CD28 Autonomous Signaling Orchestrates IL-22 Expression and IL-22-Regulated Epithelial Barrier Functions in Human T Lymphocytes

IL-22 is a member of the IL-10 cytokine family involved in host protection against extracellular pathogens, by promoting epithelial cell regeneration and barrier functions. Dysregulation of IL-22 production has also frequently been observed in acute respiratory distress syndrome (ARDS) and several chronic inflammatory and autoimmune diseases. We have previously described that human CD28, a crucial co-stimulatory receptor necessary for full T cell activation, is also able to act as a TCR independent signaling receptor and to induce the expression of IL-17A and inflammatory cytokines related to Th17 cells, which together with Th22 cells represent the main cellular source of IL-22. Here we characterized the role of CD28 autonomous signaling in regulating IL-22 expression in human CD4+ T cells. We show that CD28 stimulation in the absence of TCR strongly up-regulates IL-22 gene expression and secretion. As recently observed for IL-17A, we also found that CD28-mediated regulation of IL-22 transcription requires the cooperative activities of both IL-6-activated STAT3 and RelA/NF-κB transcription factors. CD28-mediated IL-22 production also promotes the barrier functions of epithelial cells by inducing mucin and metalloproteases expression. Finally, by using specific inhibitory drugs, we also identified CD28-associated class 1A phosphatidylinositol 3-kinase (PI3K) as a pivotal mediator of CD28-mediated IL-22 expression and IL-22–dependent epithelial cell barrier functions.


INTRODUCTION
IL-22 has been discovered in 2000 as a member of the IL-10 family that affects the functions of several non-hematopoietic cells, such as epithelial cells, keratinocytes, and hepatocytes (1). IL-22 is mainly produced by Th17 cells, a subset of CD4 + T helper (Th) cells that regulate efficient immune defense against extracellular pathogens (2) and by a unique CD4 + Th cell subset, named Th22 (3)(4)(5). Other cell types have been described to produce IL-22 such as natural killer (NK) cells, ILC3 subset of innate lymphoid cells and at lower extent macrophages (6)(7)(8). IL-22 elicits its main effect on epithelial cells by stimulating their regeneration and proliferation, and by promoting epithelial barrier functions essential for the host defense against extracellular microbes at mucosal surfaces (9)(10)(11)(12)(13)(14). However, high levels of IL-22 and IL-17 were detected in plasma of patients with sepsis-induced acute respiratory distress syndrome (ARDS) and were correlated with poor prognosis (15). More recently, a novel severe acute respiratory syndrome-related coronavirus 2 (SARS-Cov2) has been identified as the etiological agent of COVID-19 (16), a respiratory pandemic disease that, in several patients, causes ARDS associated with high levels of pro-inflammatory cytokines (17) including Th17 cytokines such as IL-22, IL-17, and IL-6 (18,19). Moreover, an excessive production of IL-22 together with other cytokines of the Th17 signature has also been involved in maintaining and amplifying the chronic state of several inflammatory and autoimmune diseases (20)(21)(22)(23)(24)(25)(26). Therefore, the identification of stimulatory molecules and associated signaling pathways regulating and/or amplifying IL-22 production may provide new insight into immune defense against pathogens and novel therapeutic targets for inflammatory and autoimmune diseases.
CD28 is one of the most important costimulatory receptor required for full T cell activation and differentiation (27). CD28 interacts with B7.1/CD80 or B7.2/CD86 on the surface of professional antigen presenting cells (APCs) and lowers the threshold of the T cell receptor (TCR), thus enhancing the early signaling events necessary for optimal cytokine production, cell cycle progression and survival (28,29). Human CD28 is also able to deliver TCR-independent signals, which leads to the production of pro-inflammatory cytokines and chemokines (30,31) in CD4 + T cells. We also evidenced that CD28 autonomous stimulation of peripheral blood CD4 + T cells from either healthy donors (HD) or type 1 diabetes (T1D) or multiple sclerosis (MS) patients promotes the production of inflammatory cytokines and chemokines related to the Th17 cell phenotype (32)(33)(34), including IL-22 (35).
In this work, we elucidated the role of CD28 autonomous stimulation in the regulation of IL-22 expression and IL-22mediated effector functions. We found that IL-22 gene expression and secretion was strongly up-regulated by CD28 in human CD4 + T cells. Similar to IL-17A, CD28-induced IL-22 gene expression was regulated by both RelA/NF-kB and IL-6-regulated STAT3 transcription factors. Co-culture experiments of CD28-activated T cells with CACO-2 epithelial cell lines evidenced a critical role of IL-22 in regulating the expression of mucins and metalloproteases. Finally, we also evidence a pivotal role of CD28-associated class 1A phosphatidylinositol 3-kinase (PI3K) in regulating both IL-22 expression and IL-22-dependent epithelial barrier functions.

