Human tolerogenic dendritic cell subtypes exert divergent effects on induction of cytotoxic CD4+ T cells

Tolerogenic dendritic cells (tolDC) are currently in clinical trials for the treatment of autoimmune diseases such as rheumatoid arthritis and multiple sclerosis. The methods for producing therapeutic tolDC vary widely, with little being known about the commonalities and differences of distinct cell products in terms of their regulatory actions on CD4⁺ T cells. We compared human monocyte-derived tolDC generated with vitamin D3 alone or in combination with dexamethasone. We found marked differences in the surface expression of HLA-DR and immune regulatory molecules, but also found commonalities, e.g. a strongly reduced capacity to produce interleukin-12 and a concomitant decreased ability to induce interferon-γ secretion by allogeneic CD4⁺ T cells. To gain a deeper understanding of how these tolDC types exert their regulatory effects, we co-cultured them with CD4⁺ T cells from rheumatoid arthritis patients or healthy controls and analysed the gene expression profile and function of the responding T cells. We found that tolDC generated with vitamin D3 alone, but not in combination with dexamethasone, induced potent cytotoxic activity in the responding CD4⁺ T cells as demonstrated by an enhanced cytotoxic gene signature, increased levels of intracellular granzyme B, and superior cytotoxic activity towards myeloid and B cells. These data identify cytotoxicity as an atypical CD4⁺ T helper cell effector function induced by some but not all tolDC types, with implications for their individual clinical applications.

Introduction

Autoimmune diseases such as rheumatoid arthritis (RA) and multiple sclerosis (MS) are debilitating conditions, caused by a breach in immune tolerance to self-tissues and leading to uncontrolled and pathogenic immune responses to auto-antigens. Current therapeutics mainly act through dampening inflammation, but do not target the underlying cause of these pathologies. To counter this, tolerogenic dendritic cell (tolDC) therapies have been developed to reinstate immune tolerance in an antigen-specific manner, and a number of these have been or are being tested in clinical trials (1, 2).

Human tolDC are commonly generated from peripheral blood monocytes and a variety of methods have been described to endow these cells with stable tolerogenic function (3). These include the use of immunosuppressive agents at various stages of the monocyte-derived DC (moDC) differentiation process to attenuate their maturation into immunogenic cells. In general, tolDC are characterised by a ‘semi-mature’ phenotype expressing variable levels of MHC II, intermediate expression of co-stimulatory CD86, increased expression of regulatory molecules (e.g., PD-L1/2, MerTK and ILT3), as well as an anti-inflammatory cytokine secretion profile (e.g., low IL-12p70 and high IL-10). However, despite these typical features, a common transcriptomic ‘signature’ amongst different tolDC types has not been found, making it likely that these cells act upon T cells through both overlapping and distinct pathways (4, 5).

Phase I clinical trials with a number of tolDC types in several autoimmune diseases (including RA, MS, type I diabetes and Crohn’s disease) have shown that this therapeutic approach is safe and well-tolerated by patients (6). However, although there is anecdotal evidence for some clinical improvements in some patients, unequivocal and long-lasting clinical benefits of this treatment have not yet been established. The current challenge for the field is, therefore, to improve the efficacy of tolDC treatment to meet its enormous theoretical potential. Although a number of promising initiatives are ongoing to achieve this, e.g. through clinical trials that compare different routes of tolDC administration (7), progress is hampered by an incomplete understanding of how these cells exert their regulatory effects. This is further compounded by the variety of methods utilised to generate tolDC and their application in different autoimmune disease settings; making comparisons between different tolDC types difficult.

Two of the most commonly used agents for the generation of tolDC are the glucocorticoid dexamethasone (Dex) and the active form of vitamin D₃ (VD₃). We have developed therapeutic tolDC for the treatment of RA, using a combination of these immunosuppressive agents (referred to as DexVD₃DC). We have shown that these tolDC have potent regulatory activity in vitro, including the induction of IL-10-producing Tr1-like cells (8–12). We have also shown the therapeutic benefit of equivalent murine tolDC, generated with Dex and VD₃, in experimental models of inflammatory arthritis (13, 14). Our phase 1 clinical trial (AuToDeCRA – Autologous Tolerogenic Dendritic Cells for RA) showed that injection of these autologous tolDC into an inflamed joint is safe (11), and we have recently completed recruitment of RA patients to a phase 2a clinical trial (AuToDeCRA-2) in which we are testing different routes of tolDC administration (intra-articular, intra-dermal and intra-nodal). DexVD₃DC are also undergoing testing in the TOLERANT trial at University Medical Centre Utrecht (15).

