Tristetraprolin regulation of interleukin-22 production

Interleukin (IL)-22 is a STAT3-activating cytokine displaying characteristic AU-rich elements (ARE) in the 3′-untranslated region (3′-UTR) of its mRNA. This architecture suggests gene regulation by modulation of mRNA stability. Since related cytokines undergo post-transcriptional regulation by ARE-binding tristetraprolin (TTP), the role of this destabilizing protein in IL-22 production was investigated. Herein, we demonstrate that TTP-deficient mice display augmented serum IL-22. Likewise, IL-22 mRNA was enhanced in TTP-deficient splenocytes and isolated primary T cells. A pivotal role for TTP is underscored by an extended IL-22 mRNA half-life detectable in TTP-deficient T cells. Luciferase-reporter assays performed in human Jurkat T cells proved the destabilizing potential of the human IL-22-3′-UTR. Furthermore, overexpression of TTP in HEK293 cells substantially decreased luciferase activity directed by the IL-22-3′-UTR. Transcript destabilization by TTP was nullified upon cellular activation by TPA/A23187, an effect dependent on MEK1/2 activity. Accordingly, IL-22 mRNA half-life as determined in TPA/A23187-stimulated Jurkat T cells decreased under the influence of the MEK1/2 inhibitor U0126. Altogether, data indicate that TTP directly controls IL-22 production, a process counteracted by MEK1/2. The TTP-dependent regulatory pathway described herein likely contributes to the role of IL-22 in inflammation and cancer and may evolve as novel target for pharmacological IL-22 modulation.

Interleukin (IL)-22 1,2 is a member of the IL-10 cytokine family sharing some fundamental structural and biological properties with IL-10, IL-20, IL-24, and IL-6. Biochemically, this is exemplified by the shared ability of aforementioned cytokines to mediate robust activation of the transcription factor signal transducer and activator of transcription (STAT)-3 and associated STAT3-dependent downstream events connecting to proliferation, anti-apoptosis, strengthening of host-defense, and regulation of inflammatory responses. A particularly striking feature of IL-22 is that this cytokine specifically targets epithelial (-like) cells, among others keratinocytes and hepatocytes as well as lung and intestinal epithelial cells. Restricted expression of the decisive IL-22 receptor chain IL-22R1 on aforementioned cell types is regarded the biological basis for this selectivity [3][4][5][6] . Leukocytic cells generally neither express IL-22R1 nor respond to IL-22. Yet, IL-22 is largely a lymphocyte-derived cytokine being efficiently produced by natural killer and related innate lymphoid cells, by invariant NK-T and γ δ T cells and a broad array of adaptive CD4 + or CD8 + T cells, the former including differentiated Th1, Th17, and Th22 subsets 4,[7][8][9][10][11] .
The role of IL-22 in disease is truly context dependent. IL-22 exerts tissue-protective/anti-microbial functions in infection-and/or injury-driven diseases at biological barriers such as intestine, lung, and liver 5 . Examples of pathological conditions with IL-22 displaying protective properties include intestinal infection by Citrobacter rodentium 12 , colitis induced by dextran sulfate sodium 13 , concanavalin A 14 -or acetaminophen-induced acute liver injury 15 as well as ventilator-induced lung injury 16 and mucosal candidiasis 17 . However, Janus-faced IL-22 also shows the potential to aggravate some aspects of pathological inflammation. Specifically, IL-22 promotes disease in experimental psoriasis 18 and arthritis 19 . Besides induction of inflammatory chemokines, a major mode of IL-22 disease-promoting functions in this context is its pro-proliferative and anti-apoptotic action targeting keratinocytes and synoviocytes, respectively. It is obvious that those two key properties likewise relate to the unfortunate role of IL-22 in cancer 20 .