Cells, Abs, and Reagents
Human primary CD4 + T cells were enriched from PBMC by negative selection using an EasySepTM isolation kit (#17952, STEMCELL Technology) and cultured in RPMI 1640 supplemented with 5% human serum (Euroclone, UK), Lglutamine, penicillin and streptomycin. The purity of the sorted population was 95% to 99%, as evidenced by staining with anti-CD3 plus anti-CD4 Abs. Human naïve CD4 + CD45RA + and effector/memory CD4 + CD45RO + T cells were sorted using a high speed cell sorter (Moflo Astrios EQ, Beckman Coulter). Purity of sorted cells was consistently > 98%, and was acquired on a cytometer (CytoFLEX, Beckman Coulter). PBMCs were derived from buffy coats or anonymous healthy blood (HD) donors provided by the Policlinico Umberto I (Sapienza University of Rome, Italy). Written informed consent was obtained from blood donors and both the informed consent form and procedure was approved by the Ethics Committee of Policlinico Umberto I (ethical code N. 1061bis/2019, 13/09/2019). CACO-2 epithelial cell line from human colon was provided by ATCC (Maryland, USA) and cultured in DMEM supplemented with 10% FBS (Euroclone, UK), L-glutamine, penicillin and streptomycin.

Real-Time PCR
CD4 + T cells were plated at 2 × 10 6 ml −1 in 24-well plate or 24well plate inserts in the experiments of co-culture with CACO-2 cells and stimulated for the indicated times with control isotype matched Abs (Ig) or crosslinked anti-CD28 (2 mg ml −1 ), or anti-CD3 (2 mg ml −1 ) or anti-CD3 plus anti-CD28 Abs (2 mg ml −1 ). Total RNA was extracted by either CD4 + T cells or CACO-2 cells using Trizol according to the manufacturer's instructions and was reverse-transcribed into cDNA by using Moloney murine leukemia virus reverse transcriptase (Thermo Fisher Scientific CA, USA). TaqMan Universal PCR Master Mix and human IL-6, IL-22, MMP9, MUC1, and GAPDH primer/probe sets were from Thermo Fisher Scientific. The relative quantification was performed using the comparative C T method. The results were expressed as arbitrary units (AU) by using the mean value of cells stimulated with control Ig as C T calibrator or fold induction (F.I.) over control Ig-stimulated cells after normalization to GAPDH.

Plasmids Construction, Cell Transfection, and Cytofluorimetric Analysis
The pIL-22 (-644)-GFP construct containing the GFP under the control of the −644 bp region upstream of the transcriptional start site and the +156 region upstream the translational start site was generated as previously described (36). Briefly, human IL-22 promoter fragment (−1069/+156) was amplified from genomic DNA isolated from human PBMCs by using AmpliTaq (Applied Biosystem). The primers used (excluding additional flanking cloning/restriction sites) were as follow: forward 5′-GGTATTTGCATTTTGATACTTGCATTTGCT-3′ and 5′-TGCAGACAATTCTAACTCGAG-3′. By using an internal AseI restriction site located at position − 644 of the human IL-22 promoter, the pIL-22 (-644)-GFP construct was obtained by subcloning the PCR fragment into AseI-Hind III sites within the CMV promoter of pEGFP-N1 vector (Clontech, UK). The sequence of human IL-22 promoter fragment (−644/+156) was verified by DNA sequencing and was as follow: 5′-ATTAATA

