results of histological examination, demonstrating an increase in cutaneous IL-4 and a decrease in IL-8 mRNA expression after IL-10 treatment [27 ].
Immunological investigations
For determination of systemic, immunological effects, blood samples were obtained by venous puncture before and at different time-points during IL-10 therapy as well as during and at the end of the 7-week post-therapeutic observation period as indicated.
Soluble Mediators
Standard laboratory investigations included the 1- and 2-h blood sedimentation rates, the plasma levels of C-reactive protein (CRP) and neopterin, and the Ig concentrations in serum (IgM, IgA, IgE, IgG, IgG1–4). For analysis, blood samples were sent to the central laboratory of the University Hospital Charité (Berlin Humboldt University, Berlin, Germany).
Nonstandard laboratory investigations included IL-10, IL-1R antagonist (IL-1RA), TGF-β1, the common subunit p40 used by IL-12 and IL-23, TNF-, IL-6, soluble IL-2R (sIL-2R), LPS-binding protein (LBP), DHEA-S, cortisol, and procalcitonin (PCT). For determination of these soluble mediators, plasma and serum supernatants were harvested from venous blood samples within 15 min and were stored at –70°C until analysis. The following commercially available ELISA kits were used for cytokine determination in plasma: IL-10 (CytoscreenTM, Laboserv, Giessen, Germany); IL-1RA and TGF-β1 (QuantikineTM, R&D Systems, Wiesbaden, Germany); and IL-12 p40 (Medgenix, Ratingen, Germany). The concentrations of TNF-, IL-6, and sIL-2R and of LBP, DHEA-S, cortisol in plasma, and serum, respectively, were determined with a semiautomatic chemiluminescence-based ELISA system (ImmuliteTM, DPC Biermann, Bad Nauheim, Germany). PCT plasma levels were measured with the LUMItestTM PCT (Brahms Diagnostica GmbH, Berlin, Germany).
Leukocyte Differentiation and Flow Cytometry
Leukocyte populations in EDTA anticoagulated venous blood were determined from differential blood smear and white blood cell counts. Flow cytometric analyses were performed using a FACScanTM machine and LYSYS IITM software (BD Bioscience, Heidelberg, Germany).
Lymphocytic subpopulations, T cell activation, and proportions of transferrin receptor (CD71)-positive monocytes were determined in a routine setting in Ficoll-Paque-separated PBMCs. The following FITC- and PE-labeled mAb were used: T cells: CD3 (clone UCHT1), CD4 (13B8.2), and CD8 (B9.11); total and CD5+ B cells: CD19 (3G8) and CD5 (BL1a); NK cells: CD16 (3G8); CD71+ monocytes: CD14 (RMO52) and CD71 (YDJ1.2.2); T cell activation and effector/memory T cells: CD11a (25.3), CD25 (B1.49.9), CD45RA (ALB11), CD45RO (UCHL1), and CD71 (YDJ1.2.2; all from Beckman Coulter, Krefeld, Germany) and HLA-DR (L243; BD Bioscience).
The expression of accessory molecules and activation antigens on monocytes was assessed in EDTA anticoagulated blood. Briefly, 50 µl whole blood was incubated with 5 µl each respective fluorescence-labeled mAb for 30 min at 4°C followed by erythrocyte lysis with FACSTM lysing solution (BD Bioscience), washing, and flow cytometric measurement of 20,000 monocytes. Monocytic surface antigen expression was assessed by electronic gating on CD14-positive leukocytes and determination of the geometric mean of fluorescence intensity (mfi). The following FITC-, PE-, or PE-Cy5-labeled mAb were used: CD14 (clone RMO52) and CD80 (MAB104; Beckman Coulter); CD40 (5C3), CD86 (FUN-1), and HLA-DR (L243; all from BD Bioscience); and CD64 (32.2; Medac, Hamburg, Germany).
Ex Vivo Functional Assays
For analysis of ex vivo LPS-induced cytokine synthesis, heparin anticoagulated whole blood was diluted 1:2 with RPMI-1640 medium (Biochrom KG, Berlin, Germany) and stimulated with endotoxin from E. coli 0127:B8 (100 ng/ml; Sigma, Deisenhofen, Germany) at 37°C and 5% CO2. Secretion of TNF- and IL-1β was measured in supernatants of 4 h culture using the ImmuliteTM system. Concentrations of IL-12 p70 and IL-1RA were determined by commercially available ELISA kits (QuantikineTM, R&D Systems) after 24 h endotoxin stimulation. Correction by monocyte numbers yielded the cytokine secretion capacity per monocyte.