Given the multilayered biological functions of IL-22, knowledge of molecular mechanisms driving its production is crucial. Previous reports indicate that transcription factors/nuclear receptors such as STAT3, retinoid orphan receptor-γ t and aryl hydrocarbon receptor 21 as well as the cAMP response element-binding protein and nuclear factor of activated T cells (NF-AT) 22 are involved in initiation of IL-22 gene transcription. However, firm knowledge on post-transcriptional molecular mechanisms regulating IL-22 expression is lacking. Sequence analysis reveals a remarkable density of adenylate-and uridylate (AU)-rich elements (ARE) in the 3′ -untranslated region (3′ -UTR) of human and murine IL-22 mRNA (Fig. 1). The presence of those elements at this location suggests post-transcriptional regulation by modulation of mRNA stability [23][24][25] . The CCCH zinc finger protein tristetraprolin (TTP) 25 has been identified as crucial trans-acting factor binding to ARE within the 3′ -untranslated region of labile mRNA molecules. Subsequent to target binding, TTP is supposed to destabilize mRNA molecules by enforcing the processes of mRNA deadenylation and decapping thereby augmenting decay by exonucleases 24,25 . Cytokines are a prime target of TTP biological activity. A well characterized example is tumor necrosis factor (TNF)-α , the expression of which is most efficiently controlled by TTP as part of a negative feedback loop aiming at control of exacerbated inflammation and/or to initiate its resolution 26,27 . Since mRNA expression of IL-22-related IL-10 25,28 and IL-6 25,29 is known to be modulated by TTP and post-transcriptional gene regulation is frequently organized in functional units 30 , we set out to investigate in detail the role of TTP in IL-22 expression.

Results
Enhanced production of IL-22 detected in TTP-deficient mice and ex vivo stimulated TTP −/− splenocytes. TTP −/− mice display a characteristic inflammatory syndrome with erosive arthritis, conjunctivitis, dermatitis, and cachexia as obvious severe symptoms 31 . In accord with the picture of uncontrolled persistent inflammation, we report for the first time on significantly elevated systemic levels of IL-22 in TTP −/− mice as compared to wildtype littermates (Fig. 2a). Likewise, serum levels of the IL-22-related and TTP-regulated 25,28,29 cytokines IL-6 ( Fig. 2b) and IL-10 ( Fig. 2c) were increased. Data are in accord with previously reported IL-22 mRNA upregulation as detected in skin and draining lymph nodes of TTP −/− mice 32 .
In order to further investigate on a cellular level IL-22 production in the context of TTP deficiency, cytokine production by ex vivo stimulated splenocytes was assessed. For that purpose, cytokine-(exposure to IL-12/IL-18) and T cell receptor (TCR)-(exposure to α CD3/α CD28) stimulated IL-22 release was evaluated in splenocytes isolated from TTP −/− mice and respective wildtype littermates. Of note, IL-18, particularly in combination with IL-12, is a most potent mediator of cytokine-based T cell activation 33 . Here we demonstrate that IL-12/IL-18- (Fig. 3a, left panel) and α CD3/α CD28- (Fig. 3b) mediated IL-22 production was potentiated in splenocytes derived from TTP −/− mice. Likewise, production of IL-6 and IL-10, determined in splenocytes exposed to IL-12/IL-18, was markedly increased in TTP −/− mice   TTP deficiency associates with augmented IL-22 mRNA half-life as detected in primary murine CD3 + T cells. In order to more directly relate TTP expression with IL-22 mRNA stability in T cells, actinomycin D experiments were performed using isolated splenic CD3 + T cells from TTP −/− mice or wildtype littermates, respectively. IL-22 mRNA induction was achieved by activating T cells with α CD3/α CD28. Notably, a 4 h incubation period was sufficient to mediate robust induction of IL-22 mRNA under those experimental conditions. In accord with aforementioned observations, T cells obtained from TTP −/− mice displayed significantly enhanced IL-22 mRNA expression as compared to wildtype littermates (Fig. 4a). Actinomycin D experiments performed subsequent to this 4 h gene induction period revealed an IL-22 mRNA half-life of approximately 29 minutes that was increased upon TTP deficiency (Fig. 4b). Data altogether demonstrate a role for TTP in the regulation of IL-22 mRNA stability.