C A A T T T T A A G A T A T A T T T A C T T C T G C C T T A A T T G T T A T G A T C A C T T A A A A A T
Plasmid vectors expressing HA-tagged human RelA, IKKa, IKKb and NIK were previously described (37,38).
CD28-positive Jurkat T cells were electroporated (at 260 V, 960 mF) in 0.45 ml RPMI-1640 supplemented with 10% FBS with 1 mg of pIL-22(-644)-GFP together with pcDNA3 control vector 10 mg HA-RelA, or HA-IKKa or HA-IKKb or 20 mg HA-NIK. Twenty-four hours after transfection, the cells were analyzed by using a BD FACScan flow cytometer (BD Biosciences, Italy). Mean fluorescence intensity was calculated on a total of 5 × 10 3 GFP-positive events by gating for SSC and FSC.

Cytotoxicity Assay
The cytotoxicity of AS605240 on CACO-2 cells was evaluated by propidium iodide (PI) staining (10 mg ml −1 ). CACO-2 cells were plated at 2.5 × 10 5 cells ml −1 in 24-well plates and treated with 10 mM AS605240 or DMSO, as vehicle control, for 48 h. Cytotoxicity was analyzed by a BD Biosciences FACScalibur (Mountain View, CA) by quantifying the percentage of PI positive cells. Results were calculated from at least three independent experiments.

Statistical Analysis
The sample size was chosen based on previous studies to ensure adequate power. Parametrical statistical analysis (mean and SEM) was performed to evaluate differences between continuous variables through Prism 5.0 (Graph Pad Software, San Diego, CA) using Student t test. For multiple group comparisons, significant differences were calculated using nonparametric Mann-Whitney U or Wilcoxon tests, and linear regression analyses were performed using the Pearson chi-squared test. For all tests, p values < 0.05 were considered significant.

CD28 Autonomous Stimulation Up-Regulates IL-22 Gene Expression and Secretion in Human CD4 + T Cells
We have recently found that CD28 stimulation induces the expression of Th17 related cytokines in CD4 + T cells from HD, relapsing-remitting MS (RRMS) and T1D patients (32)(33)(34). More recent preliminary data also evidenced that CD28 stimulation up-regulated IL-22 gene expression in peripheral CD4 + T cells from stable RRMS patients (35). In order to verify whether CD28-induced IL-22 expression was also effective in HD and clarify the molecular basis of this activity, we performed a kinetic analysis of IL-22 gene expression by stimulating human CD4 + T cells from HD with an agonistic anti-CD28 Ab (CD28.2) binding the same epitope recognized by B7 molecules (40) or anti-CD3 (UCHT1) or anti-CD3 plus anti-CD28 Abs. CD28 stimulation of CD4 + T cells from one representative HD induced a significant increase (p < 0.01) of IL-22 gene expression within 6 h that further increased 24 to 48 h and decreased 72 h after stimulation ( Figure 1A). No significant differences in neither IL-22 mRNA levels nor kinetics were observed between CD28 and CD3 plus CD28 stimulation, whereas CD3 stimulation alone induced only a slight increase of IL-22 gene expression after 48 h ( Figure 1A). Similar results were obtained by analyzing IL-22 gene expression in CD28-stimulated CD4 + T cells from a larger sample size (n = 7) (Figures 1B, C). CD28-induced IL-22 gene expression was also associated with a strong increase of IL-22 cytokine secretion after 24 to 48 h from stimulation ( Figures 1D,  E). In the human system, IL-22 gene expression has been described to be induced by IL-6 (41). Consistently with our previous data (33), CD28 stimulation induced a strong IL-6 secretion ( Figure 1F). IL-6-mediated signaling alone was not sufficient for inducing IL-22 gene expression ( Figure 1G). However, the blockade of IL-6-mediated signaling by a neutralizing anti-IL-6 Ab strongly inhibited CD28-induced IL-22 gene expression ( Figure 1H) and secretion ( Figure 1I) in CD4 + T cells from HD. As we previously observed for IL-17A (33) and other pro-inflammatory cytokines (31), the analysis of CD28-induced IL-22 secretion in purified naïve CD4 + CD45RA + (Figure 2A) or effector/memory CD4 + CD45RO + ( Figure 2B) evidenced no significant differences between the two population ( Figure 2C).
These data evidence that CD28 intrinsic signaling regulates IL-22 gene expression and secretion in both naïve and effector/ memory CD4 + T cells in a IL-6-dependent manner.