For analysis of Con A-triggered type 1 and type 2 cytokine secretion, heparin anticoagulated whole blood was diluted 1:5 with RPMI-1640 medium and stimulated with Con A (100 µg/ml; Sigma) at 37°C and 5% CO2 for 24 h. Cytokine concentrations in culture supernatants were determined by the respective commercial ELISAs: IFN- (Medgenix) and IL-2 and IL-4 (both CytoscreenTM, Laboserv).
For analysis of intracellular type 1 cytokine formation in T cells and NK cells, ex vivo stimulation of whole blood with subsequent intracellular staining of IFN- (T cells and NK cells), TNF- (T cells and NK cells), IL-2 (T cells), and IL-4 (T cells) followed by flow cytometric measurement was used. Briefly, heparinized whole blood was diluted 1:2 with RPMI-1640 medium (Biochrom KG) and stimulated with PMA (25 ng/ml; Sigma) and the calcium ionophore ionomycin (1 µg/ml; Sigma) in the presence of the cytokine secretion inhibitor Brefeldin A (10 µg/ml; BD Bioscience) for 4 h at 37°C and 5% CO2. Thereafter, lymphocyte surface antigens were stained using FITC-labeled anti-CD2 and PE-Cy5-labeled anti-CD3 mAb (both from Beckman Coulter) in a final concentration of 5% (v/v) for 20 min at 4°C in the dark. Then, erythrocyte lysis and cell fixation were performed using FACSTM lysing solution. After permeabilization of leukocytes by FACSTM permeabilizing solution, cells were stained intracellularly with PE-labeled anti-IFN-, anti-TNF-, anti-IL-2, and anti-IL-4 mAb, respectively [each at 20% (v/v); Becton Dickinson, San Jose, CA, USA], for 30 min at room temperature in the dark (all reagents from BD Bioscience). Following a washing step, flow cytometric measurement and analysis of T cells (CD2- and CD3-positive lymphocytes) and NK cells (CD2-positive and CD3-negative lymphocytes) for assessment of the geometric mfi of the respective population and the percentage of positive cells for each cytokine were done.
Statistical analysis
Statistical analyses were performed using SPSSTM software (SPSS Inc., Chicago, IL, USA). Significant changes over time (during and after IL-10 therapy) were assessed by Wilcoxon test for paired samples versus the values before therapy. At each time-point, comparison between the two IL-10 therapy regimes was performed using the Mann Whitney U test. Significant changes over time were tested using the Friedman test. For some parameters, single-point values are missing. If the Friedman test was not applicable as a result of missing values, the respective lowercase numbers are given. In all tests, an error probability of P < 0.05 was considered significant. Data are presented as the mean ± SD or SEM as indicated.
RESULTS
Concentrations of anti- and proinflammatory cytokines in plasma
IL-10 plasma concentrations strongly increased during therapy with rhIL-10 (Table 1
). Whereas daily IL-10 application (8 µg/kg body weight) resulted in permanent, strongly elevated IL-10 plasma levels, treatment 3x/week (20 µg/kg body weight) led to higher peak values (more than 500 pg/ml) but also to heavy fluctuations (Table 1)
. In the latter group, the marked loss of measured IL-10 plasma concentrations beginning with Day 7 resulted, at least partly, from the long interval (72 h) between the last IL-10 application and blood withdrawal in the weekly determination intervals. Interestingly, within the therapeutic period, significantly falling IL-10 concentrations (P<0.05; Friedman test) were also observed in the daily application group, indicating a changing IL-10 distribution/elimination pattern. After the end of therapy, however, the plasma concentrations of endogenous IL-10 did not differ from those before therapy (Table 1)
.
|
View this table: [in a new window] |
Table 1. Group Differences in IL-10 and sIL-2R Plasma Concentrations
|
When investigating the plasma levels of other anti-inflammatory cytokines, we observed a strong rise in the IL-1RA plasma concentration during the initial days of IL-10 therapy (Fig. 1A ; for group-specific data, see Supplemental Fig. 1). Surprisingly, the total (latent and active) plasma levels of TGF-β1 dropped sharply and remained suppressed for the duration of the IL-10 application (Fig. 1B ; for group-specific data, see Supplemental Fig. 3B). After the end of IL-10 therapy, TGF-β1 plasma concentrations, after a short re-increase, were diminished significantly until the end of the 7-week observation period, indicating lasting alterations of TGF-β production and/or turnover (Fig. 1B) .