Addition of VD₃ (in the absence of Dex) during differentiation of human monocytes into DC is another common method for the generation of tolDC (VD₃DC). These VD₃DC have decreased expression of co-stimulatory molecules and IL-12p70, with reduced ability to induce T cell proliferation and prime Th1 responses (16, 17). VD₃DC have also been shown to induce antigen-specific hyporesponsiveness in CD4⁺ T cells and to promote the induction of Tr1-like cells or FoxP3⁺ Tregs in vitro (18–20). Equivalent murine tolDC, generated with VD₃ as the sole tolerogenic agent, suppressed disease symptoms in the experimental autoimmune encephalomyelitis model and improved the balance between regulatory and pathogenic T cells (21–24). VD₃DC have been trialled in MS patients (7).

Here, we investigated the regulatory properties of human monocyte-derived DexVD₃DC and VD₃DC. We performed phenotypic profiling of these tolDC types and investigated their regulatory actions on CD4+ T cells. We performed transcriptomic analyses of CD4⁺ T cells from RA patients that had been primed by tolDC, as well as functional studies. Our data show overlapping but also divergent properties of these tolDC types, particularly regarding the induction of cytotoxic activity in CD4⁺ T cells.

Materials and methods

The minimum information about tolerogenic antigen presenting cells (MITAP) checklist was used for this paper (25).

Peripheral blood samples and ethical approval

Ethical approval for the use of leukocyte reduction system (LRS) cones from platelet donations to the National Health Service Blood and Transplant (NHSBT) was provided by the Faculty of Medical Sciences Ethics Committee. For work carried out at GlaxoSmithKline, research use was in accord with the terms of the informed consents under an IRB/EC approved protocol. Ethical approval for the use of peripheral blood donations obtained from healthy volunteers with informed consent was provided by The Animal Welfare and Ethical Review Body (AWERB), Newcastle University. RA blood samples were obtained with informed consent from the Northeast Early Arthritis Clinic (NEAC) (Musculoskeletal Unit, Freeman Hospital, Newcastle-upon-Tyne Hospital Trust). Ethical approval was given by the Newcastle & North Tyneside 2 Research Ethics Committee under the project titled ‘Prognostic and therapeutic biomarkers in an observational inception cohort: the Northeast Early Arthritis Cohort’ (REC reference 12/NE/0251). Stored PBMC samples from four treatment-naïve early arthritis female patients, age range 46-68, with a confirmed diagnosis of RA were used.

Generation of monocyte-derived mature and tolerogenic DC subtypes

Peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation on Lymphoprep (StemCell Technologies, Vancouver, Canada). CD14⁺ monocytes were freshly isolated from PBMC by magnetic positive selection using anti-CD14 MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany), as previously described in detail (26). Monocytes were seeded at 0.5x10⁶ cells/ml in a 24-well plate (1ml/well) in RF10; RPMI-1640 supplemented with 10% v/v foetal bovine serum (FBS) (Gibco, Waltham, MA, USA), 100U/ml penicillin, 100µg/ml streptomycin and 2mM L-glutamine (Sigma-Aldrich, Burlington, MA, USA). To generate moDC, cells were cultured in the presence of GM-CSF and IL-4 (each 50 ng/ml; Immunotools, Friesoythe, Germany) and incubated at 37 °C, 5% CO₂ for 7 days. Medium supplemented with cytokines (50ng/ml) was refreshed on day 3. On day 6, cells were either stimulated for 24hrs with 100ng/ml Lipopolysaccharide (LPS) (Sigma) to generate mature (Mat)DC or were left untouched as immature (Im)DC. To generate DexVD₃DC, dexamethasone (Dex) (Sigma) was additionally added at 1μM on days 3 and 6 of culture and 0.1nM 1,25-dihydroxyvitamin D3 (Calcitriol) (VD₃) (Tocris, Bristol, UK) was added alongside Dex and LPS on day 6 of culture. VD₃DC were generated as described above with the addition of 10nM VD₃ on days 0 and 3 of culture and LPS on day 6. On culture day 7, DC were harvested and washed extensively for further processing (see below). Viability and cell number was determined by trypan blue exclusion (Sigma).

T cell isolation

CD4⁺ T cells were isolated from LRS cones using the RosetteSep human CD4⁺ T cell enrichment cocktail (StemCell) according to manufacturer’s instruction with the exception of adding 75µl cocktail/ml of sample. The sample was diluted 1:2 with PBS supplemented with 2% FBS (Sigma) and CD4⁺ T cells were separated by density gradient centrifugation on Lymphoprep (StemCell). For further isolation of naïve CD4⁺ T cells, CD45RO⁺ cells were depleted using CD45RO MicroBeads (Miltenyi Biotec), according to manufacturer’s instruction. Purity of naïve CD4⁺ T cells was confirmed by flow cytometry and was typically >85% (data not shown). Cells were cryopreserved using FBS (Sigma) supplemented with 10% dimethyl sulfoxide (DMSO) (Wak-Chemie Medical GmbH, Steinbach, Germany).