The destabilizing potential of the IL-22-3′-UTR. Luciferase-reporter assays were performed in Jurkat T and HEK293 cells to investigate mechanisms regulating IL-22 mRNA stability in detail. Since murine and human IL-22-3′ -UTR ARE sequences display extensive homology (Fig. 1) and aforementioned cells lines are of human origin, further experiments were performed using human IL-22-3′ -UTR sequences. For that purpose, luciferase-reporter constructs were transfected with variants of the human IL-22-3′ -UTR cloned next to a luciferase-reporter gene (Fig. 5a). Since luciferase enzyme activity can be readily determined, this experimental approach allows to straightforwardly evaluate the potential for post-transcriptional regulation by ARE originally located in the IL-22-3′ -UTR. Using human Jurkat T cells, we demonstrate that the full length IL-22-3′ -UTR (transfection of wt_UTR_IL22) in fact displays mRNA destabilizing characteristics. Transfection of ARE37_IL22, containing a 206 nt ARE-rich deletion fragment representing the distal part of the IL-22-3′ -UTR, likewise reduced luciferase-reporter activity. This capability was lacking in case of transfection with ARE_del_IL22, containing a deleted fragment of the IL-22-3′ -UTR without ARE (Fig. 5b). Interestingly, concomitant activation of Jurkat T cells  (e) HEK293 cells were transfected for 16 h with indicated luciferase reporter plasmids together with either a TTP-expression (pZEO_hTTP)-or a control-plasmid (ctrl). Luciferase activity is depicted (as % of cells transfected with the same luciferase reporter plasmid plus ctrl-plasmid) and expressed as means ± SD (n = 4; *p = 0.02 for wt_UTR and ctrl or TTP, *p = 0.028 for ARE37 and ctrl or TTP). Statistical analysis on raw data, Student's t-test. (f) TTP overexpression was confirmed by immunoblot analysis of lysates obtained from (e). One representative of four independently performed experiments is shown. by 12-O-tetradecanoylphorbol-13-acetate (TPA)/A23187 (calcium ionophore) nullified the inhibitory action of the IL-22-3′ -UTR (Fig. 5c). In accord with a previous report 34 , constitutive expression of TTP was detectable by immunoblot analysis in Jurkat T cells. Levels of TTP in these cells were not further increased under the influence of TPA/A23187 (Fig. 5d). Data thus suggest that endogenously expressed TTP present in Jurkat T cells contributes to post-transcriptional gene regulation achieved by the IL-22-3′ -UTR. SV40-driven target gene overexpression in HEK293 cells was employed to further address the role of TTP in gene regulation via the IL-22-3′ -UTR. As shown in Fig. 5e, overexpression of human TTP reduced luciferase activity (when achieved through a plasmid containing ARE derived from the IL-22-3′ -UTR -transfection with wt_UTR_IL22 or ARE37_IL22). In contrast, overexpression of TTP (see Fig. 5f) in combination with ARE_del_IL22 or with a luciferase expression plasmid entirely lacking IL-22-3′ -UTR (and thus ARE) sequences did not inhibit but tended to increase luciferase reporter activity. This observation excludes suppressive effects of TTP overexpression acting on the level of luciferase enzyme transcription or activity.
Finally, in vitro binding assays were performed that demonstrate physical binding of TTP to an RNA sequence derived from the IL-22-3′ -UTR but not to a mutated counterpart (Fig. 6). This RNA oligonucleotide was specifically selected and spans the region of human ARE5/6. Notably, the whole IL-22-3′ -UTR sequence covered by this RNA oligonucleotide (45 nt) is conserved between mice and humans displaying 93.3% identity. Data altogether indicate that TTP is able to regulate reporter gene expression by interacting with adjacent IL-22-3′ -UTR sequences and thus by destabilizing target mRNA.  35 . In order to extend those previous data, TPA/A23187-activated Jurkat T cells were coincubated with a panel of pharmacological inhibitors affecting the mitogen-activated protein kinase (MAPK)/extracellular-signal-regulated kinases (ERK) signaling pathway. As shown in Fig. 7a, PD98059 (targeting MEK1/2 albeit with less potency compared to U0126) and FR180204 (targeting ERK1/2) as well as SB203580 (targeting p38 MAPK) but not SP600125 (targeting c-jun N-terminal kinases) significantly inhibited IL-22 mRNA expression. Those experiments