 View larger version (25K): [in a new window] |
Figure 1. Concentrations of anti- and proinflammatory cytokines in plasma. In 10 patients with psoriasis, plasma concentrations of IL-1RA (A), total TGF-β1 (B), IL-12/IL-23 p40 (C), TNF- (D), and IL-6 (E) were determined before (open bars), during (solid bars), and after (shaded bars) a 49-day therapy with 8 µg rhIL-10/kg body weight daily (n=3–5) or 20 µg rhIL-10/kg body weight 3x/week (n=3–5). Data are given as the mean ± SEM consolidated from both treatment groups [**, P<0.01; *, P<0.05, vs. before therapy; Wilcoxon test for paired samples; the Friedman test for multiple paired samples (whole period) was significant for total TGF-β1 (P<0.05); for IL-12/IL-23 p40, the Friedman test could not be applied as a result of missing values].
|
secretion (see 
Fig. 6A
). Surprisingly, the plasma levels of TNF- showed a small but significant elevation for the duration of the IL-10 therapy (Fig. 1D
; for group-specific data, see Supplemental Fig. 3D). Remarkably, withdrawal of IL-10 led to a strong, additional, temporary increase of TNF- plasma concentrations (Fig. 1D)
. The pattern of IL-6 (Fig. 1E
; for group-specific data, see Supplemental Fig. 3E) seems to reflect an IL-10-triggered induction and was confirmed by increases in the plasma concentrations of acute-phase reactants (see Table 2
).
 View larger version (28K): [in a new window] |
Figure 2. Serum levels of IgE and IgA. (A) Serum concentrations of IgE were determined before, during (highlighted band at the x-axis), and after a 49-day therapy with 8 µg rhIL-10/kg body weight daily (n=2–5, open bars) or 20 µg rhIL-10/kg body weight 3x/week (n=2–5, shaded bars). Relative values in comparison with Day 0 (100%) are given as the mean ± SD (#, P<0.05, vs. before therapy in patients treated 3x/week; Wilcoxon test for paired samples; +, P<0.05, between patients groups; Mann Whitney U test). (B) In 10 patients with psoriasis, serum concentrations of IgA were determined before (open bar), during (solid bars), and after (shaded bars) a 49-day therapy with 8 µg rhIL-10/kg body weight daily (n=2–5) or 20 µg rhIL-10/kg body weight 3x/week (n=2–5). Data are given as the mean ± SEM consolidated from both treatment groups [**, P<0.01; *, P<0.05, vs. before therapy; Wilcoxon test for paired samples; the Friedman test for multiple paired samples (whole period) was significant for IgA (P<0.001)].
|
 View larger version (18K): [in a new window] |
Figure 3. Monocytic expression of HLA-DR and coaccessory molecules. In 10 patients with psoriasis, the monocytic expression of HLA-DR (A) and the coaccessory molecules CD86 (B), CD80 (C), and CD40 (D) were determined flow cytometrically before (open bars), during (solid bars), and after (shaded bars) a 49-day therapy with 8 µg rhIL-10/kg body weight daily (n=3–5) or 20 µg rhIL-10/kg body weight 3x/week (n=3–5). Geometric mfi are given as the mean ± SEM consolidated from both treatment groups [**, P<0.01; *, P<0.05, vs. before therapy; Wilcoxon test for paired samples; the Friedman test for multiple paired samples (whole period) was significant for monocytic HLA-DR (P<0.001), CD80 (P<0.001), and CD40 (P<0.001)].
|
 View larger version (20K): [in a new window] |
Figure 4. Ex vivo endotoxin-triggered cytokine secretion. In 10 patients with psoriasis, the secretion of TNF- (A), IL-1β (B), IL-12 p70 (C), and IL-1RA (D) was determined after 4 h (TNF- and IL-1β) and 24 h (IL-12 p70 and IL-1RA) endotoxin stimulation of whole blood samples drawn before (open bars), during (solid bars), and after (shaded bars) a 49-day therapy with 8 µg rhIL-10/kg body weight daily (n=3–5) or 20 µg rhIL-10/kg body weight 3x/week (n=3–5). Data are shown as the mean ± SEM consolidated from both treatment groups [**, P<0.01; *, P<0.05, vs. before therapy; Wilcoxon test for paired samples; the Friedman test for multiple paired samples (whole period) was significant with P<0.001 for all parameters].