Cell surface phenotyping of moDC

Cells were stained for viability using Zombie Aqua™ or Zombie NIR™ (Biolegend, San Diego, CA, USA) according to manufacturer’s instruction. The reaction was quenched with cold staining buffer (Ca²⁺ and Mg²⁺ free PBS supplemented with 3% FBS, 1mM EDTA and 0.01% sodium azide). To prevent non-specific antibody binding, cells were centrifuged and resuspended in staining buffer supplemented with 10% TruFcX Fc receptor block (Biolegend) for 20 minutes at 4 °C. Cells were stained for phenotypic analysis using the following mouse anti-human antibodies: CD83 - FITC (HB25e), CD86 - BV785 (IT2.2), HLA-DR - BV510 (L243), PD-L1 - PE-Cy7 (MIH1), MERTK - PE (590H11G1E) and ILT3 - APC (ZM4.1) in staining buffer supplemented with 10% Brilliant Buffer Plus and 5% True Stain Monocyte Blocker (Biolegend) for 30 minutes at 4°C. Cells were acquired immediately using the Attune NxT flow cytometer. UltraComp eBeads™ (Invitrogen) were used for compensation according to manufacturer’s instruction. Cells were protected from light throughout staining and acquisition. Data was analysed using FlowJo software version 10 (Treestar Inc, OR, USA). Fluorescence minus one (FMO) controls were used to gate live cells following exclusion of debris and doublets.

Cytokine analysis

Concentrations of cytokines in culture supernatants were quantified using either the 18plex ProcartaPlex™ Human Th1/Th2/Th9/Th17 Cytokine Panel (Thermo) or the V-plex Plus Human Proinflammatory Panel 1 Kit (Meso Scale Discovery, Rockville, MD, USA) according to manufacturer’s instruction.

MoDC/CD4⁺ T cell co-cultures

Cryopreserved CD4⁺ T cells were thawed, labelled with 0.5 µM CellTrace^™ Far Red proliferation tracking dye (Invitrogen, Waltham, MA, USA) according to manufacturer’s instruction. Labelled CD4⁺ T cells were then co-cultured with freshly generated moDC (1x10⁴/well). Co-culture was carried out in 96-well U bottom plates at a DC: T cell ratio of 1:10 (1x10⁵ T cells/well) in RF10 at 37°C, 5% CO₂ for 6 days. On day 6, supernatants were collected and stored at -80 °C. Cells were stained for viability as described above and surface stained for flow cytometry using mouse anti-human CD3 - BUV395 (UCHT1) and CD4 - PE-Cy7 (RPA-T4) in staining buffer supplemented with 4µg/ml human IgG (Hizentra^®; CSL Behring, King of Prussia, PA, USA) to prevent non-specific antibody binding. Cells were incubated for 30 minutes at 4°C. Cells were centrifuged and resuspended in staining buffer to be acquired immediately on the BD LSR Fortessa X20^™ or fixed in PBS supplemented with 1% formaldehyde. Fixed cells were stored at 4 °C protected from light and acquired within 7 days. BD^™ CompBeads (BD Biosciences) were used for compensation according to manufacturer’s instruction. Data was analysed using FlowJo software version 10 (Treestar Inc, OR, USA).

Gene expression profiling by Nanostring

Cryopreserved PBMC from early arthritis patients were thawed and labelled with CellTrace^™ as above and co-cultured with freshly generated allogeneic healthy donor moDC at a 1:10 DC: PBMC ratio (1x10⁵ moDC and 1x10⁶ PBMC/well) in a 24-well plate in RF10 at 37°C, 5% CO_2. On day 6, the cells were harvested and stained for fluorescence activated cell sorting (FACS) using viability dye Zombie NIR^™ and the following mouse anti-human antibodies: CD3-BUV395 (UCHT1), CD4-PE-Cy7 (RPA-T4), CD11c-PE/Dazzle™ 594 (BU15) and CD14-FITC (M5E2) in staining buffer supplemented with human IgG (Hizentra^®). Cells were strained through a 35μm cell strainer and acquired at 4 °C on the BD FACSAria^™ II using a 70μm nozzle. Events were gated sequentially as follows: cells, single cells, live cells (Zombie NIR^-), proliferated cells (CellTrace^-), CD11c^-, CD14^- and CD3⁺CD4⁺. Approximately 3 x 10⁵ cells were sorted into cold RF10 per sample. Sorted CD4⁺ T cells that had proliferated in response to matDC (T-MatDC), DexVD₃DC (T-DexVD₃DC) or VD₃DC (T-VD₃DC) were lysed in RLT buffer containing 1% β-mercaptoethanol, passed through a QIAshredder column (Qiagen, Hilden, Germany) and stored at -80°C. Total RNA was extracted using the RNeasy^® Micro Kit (Qiagen) according to manufacturer’s instruction. RNA concentration was determined using Qubit (Invitrogen) and 100ng RNA per sample was analysed on the NanoString nCounter^® platform using the Immune Exhaustion panel (NanoString Technologies, Seattle, WA, USA). Raw count data is provided in Supplementary Data Sheet 1.