were performed under conditions not affecting cell viability (Fig. 7b). Alike IL-22, also expression of related IL-6 was potently inhibited by U0126 in TPA/A23187-stimulated Jurkat T cells (Fig. 7c). Since U0126 was, by far, the most effective inhibitor of IL-22 expression (Fig. 7a and ref. 22) and TPA/A23187 potently activated the MEK/ERK Figure 6. Binding of TTP to ARE located in the IL-22-3′-UTR as detected in vitro by RNA-EMSA. In vitro translated TTP was incubated together with a 32 P-γ -ATP-labelled RNA oligonucleotide probe that includes the ARE5/6 region of the human IL-22-3′ -UTR (see Fig. 1). In addition to this 'wildtype' (wt) oligonucleotide, a 'mutated' (mut) oligonucleotide was used lacking regular ARE sequences (see methods section). Reaction mixtures were subjected to native polyacrylamide gel electrophoresis. The brace indicates positions of retarded complexes present when the 'wildtype' but absent when the 'mutated' oligonucleotide was used. This retarded signal allegedly represents RNA/TTP complexes. Ctrl, denotes a control-in vitro translation setup with expression of an unrelated protein (firefly-luciferase) serving as control for unspecific protein/RNA interactions. Inset: Immunoblot analysis of in vitro translated TTP. One representative of three independently performed experiments is shown.  pathway in Jurkat T cells (Fig. 7d), we chose to focus on this inhibitor in subsequent experiments. Notably, U0126 likewise suppressed IL-22 mRNA induction (Fig. 7e) and protein release (Fig. 7f) as well as IL-10 mRNA induction (Fig. 7g) by human α CD3-stimulated primary T cells.
Although MAPK signaling is mandatory for maximal cellular activation by NF-AT 36 and thus likely involved in IL-22 promoter activation 22 , effects of U0126 on IL-22 mRNA stability were assessed in Jurkat T cells after a 4 h induction period using TPA/A23187. Figure 8a confirms robust IL-22 mRNA induction by TPA/A23187-stimulated Jurkat T cells 22 . As determined by actinomycin D-mediated transcriptional blockage performed in those same experiments, Jurkat T cells displayed significantly diminished IL-22 mRNA stability under the influence of U0126 (Fig. 8b). Those experiments were performed under conditions not affecting cell viability (Fig. 8c).
Since TTP is a known target of the MEK/ERK pathway 25 , TPA/A23187 activates ERK1/2 in Jurkat T cells (Fig. 7d) as well as HEK293 cells (Fig. 9a), and the MEK1/2 inhibitor U0126 reduced IL-22 mRNA half-life in TPA/A23187-stimulated Jurkat T cells (Fig. 8b), effects of U0126 were investigated in the context of luciferase-reporter activity under the control of the IL-22-3′ -UTR. For that purpose, HEK293 cells were transfected with wt_UTR_IL22 alone or in combination with the TTP expression plasmid. Cells were additionally stimulated with TPA/A23187 alone or in combination with U0126. As already shown (Fig. 5e), TTP overexpression reduced IL-22-3′ -UTR-directed luciferase-reporter activity (Fig. 9b). Suppression by TTP was, in accord with data on Jurkat T cells (Fig. 5c), reversed under the influence of TPA/A23187 (Fig. 9b). Since TPA/A23187 did not significantly affect luciferase reporter activity in the absence of IL-22-3′ -UTR (and thus ARE) sequences (111.7 ± 12.1% for TPA/A23187 versus unstimulated HEK293 cells with luciferase activity of unstimulated cells set as 100%, n = 3), stimulatory effects of TPA/A23187 acting on the level of luciferase enzyme transcription or activity can be excluded in these experiments. Notably, this TPA/A23187 effect was nullified by coincubation with U0126 (Fig. 9b). Data altogether suggest that the inhibitory action of TTP on IL-22-3′ -UTR directed luciferase gene expression is counteracted by TPA/A23187-stimulated MEK/ERK signaling.  37 . These cytokines display tissue protective properties 15,38,39 that, however, come along with the potential to drive tumor growth 20,40,41 . This connects to immunosuppressive 37,41 and oncogenic functions 42 of STAT3 in leukocytes and cancerous cells, respectively. In addition, whereas functional IL-10 receptors lack on hepatocytes 43 , IL-6 and IL-22 are pivotal mediators of the STAT3-driven acute phase response 44 . Here, we characterize regulation by TTP as further IL-22 characteristic being shared with IL-6 25,29 and IL-10 25,28 .