|
 View larger version (20K): [in a new window] |
Figure 5. Alterations in lymphocytic subpopulations and activated T cells. In 10 patients with psoriasis, lymphocyte subpopulations (T cells, B cells, and NK cells) and T cell activation antigen expression were flow cytometrically determined before (open bars), during (solid bars), and after (shaded bars) a 49-day therapy with 8 µg rhIL-10/kg body weight daily (n=4–5) or 20 µg rhIL-10/kg body weight 3x/week (n=4–5). The proportions of HLA-DR+ T cells (A), CD16+ lymphocytes (NK cells; B), and the CD5+ B cell subset as a percentage of all B cells (C) and absolute numbers (D) are shown as the mean ± SEM values consolidated from both treatment groups [**, P<0.01; *, P<0.05, vs. before therapy; Wilcoxon test for paired samples; the Friedman test for multiple paired samples (whole period) was significant for all parameters (P<0.05 for HLA-DR+ T cells; P<0.001 for all other parameters)].
|
 View larger version (28K): [in a new window] |
Figure 6. Ex vivo Con A-triggered secretion of lymphokines (A–C) and PMA/ionomycin-induced, intracellular IFN- and TNF- production by NK cells (D and E). In 10 patients with psoriasis, ex vivo lymphokine production was determined before (open bars), during (solid bars), and after (shaded bars) a 49-day therapy with 8 µg rhIL-10/kg body weight daily (n=4–5) or 20 µg rhIL-10/kg body weight 3x/week (n=3–5). For assessing the type 1/type 2 cytokine balance, IFN- (A), IL-2 (B), and IL-4 (C) secretion was determined after 24 h Con A stimulation in whole blood culture. Data are shown as the mean ± SEM consolidated from both treatment groups [**, P<0.01; *, P<0.05, vs. before therapy; Wilcoxon test for paired samples; the Friedman test for multiple paired samples (whole period) was significant for IL-2 (P<0.001)]. The cytokine production capacity of NK cells was flow cytometrically assessed by determining the intracellular IFN- (D) and TNF- (E) production after 4 h PMA/ionomycin stimulation in whole blood. The geometric mfi is given as the mean ± SEM consolidated from both treatment groups (**, P<0.01; *, P<0.05, vs. before therapy; Wilcoxon test for paired samples).
|
|
View this table: [in a new window] |
Table 2. Blood Sedimentation Rate and Plasma Levels of Acute-Phase Reactants and DHEA-S
|
The serum levels of the adrenal androgen DHEA-S and the DHEA-S/cortisol ratio have been shown recently to decrease in several states of immunodepression [31 32 33 ], and DHEA-S therapy was demonstrated to augment immunoreactivity [34 ]. Interestingly, we found a significant decrease of DHEA-S concentrations during IL-10 therapy, which persisted, at least partly, until the end of the 7-week observation period (Table 2 ; for group-specific data, see Supplemental Fig. 1G). In contrast to this, the systemic levels of the adrenal steroidal stress hormone cortisol did not change significantly (data not shown), indicating a selective suppression of DHEA-S synthesis.
When looking at the Ig serum levels, we found no changes in the concentrations of IgM, total IgG, and IgG1–4 (data not shown). In contrast to this, during the IL-10 application period, IgE serum concentrations increased in patients treated daily (tendency with P=0.068 at Days 14, 28, 35, 42, 49, and 56) and in patients treated 3x/week (Fig. 2A ), and we found consistent differences between both treatment groups [Fig. 2A ; as a result of great interindividual variances before therapy (6.7–234.0 U/ml), data in Fig. 2A were given as mean relative changes of IgE levels, and absolute mean values were shown in the Supplemental Fig. 4A; data consolidated from both treatment groups revealed a highly significant increase in IgE serum levels during IL-10 therapy (data not shown)]. There was also a significant enhancement of IgA plasma levels, which remarkably sustained until the end of the observation period (Fig. 2B ; for group-specific data, see Supplemental Fig. 4B).
Numbers of circulating leukocytes
In the initial days of IL-10 therapy, a temporary increase in the total number of leukocytes [6.5±2.29 gigapartikel pro liter (Gpt/l) before therapy; 7.9±2.37 Gpt/l on Day 1; mean±SD; P=0.07; Wilcoxon test], the monocyte count (Table 3
), and the neutrophil count (4.3±1.53 Gpt/l before therapy; 5.6±1.69 Gpt/l on Day 1; mean±SD; P<0.05; Wilcoxon test) occurred. As the proportion of immature neutrophils (3.6±2.27% of leukocytes before therapy; mean±SD) did not increase, these phenomena most likely resulted from the release of adhered/sequestrated leukocytes into the circulation. The numbers of eosinophilic and basophilic granulocytes and of lymphocytes did not change significantly (0.23±0.110 Gpt/l; 0.05±0.053 Gpt/l; and 1.7±0.94 Gpt/l before therapy, respectively; mean±SD).