STAT3 and RelA/NF-kB Transcription Factors Regulate CD28-mediated IL-22 Gene Expression
The human IL-22 gene promoter contains two putative STAT3 responsive elements (STAT3 I and II) and two NF-kB binding sites (NF-kB I and II) upstream of the transcription start site that have been involved in IL-22 promoter trans-activation (42, 43) ( Figure 3A). We have previously demonstrated that human CD28 individual stimulation induces the activation and nuclear translocation of RelA (31,44) and, more recently, we evidenced that CD28 induces the phosphorylation on Tyr705 of STAT3 (pSTAT3) and its nuclear translocation in a IL-6dependent manner. pSTAT3 and RelA/NF-kB in turn cooperate for inducing the promoter trans-activation and transcription of IL-17A (33). To assess if pSTAT3 and RelA were also involved in CD28-mediated up-regulation of IL-22, we performed chromatin immunoprecipitation (ChIP) assays in peripheral CD4 + T cells stimulated with agonistic anti-CD28.2 Abs. To do that, three distinct oligonucleotide probes were used; probe I (−338 to −203) for both STAT3 I and NF-kB I binding sites, probe II (−213 to −110) for NF-kB II binding site and probe III (−131 to −20) for both STAT3 II binding site and TATA box ( Figure 3A). The kinetic analysis of the specific recruitment to the IL-22 promoter evidenced that RelA efficiently bound NF-kB I but not NF-kB II consensus sequences within 6 h from stimulation and persisted over 24 h, whereas no significant recruitment of RelB was observed ( Figure 3B). pSTAT3 was significantly recruited on both STAT3 I and STAT3 II consensus sequences on the IL-22 promoter with similar kinetics ( Figure 3C). The specific recruitment of both RelA on the NF-kB I and pSTAT3 on both I and II binding sites was confirmed in CD4 + T cells from a larger sample size (n = 9) and was associated to CD28-induced transcriptional activation of IL-22 promoter, as evidenced by RNA polymerase II (pol II) promoter occupancy ( Figure 3D). The significant inhibition of IL-22 gene expression exerted by the STAT3 inhibitor S31-201 (45,46) and the NF-kB inhibitor PS1145 (47) (Figure 3E), confirmed the mutual cooperation of both transcription factors in regulating CD28induced IL-22 expression.