To specify the role of TTP for IL-22 expression, TTP −/− mice that suffer from life-shortening harsh inflammation 31 were investigated. We detected increased serum IL-22 in TTP −/− mice as compared to wt littermates which was paralleled by a bias observed in TTP-deficient cultured splenocytes and isolated splenic CD3 + T cells to express elevated levels of IL-22. Data demonstrate the capability of TTP to modulate IL-22 expression in primary murine T cells. Notably, this conclusion is at variance with a previous report not observing upregulation of IL-22 secretion in TTP-deficient isolated T cells 32 . However, fundamental different protocols were used. The latter study used naïve CD4 + T cells differentiated in presence of α CD3/α CD28 towards T0, Th1, Th17, and Th22 using 3-day-incubation-protocols. Thereafter, IL-22 secretion was detectable only in Th22 cells and did not differ between genotypes 32 . Herein, we specifically decided to use whole splenic CD3 + T cells that include memory T cell subsets prone to efficiently express IL-22. After only 4 h of stimulation by α CD3/α CD28, brisk upregulation of IL-22 mRNA was detected. This brief incubation period precludes indirect effects of extended stimulation protocols and enables efficient determination of IL-22 mRNA half-life. In fact, splenic TTP −/− CD3 + T cells displayed significantly prolonged IL-22 mRNA half-life in the context of α CD3/α CD28 stimulation.
Using Jurkat T and HEK293 cells, luciferase reporter assays were performed in order to deepen knowledge on the relevance of the IL-22-3′ -UTR for IL-22 expression. Experiments revealed a strong ARE-dependent mRNA destabilizing potential of the IL-22-3′ -UTR that was nullified in response to TPA/A23187. Notably, overexpression of TTP inhibited luciferase activity under the control of the IL-22-3′ -UTR. In vitro assays furthermore demonstrated physical binding of TTP to conserved ARE within the human IL-22-3′ -UTR. Altogether, data indicate that TTP directly regulates IL-22 in activated T cells. Notably, functional TTP is constitutively expressed in Jurkat T cells 34 and, after polyclonal activation, rapidly induced and biological active in primary human T cells 45,46 .
Previously, we observed in TPA/A23187-stimulated Jurkat T cells suppression of IL-22 mRNA by the MEK1/2 inhibitor U0126 22 . To expand this observation, a panel of MAP kinase inhibitors was evaluated herein. With the exception of SP600125, all compounds reduced IL-22 mRNA. Among those inhibitors, which included the p38 MAP kinase inhibitor SB203580, U0126 was the most potent. The specificity of this U0126 action is emphasized by the notion that PD98059, also targeting MEK1/2, and FR180204, targeting the MEK1/2 downstream kinases ERK1/2 47 , were as well capable of modulating IL-22 mRNA in Jurkat T cells. U0126 likewise suppressed production of IL-22 by primary human CD3 + T cells, along with that of related IL-10. Observations correspond to a recent report demonstrating inhibition of IL-22 secretion by PD98059 as detected in Th17 cells generated from naïve T cells under the influence of IL-1β , IL-6, and IL-23 as well as anti-IFNγ and anti-IL-4. However, the capacity of MEK1/2 inhibition to directly impair IL-22 gene expression was not assessed 48 .
Activation of the MEK/ERK axis, particularly in cooperation with p38 MAP kinase, has the capability to target and inhibit TTP function 25,[49][50][51] . Modulation of TTP biological function by phosphorylation is supposed to be counteracted by the phosphatase PP2A 25 . Actually, we observed significant inhibition of IL-22 mRNA half-life by U0126 as detected in TPA/A23187-stimulated Jurkat T cells. Notably, ERK1/2 (Fig. 7d) as well as p38 MAP kinase 52 are being activated in Jurkat T cells under those experimental conditions. A regulatory interplay between the MEK/ERK axis and TTP was substantiated by performing luciferase reporter assays in TTP-overexpressing HEK293 cells. Under those conditions, U0126 reversed transcript stabilization achieved by TPA/A23187. Data demonstrate that the MEK/ERK pathway has the capability to antagonize TTP functions that aim at destabilizing IL-22 mRNA. In light of a broader context, promoting IL-22 expression feeds into the role of the MEK/ERK axis to support induction of genes associated with immunoactivation and inflammation 47 .