|
View this table: [in a new window] |
Table 3. Monocytes and Monocytic Subpopulations
|
RI CD64, which is induced by IL-10 in vitro [40
, 41
], increased significantly during IL-10 treatment (Table 3
; for group-specific data, see Supplemental Fig. 2, D and E). As IL-10 is known to inhibit the antigen-presenting capacity of monocytes in vitro [42 , 43 ], we investigated the monocytic expression of the MHC class II molecule HLA-DR and of the coaccessory molecules CD86, CD80, and CD40. Interestingly, each of the accessory molecules exhibited a distinct expression pattern. The monocytic expression of HLA-DR, an important parameter of cellular immunocompetence [44 ], was inhibited during IL-10 treatment and recovered to pretherapeutic values after IL-10 withdrawal (Fig. 3A ; for group-specific data, see Supplemental Fig. 5A). Similar to HLA-DR, CD86 expression decreased rapidly during IL-10 application (Fig. 3B ; for group-specific data, see Supplemental Fig. 5B). Remarkably and in contrast to short-term treatment [41 ], it remained low after withdrawal of IL-10 therapy (Fig. 3B) . A progressive decline was observed for CD80, a molecule expressed in much lower intensity compared with CD86, reaching the significance level after the end of IL-10 therapy (Fig. 3C ; for group-specific data, see Supplemental Fig. 5C). The monocytic expression of CD40 was increased significantly in the first weeks of IL-10 application and thereafter, gradually decreased. As for CD80 expression, a significantly diminished monocytic CD40 expression was found in the post-treatment observation period (Fig. 3D ; for group-specific data, see Supplemental Fig. 5D).
The secretion capacity for TNF-, IL-1β, IL-12 p70, and IL-1RA was determined in 4 h (TNF- and IL-1β) and 24 h (IL-12 p70 and IL-1RA) of whole blood endotoxin stimulation assays. As expected, IL-10 treatment strongly inhibited the production of the proinflammatory cytokines TNF-, IL-1β, and IL-12 p70 (Fig. 4 A-C
; for group-specific data, see Supplemental Fig. 6A-C). Interestingly, however, some re-increase was observed during the treatment period, indicating endogenous counter-regulatory mechanisms. Remarkably, the secretion capacity for IL-1RA, an anti-inflammatory cytokine antagonist, which is induced by IL-10 in vitro [45
], was also diminished significantly during long-term IL-10 therapy (Fig. 4D
; for group-specific data, see Supplemental Fig. 6D).
No other principal results were obtained after reviewing the data about monocyte or phagocyte (monocyte and neutrophil) numbers (data not shown).
Phenotypes and functions of circulating lymphocyte subpopulations
During the course of IL-10 therapy, no significant changes in the proportions of CD3+ lymphocytes (74.3±8.51% before therapy; mean±SD), CD4+ and CD8+ lymphocytes (48.1±7.79% and 32.6±4.02%, respectively; mean±SD before therapy), and the CD4/CD8 ratio (1.5±0.38; mean±SD) were observed. Moreover, as revealed by the determination of CD11a expression and CD45RA+ and CD45RO+ T cell subsets [46
, 47
], no significant changes were observed in the proportions of naive and effector/memory cells for the CD4+ and CD8+ T cell subsets (data not shown). The low proportions of T cells positive for the activation antigens CD71 (1.0±0.70% before therapy; mean±SD) and CD25 (3.9±1.86% before therapy; mean±SD) did not change significantly over the study course. In contrast, there was a slight drop of HLA-DR+ T cells during IL-10 treatment, which continued in the early post-therapeutic observation period (Fig. 5A
; for group-specific data, see Supplemental Fig. 7A). A more pronounced and lasting decrease without normalization in the 7-week post-treatment period was observed for CD16+ NK cells (Fig. 5B
; for group-specific data, see Supplemental Fig. 7B). In contrast, CD19+ B cells increased significantly from 11.0 ± 1.2% of lymphocytes before therapy to a maximum of 14.6 ± 1.3% at Day 84 (mean±SD; P<0.05; Wilcoxon test). This increase was mainly a result of a relative and absolute expansion of the CD5+ B cell subset after IL-10 therapy and remained enhanced until the end of the study (Fig. 5 C and D
; for group-specific data, see Supplemental Fig. 7, C and D).