Functional Relevance of CD28-Mediated IL-22 Production in Epithelial Cell Functions
The main functional effects of IL-22 are exerted on epithelial cells, such as intestinal cells, where IL-22 promotes the production of anti-microbial peptides, mucins and matrix metalloproteases (MMPs) required for epithelial barrier functions and protection against extracellular pathogens (8,9). We next analyzed the functional role of CD28-mediated activation of IL-22-producing T cells on intestinal epithelial cells. To do that, CD4 + T cells were stimulated with control isotype matched Abs (Ig), or anti-CD28, or anti-CD3 or anti-CD3 plus anti-CD28 Abs and co-cultured in trans-well plates with CACO-2 cells for different times. The gene expressions of mucin 1 (MUC1), MMP9 and S100A9 anti-microbial peptide were then analyzed. The kinetic analysis evidenced that both MMP9 ( Figure 4A) and MUC1 ( Figure 4B) gene expressions were strongly up-regulated in CACO-2 cells after 48 h of coculture with CD28-stimulated CD4 + T cells, whereas no significant increase in S100A9 was observed ( Figure 4C). Similar results were obtained by analyzing MMP9 ( Figure 4D) and MUC1 ( Figure 4E) gene expression in CACO-2 cells cocultured with CD28-stimulated CD4 + T cells from a larger sample size (n = 9). Moreover, the up-regulation and MUC1 gene expression induced by co-culturing CACO-2 cells with CD28-stimulated CD4 + T cells was also associated with a strong increase of MUC1 protein content ( Figure 4F) after 48 h from stimulation. In contrast to the gene expression data ( Figure 4D), when we looked at MMP9 secretion we found high levels of MMP9 in culture supernatants from CACO-2 cells cocultured with CD4 + T cells stimulated with control Ig that further increased following CD28 stimulation (Supplementary Figure S1A), thus suggesting that CD4 + T cells also produced MMP9, as previously reported (48). CD4 + T cells secreted, indeed, high levels of MMP9 in culture supernatant and no significant differences were observed following CD28 stimulation (Supplementary Figure S1B).  Figure 1H) and secretion ( Figure 1I), MUC1 gene expression was strongly impaired by anti-IL-6 neutralizing Abs in CACO-2 cells cocultured with CD28-stimulated CD4 + T cells ( Figure 4H).
We have recently demonstrated that class 1A PI3K activity is crucial for CD28-induced up-regulation of pro-inflammatory cytokines related to Th17 cell phenotype in both HD and RRMS patients (33,35). Consistently, a non-cytotoxic dose of the class 1A PI3K inhibitor AS605240 (52) (Figures 5A, B), significantly impaired MMP9 and MUC1 gene expression in CACO-2 cells co-cultured with CD28-stimulated CD4 + T cells ( Figure 5C) as well as CD28induced IL IL-22 ( Figure 5D) and IL-6 secretion ( Figure 5E). Moreover, the inability of recombinant IL-6 to restore impaired IL-22 gene expression in AS605240-treated CD28-stimulated CD4 + T cells, strongly support a role of class 1A PI3K downstream CD28. Altogether these data support a crucial role of CD28associated PI3K in IL-22-regulated epithelial barrier functions. Frontiers in Immunology | www.frontiersin.org October 2020 | Volume 11 | Article 590964 DISCUSSION IL-22 is an important cytokine that strengthen epithelial barrier functions against extracellular pathogen and regulates tissue homeostasis and repair. However, excessive production of IL-22 has been also associated with severe complications in several pathogen infections (12), including the recently identified Sars-Cov2 (18) as well as with inflammatory bowel diseases such as Crohn's disease (23), neurological disorders such as MS and Guillain-Barrè syndrome (25,26), and rheumatoid arthritis (53,54). Therefore, the identification of receptors and associated signaling mediators regulating IL-22 expression could represent an important goal of the ongoing research in inflammation and autoimmunity.  In mouse T cells, IL-22 is mainly regarded as a Th17 cytokine (41,55,56), whereas in human CD4 + T cells IL-22 expression often does not correlate with the expression of IL-17 or the master transcriptional regulator of Th17 cells retinoic acid receptor-related orphan nuclear receptor gt (RORgt) (57,58) and is produced by a unique memory Th22 cell subset (3,4). Herewith, we evidence that CD28 stimulation induced IL-22 expression and production in the absence of TCR engagement (Figure 1). The up-regulation of IL-22 transcription by CD28 autonomous stimulation of peripheral CD4 + T cell followed a kinetic ( Figures 1A, B) similar to IL-17A (33) and no significant differences were observed between naïve (CD45RA) and effector/ memory (CD45RO) CD4 + T cells ( Figure 2). Moreover, we found that CD28-induced up-regulation of IL-22 depended on CD28-induced IL-6 ( Figure 1F), as demonstrated by the strong inhibitory effects exerted by neutralizing anti-IL-6 Abs ( Figures  1H, I). However, although IL-6 itself is able to prime IL-22 production in activated human CD4 + T cells (41), stimulation of CD4 + T cells with IL-6 alone is not sufficient for up-regulating IL-22 gene expression ( Figure 1G) but requires a second signal delivered by CD28. In this contest, among the most promising therapeutic strategies for dampening excessive inflammatory responses and ARDS in SARS-Cov2 infection, tocilizumab, an anti-IL-6R mAb inhibitor that interferes with IL-6-mediated signaling and has already approved for RA treatment (59,60) resulted effective in preventing the cytokine storm, ARDS and to reduce mortality in COVID-19 patients (61)(62)(63).