Altogether, this is the first report indicating that TTP by interaction with the IL-22-3′ -UTR directly regulates IL-22 gene expression in a T cell autonomous fashion. Although not addressed herein, it tempting to speculate that IL-22 regulation by TTP is part of an interdependent regulatory network enforcing posttranscriptional regulation by diverse mechanisms comprising of further ARE-binding proteins such as human antigen R (HuR), KH-type splicing regulatory protein (KSRP), and AU-binding factor 1 (AUF1) as well as microRNA populations 23,24 . This layer of gene regulation may open the avenue towards novel pharmacological approaches. In that context it is noteworthy that pharmacological regulation of TTP availability by using an oligonucleotide targeting insulin receptor substrate-1 associates with modulation of endothelial cell steady-state levels of a set of ARE-containing mRNA molecules coding for TNFα , vascular endothelial growth factor, (VEGF), IL-1β , IL-8, IL-12, and IL-22 53 . However, the biochemical basis of this association, specifically the relationship between TTP and IL-22 mRNA stability as well as interactions between TTP and the IL-22-3′ -UTR, was not addressed in that report.
Posttranscriptional regulation as detected herein is frequently organized in functional units 30 . In case of TTP, control of pro-inflammatory genes prevails which is clearly documented in vivo by overwhelming TNFα -and IL-23-dependent inflammation in TTP −/− mice 31,32 . Beyond that, it has become evident that TTP is able to inhibit in quasi coordinated manner a set of cytokines and growth factors that are supposed to promote diverse aspects of carcinogenesis 25,54 . Those include obvious candidates such as vascular endothelial growth factor 55 , IL-1 56 , IL-8 57 , IL-6 40 , IL-10 41 , and IL-22 20 . Interestingly, expression of TTP is characteristically low in cancerous tissue 54 . Data presented herein indicate that therapeutic strategies aiming at upregulation of TTP biological activity may, among others, oppose the ill-fated role of IL-22 in colon, liver, lung, and skin carcinogenesis 20,58-60 .
Animals. All mice were housed in accordance with standard animal care requirements and maintained under specified pathogen-free conditions on a 12/12-h light/dark circle. Throughout the study [14][15][16] week-old mice were used. Water and food were given ad libitum. TTP +/− mice (a kind gift by Dr. Blackshear 27 , NIEHS, National Institutes of Health, Research Triangle Park, NC, USA) had a C57BL/6 background. TTP −/− and TTP +/+ mice were obtained by mating TTP +/− animals. Genotyping of mice was performed by polymerase chain reaction (PCR), using primers that span the regions of the wild type genes disrupted by the targeting vectors. The following oligonucleotides (Sigma-Aldrich) were used for genotyping the Ttp locus: TTP-wt/ko-for, 5′ -GAGGGCCGAAGCTG CGGTGGGT-3′ ; TTP-wt-rev, 5′ -GGCTGGCCAGGGAGAGCTAGGTC-3′ ; and TTP-ko-rev, 5′ -CTGTTGTGCCCAGTCAT AGCCG-3′ . Animal studies were performed in accordance with German animal protection law. were maintained in DMEM supplemented with 100 units/ml penicillin, 100 μ g/ml streptomycin, and 10% heat-inactivated FCS (Life Technologies). For experiments, HEK293 cells were seeded on 6-well polystyrene plates (Greiner) in the aforementioned culture medium. Human CD3 + T cells -For isolation of peripheral blood mononuclear cells (PBMC), written informed consent was obtained from healthy donors, and blood was taken. All experimental protocols were approved by the 'Ethik Kommission' of the University Hospital Goethe-University Frankfurt. The methods were carried out in accordance with the approved guidelines. Healthy donors had abstained from taking drugs/medication for 2 weeks before the study. PBMC were isolated from peripheral blood using Histopaque-1077 (Sigma-Aldrich) according to the manufacturer's instructions. The untouched CD3 + T cell population of PBMC was isolated