The secretion of type 1 (IFN- and IL-2) and type 2 (IL-4) cytokines by lymphocytes was determined after 24 h Con A stimulation in whole blood. In contrast to short-term treatment [48
], Con A-triggered IFN- secretion was reduced by IL-10 treatment only during the initial days, and then recovered to pretherapeutic values (Fig. 6A
; for group-specific data, see Supplemental Fig. 8A). Interestingly, in contrast to in vitro data [49
], IL-10 treatment did not exhibit suppressive effects on ex vivo IL-2 secretion but led to its incremental enhancement, which persisted partly in the observation period (Fig. 6B
; for group-specific data, see Supplemental Fig. 8B). A strong, progressive increase was found for mitogen-triggered IL-4 secretion capacity during IL-10 therapy (Fig. 6C
; for group-specific data, see Supplemental Fig. 8C). Remarkably, there was a tendency for even higher values after IL-10 withdrawal (Fig. 6C)
.
For selective evaluation of the cytokine production capacity of T cells and NK cells, flow cytometric determination of intracellular cytokine formation was performed in T cells (IFN-, TNF-, IL-2, and IL-4) and NK cells (IFN- and TNF-) after 4 h stimulation with PMA and ionomycin. Whereas we did not find a significant influence of IL-10 therapy on intracellular cytokine production mediated by T cells (geometric mfi as well as percent-positive cells; data not shown), we found an increasing mfi for IFN- in NK cells during prolonged IL-10 application. Moreover, the highest IFN- values were observed in the late post-therapeutic observation period, i.e., 5–7 weeks after IL-10 withdrawal (Fig. 6D
; for group-specific data, see Supplemental Fig. 8D). For intracellular TNF- production by NK cells, no significant changes were observed (Fig. 6E
; for group-specific data, see Supplemental Fig. 8E). Analysis of NK cell data with respect to percent-positive cells did not reveal significant changes (data not shown).
DISCUSSION
Here, we report systemic immunological effects of prolonged IL-10 application observed in psoriatic patients during a period of 7 weeks IL-10 treatment followed by a 7-week post-treatment observation period [22 ]. IL-10 treatment was well-tolerated, and antipsoriatic effects were found in nine out of 10 patients, resulting in a significant decrease of the PASI score [22 ]. Clinical unresponsiveness seen in one of our patients was not a result of a general IL-10 unresponsiveness, as the patient did respond with respect to IL-10-induced systemic immune alterations (data not shown separately). However, the pre-existence or development of neutralizing anti-IL-10 antibodies has to be kept in mind when treating patients with rhIL-10 [50 , 51 ].
Prolonged IL-10 treatment induced partly unexpected immunological effects and demonstrated differences between in vitro and in vivo as well as short-term versus long-term findings, which will be discussed in detail. Although the complex immune alterations might widely reflect general regulatory principles for interference with the cytokine network, it has to be kept in mind that IL-10 was applied to patients with an active autoimmune disease, which might have modulated the response profile.
IL-10-induced activation of unspecific immune defense
It is now well-established that IL-10, in addition to its anti-inflammatory and immunosuppressive activities, exerts activating effects on the unspecific defense system. So, IL-10 has been shown to trigger the production of acute-phase proteins by hepatocytic cells [11
]. Accordingly, in our study, we observed enhanced plasma levels of IL-6, CRP, and LBP as indicators for an acute-phase response. Moreover, strong increases in the 1- and 2-h blood sedimentation rates were observed during IL-10 application. Even more surprisingly, IL-10 also significantly increased the plasma levels of PCT, a parameter that usually indicates (systemic) bacterial/mycotic infections but may also reflect other causes of endotoxinemia, e.g., endotoxin translocation [52
, 53
]. According to the manufacturer’s information, the rhIL-10 preparation used was highly purified and sterile and did not induce toxicity (neither clinically nor laboratory-confirmed) in mice treated with a dose 500-fold higher than in our daily treated patients, indicating that no relevant contamination with LPS or other inflammatory stimuli could be responsible for these observations [28
]. Indeed, LAL testing of rhIL-10 revealed that the theoretical maximum LPS exposition per application would be 0.008 IE/kg body weight, which is far below the concentrations described for the induction of an inflammatory response in humans [29
, 30
].
Another possible cause for the observed acute-phase response might have been the repeated local traumata at the injection sites. However, by measuring sIL-2R in psoriasis patients undergoing a placebo-controlled trial with s.c. IL-10 application, we did not observe activation of unspecific immune defense in the placebo group [54 ].