IL-6 regulates the expression of both IL-17A and IL-22 through IL-6 receptor-associated STAT3 (41,55,(64)(65)(66)(67). We have recently demonstrated that CD28 stimulation induced a strong and persistent STAT3 phosphorylation on Tyr705 and its nuclear translocation in CD4 + T cells. We also found that activated STAT3 was also recruited to the proximal human IL-17A promoter and induced its trans-activation in response to CD28 stimulation (33). In T cells from STAT3 deficient mice, IL-22 expression is impaired (42,68), whereas the overexpression of constitutively active STAT3 up-regulates IL-22 expression (66). For instance, two functional STAT3 binding sites have been identified upstream of the transcription start site within the mouse IL-22 promoter (42). The human IL-22 promoter contains similar high consensus sequences ( Figure 3A) and we found that CD28 stimulation promoted the recruitment of STAT3 on both binding sites in human CD4 + T cells ( Figures  3C, D). These data together with the strong inhibition of CD28induced IL-22 expression by the STAT3 inhibitor S31-201 ( Figure 3E), support a direct role of STAT3 in regulating IL-22 gene expression in CD28-stimulated CD4 + T cells.
IL-22 regulates essential functions in host defense against pathogens by mediating the crosstalk between the immune system and epithelial cells. In fact, IL-22 regulates the expression of several genes that maintain the barrier functions of intact epithelium (9). These include molecules involved in tight and gap junctions such as claudins that keep epithelial cells together and regulate the permeability of intestinal epithelial barrier to water, nutrients and ions (73), anti-microbial proteins such as S100A (55,74) and mucins. MUC1, a membrane-bound mucin, is the main component of the intestinal mucus layer that provides a physiological barrier to prevent close contacts between epithelial cells and microbes (75). In addition to its physiological barrier functions, MUC1 has been described to modulate the functions of inflammatory T cells. In MUC1-deficient mice, Th17 cells and IL-17-producing ILC3 cells were expanded and exacerbated colitis (76). Similarly, a reduced glycosylation of MUC1 increased both susceptibility to colitis and progression to colon cancers (77,78). In both human and mouse colonic epithelial cell lines, IL-22 has been reported to promote MUC1 expression (79,80). We extend these data by evidencing an important role of CD28 in regulating epithelial barrier functions. During CD28 stimulation, RelA/NF-kB and IL-6-associated STAT3 cooperate for inducing IL-22 expression and secretion ( Figure 3) that in turn acts on epithelial cells by promoting the up-regulation of MUC1 ( Figure 4). Moreover, IL-22 secreted by CD28-activated CD4 + T cells also stimulate MMP9 expression in CACO-2 cells (Figure 4). Interestingly, Pujada et al. recently showed that MMP9 expression in colonic epithelium was associated with a more efficient epithelial barrier functions by maintaining epithelial integrity, microbiota and immune homeostasis (81). However, MMP9 is also up-regulated in inflamed intestine of IBD patients and has been recently suggested as a predictor marker of disease exacerbation in CD (82,83). Moreover, high levels of MMP9 were also found in the serum and cerebrospinal fluid of MS patients and increase during relapses (84,85). Thus, anti-inflammatory compounds able to dampen excessive production of MMPs might represent potential therapeutic strategies for inflammatory diseases. Our findings on the ability of class 1A PI3K inhibitor to impair both CD28-induced IL-6 and IL-22 production as well as IL-22-dependent MMP9 and MUC1 expression in CACO-2 cells ( Figure 5) are consistent with data from Yan et al. who identified the PI3K/Akt/mTORC1 as a crucial pathway in regulating both Th17 cell-mediated inflammation and lung injury in ARDS (15).
Our results provide novel insights on the role of CD28 and associated signaling mediators in IL-22 regulation in human CD4 + T cells and may provide the biological bases for the development of new therapeutic strategies to dampen inflammatory and autoimmune disorders.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Ethical Committee of the Policlinico Umberto I (ethical code N. 1061bis/2019, 13/09/2019). The patients/ participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
MK performed most of the experiments, analyzed the data, interpreted the results, and helped in writing the manuscript. CA, SF, SC, MS, and MB performed parts of the experiments and data analyses. SA contributed with human samples for the study. MK, CA, SF, MS, and LB were involved in the discussion about the data. LT designed the study, coordinated the work, and wrote the manuscript. All authors contributed to the article and approved the submitted version.