using the Pan-T-cell isolation kit according to the manufacturer's instructions (Miltenyi, Bergisch Gladbach, Germany). Cells were resuspended in RPMI 1640 supplemented with 10 mM HEPES, 100 units/ml penicillin, 100 μ g/ml streptomycin, and 1% human serum (Life Technologies) and seeded at 3 × 10 6 cells/ml in round-bottom polypropylene tubes. To assess successful isolation, FACS analysis (FACS Canto, BD Biosciences, Heidelberg, Germany) was performed with the following antibody: mouse monoclonal anti-human CD3-PerCP/Cy5.5 (BioLegend). CD3 + T cell isolation resulted in a mean purity of 98.2 ± 1.7% (n = 6). Murine splenocytes and splenic CD3 + T cells -Isolation of splenocytes from TTP −/− mice and wt littermates as well as further isolation of untouched CD3 + T cells (Pan-T-cell isolation kit, Miltenyi) was performed according to the manufacturer's instructions. For experiments, 5 × 10 6 CD3 + T cells or splenocytes were seeded on 12-well polystyrene plates in RPMI 1640 culture medium supplemented with 10% FCS, 100 units/ml penicillin, and 100 μ g/ml streptomycin. CD3 + cell isolation was evaluated by FACS analysis (FACS Canto) using hamster monoclonal anti-mouse CD3-PerCP/Cy5.5 (BioLegend). Purity of CD3 + cells was 97.9 ± 1.5% (n = 30).

Cloning of the human IL-22-3′-UTR, transient transfection of Jurkat T cells and HEK293 cells, and luciferase reporter assays.
To generate luciferase reporter constructs, we amplified 3′ flanking regions of the IL-22 mRNA (NM_020525) from cDNA generated from Jurkat T cell mRNA, using Phusion polymerase (Thermo Scientific, Waltham, USA). The following primers (excluding an additional NotI cloning site) were used: wt_UTR_IL22 (553 bp), ARE_del_IL22 (206 bp): forward 5′ -CCAGAGCAAAGCTGAAAAATG-3′ ; ARE37_IL22 (206 bp): forward 5′ -GTTTCCATAATCAGT ACTTTATATTTATAA-3′ . The reverse primers (excluding an additional flanking XhoI cloning/restriction site) were: wt_UTR_IL22, ARE37_IL22: reverse 5′ -GGATATCCAAGTGTTTATTGAGG-3′ ; ARE_ del_IL22: reverse 5′ -TATGCTTAGAAAGTCTACC-3′ . Fragments were cloned into psiCheck2 (Promega, Mannheim, Germany) and sequenced thereafter (MWG, Ebersberg, Germany). Transfection of Jurkat T cells -psiCheck2-plasmids were transiently transfected into Jurkat T cells using DMRIE-C reagent (Life Technologies). For each reaction, 4 μ g of indicated plasmid were transfected into 2.5 × 10 6 Jurkat T cells according to the manufacturer's instructions. The transfection was stopped after 5 h by adding 2 ml of Jurkat T culture medium (aforementioned) supplemented with 5% heat-inactivated FCS. After 16 h of rest, cells were harvested, further kept as unstimulated control or stimulated as described in the respective figure legend and harvested thereafter. Transfection of HEK293 cells -24 h before transfection, HEK293 cells were seeded on Poly-L-lysine (Sigma-Aldrich) coated 6-well polystyrene plates (Greiner) in the aforementioned culture medium. Transfection was conducted with 0.5 μ g of luciferase reporter plasmid and 2 μ g of SV40-driven pZeo_hTTP expression plasmid 61 or empty vector as indicated using Lipofectamine2000 (Life Technologies) according to the manufacturer's instructions. Determination of luciferase activity -After 16 h of resting, cells were harvested or further kept as unstimulated control or stimulated as described in the respective figure legend and harvested thereafter. Luciferase activity was determined using the dual luciferase reporter gene system (Promega) and an automated chemiluminescence detector (GloMax ® , Promega). Detector, Life Technologies): Two initial steps at 50 °C (2 min) and 95 °C (20 sec) were followed by 40 cycles at 95 °C (3 sec) and 60 °C (30 sec). Detection of the dequenched probe, calculation of threshold cycles (C T values), and data analysis were performed by the Sequence Detector software. Relative changes in mRNA expression compared to unstimulated control and normalized to GAPDH were quantified by the 2 −ΔΔCT method.