Recently, the so-called monocyte-deactivating cytokine IL-10 has been demonstrated to stimulate monocytic phagocytosis and to support the development of monocytes to macrophages [7
, 36
, 55
56
57
58
]. In our study, we observed phenotypic alterations, indicating a maturation of monocytes to circulating premacrophages during prolonged IL-10 application. So, an increased monocytic expression of CD71 and CD64 was found. Moreover, enhanced plasma concentrations of neopterin, which is involved in antimicrobial defense of monocytes [59
, 60
], indicated monocyte activation. An unexpected finding was the small but significant IL-10-induced increase in TNF- plasma concentrations, which was also seen in leukemia patients treated with rhIL-10 [61
]. In addition to monocyte activation, this may have been caused by an increase in soluble TNF- receptors, as already shown in vitro [62
, 63
] and in vivo [64
], and indeed, we could demonstrate enhanced plasma levels of sTNF-RI and sTNF-RII during IL-10 therapy in three out of three patients investigated (data not shown).
Lastly, despite falling numbers of circulating NK cells, a selective activation of this lymphocyte subpopulation was demonstrated by increasing levels of intracellular IFN- accumulation after PMA/ionomycin stimulation. The stimulatory activity of IL-10 on NK cells is well-established in the literature [12
, 65
] and may account for beneficial effects of IL-10 in tumor models [5
] and adverse effects of high-dose IL-10 application [17
].
The parameter showing strongest correlation to the IL-10 application regime was the sIL-2R plasma level. Significantly higher concentrations in the daily application group with more consistently elevated IL-10 levels indicated that it may be induced directly by IL-10. As in vitro treatment of whole blood with rhIL-10 did not lead to sIL-2R release into the supernatants (data not shown), other sources, e.g., endothelial cells, Kupffer cells, or hepatocytes, should be considered.
IL-10-induced modulation of the proinflammatory and type 1 immune response
As expected, IL-10 application led to a rapid suppression of the proinflammatory and type 1 cytokine response (suppression of LPS-triggered secretion of TNF-, IL-1β, and IL-12 p70 and of mitogen-induced lymphocytic IFN- production). Interestingly, however, this inhibition was not consistent but showed fluctuations, as in the case of the LPS-induced cytokines, or even disappeared completely after a few days, as seen in IFN- production. Besides diagnostic blood draws as late as 72 h following the last IL-10 application in the 3x/week application group [15
], counter-regulatory mechanisms, which may include the IL-10-induced activation/differentiation of monocytes to circulating premacrophages, and the increased production of IFN- by NK cells (see above) may be responsible for this observation. As it is also discussed that IFN- has immunoregulatory properties [65
], the observed pattern of IFN- production by NK cells might even contribute to the beneficial effects of IL-10 seen in our psoriasis patients [22
]. The nevertheless potent, ongoing, anti-inflammatory action of IL-10 is well-illustrated by the rebound enhancement of plasma levels of the common subunit p40 shared by IL-12 and IL-23 and TNF- plasma levels after the cessation of IL-10 therapy.
An interesting finding was the successive increase of Con A-induced IL-2 production, which persisted partly even after the end of IL-10 therapy. IL-2 is considered to be an important T cell growth factor and type 1 cytokine. Data from IL-2 knockout mice, which show an expansion of activated T cells and the development of spontaneous autoimmune arthritis, in addition to recent publications about the essential role of IL-2 for the induction of regulatory T cells and for the Fas/Fas ligand-mediated activation-induced cell death of T cells, however, also demonstrated an important immunoregulatory role of IL-2 [66
]. Concluding from this, the observed increase in mitogen-induced IL-2 production may contribute to the partial therapeutic efficacy of IL-10 therapy in the autoimmune disease psoriasis similar to enhanced IFN- production by NK cells (see above).
Evidence for a lasting shift to a type 2 immune response and lasting diminished immunocompetence
We observed several hints of a strengthened type 2 cytokine response during and beyond IL-10 application. So, significant increases were demonstrated for the numbers of CD5+ B cells, which are known to depend on IL-10 and to expand in type 2 autoimmune diseases such as Sjögren’s syndrome [67
], and mitogen-triggered IL-4 production. Furthermore, during the IL-10 application period, IgE serum levels increased, a result we had seen already in another study [25
]. In contrast to IgE levels, which declined after cessation of IL-10 therapy, the expansion of CD5+ B cells and the increased lymphocytic IL-4 production capacity persisted until the end of the 7-week observation period, indicating a lasting modulation of the type 1/type 2 balance in favor of a type 2 immune response. This may represent an important re-regulation of the immune system after prolonged IL-10 therapy. Compatible with this hypothesis are our findings in psoriatic patients treated with IL-10 for relapse prevention: A significant, decreased relapse rate could be seen in parallel with a type 2 shift of the immune system [54
], indicating that IL-10 application may be successful in preventing rather than in reverting an already-established inflammatory immune response. Taking into consideration the short plasma half-life of 2–3 days for IgE, the fast increase and normalization of IgE serum levels may have been caused directly by changes in IgE production. Other mechanisms, however, such as IL-10 induced down-regulation of high-affinity IgE receptor expression on basophils, and mast cells, as described recently for mouse mast cells [68
], may also have contributed to these changes.
A lasting, altered immunoregulation was also indicated by changes in the monocytic coaccessory molecule expression. Whereas HLA-DR expression on monocytes was reconstituted completely after the cessation of IL-10 therapy, the expression of the coaccessory molecules CD86, CD80, and CD40 showed a significant reduction at the end of the 7-week observation period. This lasting reduction, especially for CD86, likely reflects a decreased competence of the cellular immune system. Sabat and co-workers [41 ] demonstrated recently that in vitro long-term IL-10 treatment of peripheral blood monocytes from healthy volunteers induced the same phenotypic changes that are reconstituted HLA-DR expression but persistently diminished CD86 expression. Indeed, these long-term, IL-10-treated monocytes showed a diminished capacity to activate an allogeneic T cell response [41 ].
When discussing possible mechanisms of this immune deviation/re-regulation induced by long-term IL-10 therapy, an enhanced use of TGF-β1 seems most likely to be the cause for at least some of the effects. A significant decrease of TGF-β1 plasma concentrations became evident already 6 h after the first IL-10 application and most importantly, persisted until the end of the observation period. Interestingly, recent unpublished observations in our group indicate a strong induction of TGF-βRII expression on monocytes by IL-10, which can induce a drop of TGF-β1 levels in PBMC culture supernatants (own unpublished data). Strengthened TGF-β effects during and after IL-10 therapy are indicated by the lasting decrease in the endocrine immunocompetence parameter DHEA-S, whose adrenal and nonadrenal production is decreased by TGF-β [69 ]; the persistent increase of IgA plasma levels, for which TGF-β is a main inducer for class switching [70 ]; and the drop of CD14 and coaccessory molecule expression on monocytes [36 , 71 , 72 ].
In summary, by comprehensive biomarker monitoring, we found manifold effects of long-term IL-10 therapy, which were far beyond the initially considered, immediate anti-inflammatory activity. The complex, time-dependent, and even largely prolonged immune effects demonstrated, and the differences seen between in vivo and described in vitro as well as long-term versus short-term effects underline the importance of careful biomarker selection, especially in interventional studies; may explain the rather disappointing, clinical effects of IL-10 therapy in exacerbated inflammation; but also, may hint to its role for a sustained immunological reshaping.
ACKNOWLEDGEMENTS
We are very thankful to Schering Plough Research Institute/Essex Pharma (Munich, Germany), which kindly provided the rhIL-10 protein. We thank Elizabeth Wallace for accurately proofreading the manuscript.
FOOTNOTES
1 Current address: Bayer Schering Pharma AG, Target Discovery, D-13342 Berlin, Germany. 
3 Current address: Intendis GmbH, D-10589 Berlin, Germany.
Received April 18, 2008; revised September 26, 2008; accepted September 28, 2008.
REFERENCES
R surface expression and cytotoxic activity. Distinct regulation of antibody-dependent cellular cytotoxicity by IFN-, IL-4, and IL-10 J. Immunol. 149,4048-4052[Abstract]-Hydroxy-dehydroepiandrosterone and immune response Ann. N. Y. Acad. Sci. 917,971-982[Medline] RI/CD64) gene and STAT protein binding to the IFN- response region (GRR) are regulated differentially in human neutrophils and monocytes by IL-10 J. Immunol. 160,911-919 and interleukin-10 have cross-regulatory roles in modulating the class I and class II MHC-mediated presentation of epitopes of Listeria monocytogenes by infected macrophages J. Interferon Cytokine Res. 16,547-554[Medline] or IL-10 J. Immunol. 151,6370-6381[Abstract], VLA-β 1, CD45RO and CD28 on T-cell subsets: evidence for multiple subsets of "memory" T cells Int. Arch. Allergy Immunol. 104,17-26[Medline]-production by suppressing natural killer cell stimulatory factor/IL-12 synthesis in accessory cells J. Exp. Med. 178,1041-1048 J. Leukoc. Biol. 58,351-358[Abstract] have distinct, but interacting, effects on differentiation-induced changes in TNF- and TNF receptor release by cultured human monocytes Cytokine 8,49-57[CrossRef][Medline] in immune cell regulation J. Leukoc. Biol. 58,373-381[Abstract]
RI expression in mouse mast cells J. Immunol. 172,3181-3188
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