Analysis of gene expression data

Analysis of nCounter^® gene expression data was performed using R software version 4.2.1 in association with the Bioconductor repository (27). The NanoTube R package (28) was used for data processing and quality control (QC). Quality control of the gene expression count data was performed as recommended by NanoString^® Gene Expression Data Analysis Guidelines. In brief, the data was normalised in two steps using positive spike-in controls and geNorm-selected housekeeping genes (29). Removal of any outlier samples was performed based on different QC flags as recommended in NanoString^® Gene Expression Data Analysis Guidelines. Differential gene expression analysis was performed by fitting negative binomial generalised linear models to the raw count data using the DESeq2 Bioconductor package (30). To compare variation between groups, a likelihood ratio test (LRT) was performed, followed by Wald tests for pairwise comparisons between T-MatDC, T-DexVD₃DC and T-VD₃DC. Genes with a raw p-value < 0.05 and a Benjamini-Hochberg adjusted p-value < 0.05 for the LRT and Wald test, respectively, were considered to be differentially expressed. Gene Set Enrichment Analysis (GSEA) was conducted using the Fast Gene Set Enrichment Analysis (FGSEA) method outlined by Korotkevich et al. (31) which estimates gene set p-values based on the adaptive multi-level split Monte-Carlo approach. Gene sets were defined based on the NanoString Immune Exhaustion panel annotations and were considered significant if they had a p-value < 0.05. Principal Components Analysis (PCA) was performed using stats::prcomp R function. For hierarchical clustering, centred Pearson correlation and complete linkage were used as distance metric and agglomeration method, respectively. The results were visualised using factoextra::fviz_pca, volcano3D::volcano3D (32) and NMF::aheatmap R functions. Transcription factor (TF) activities were inferred using the Univariate Linear Model (ULM) method within the decoupleR package. Using the gene-level Wald statistics and the human CollecTRI regulatory network as input, UML regresses the TF-Gene interaction signed weights from the regulatory network against the differential expression statistics to compute activity scores per TF.

Intracellular staining for granzyme B

Freshly generated moDC (1x10⁴/well) were co-cultured with thawed, allogeneic naïve T cells in 96-well U bottom plates at a 1:10 DC: T cell ratio in RF10 at 37 °C, 5% CO₂ for 6 days. On day 6, cells were harvested, stained for viability as previously mentioned and fixed using Fixation/Permeabilisation Buffer (eBioscience) for 30 minutes at 4°C. Cells were permeabilised using Permeabilisation Buffer (eBioscience) for 15 minutes at 4°C. To block non-specific staining, cells were blocked with 2% mouse serum (Sigma) for 15 minutes prior to a 30-minute incubation with mouse anti-human granzyme B antibody - AF647 (GB11) at 4°C. Cells were washed, resuspended in staining buffer and acquired immediately on the BD FACSymphony™ A5. Data was analysed using FlowJo software version 10 (Treestar Inc, OR, USA).

Cytotoxicity assay

T cells were primed in co culture with moDC for 6 days as mentioned above. On day 6, primed T cells were labelled with Tag-it Violet^™ (Biolegend) to differentiate them from the target PBMC and 1x10⁵/well were co-cultured with thawed PBMC from the original moDC donor in 96-well U bottom plates at a 2:1 T cell: PBMC ratio in RF10 at 37 °C, 5% CO₂ for 24 hours. After 24 hours, cells were stained with Zombie Aqua^™ and Annexin V (Biolegend) for evaluation of apoptosis. To identify different cell populations, cells were stained with the following mouse anti-human antibodies: CD3 - BUV395 (UCHT1), CD19 - BV785 (HIB19) and CD11c - PE/Dazzle™ 594 (BU15) alongside human IgG (Hizentra^®). Cells were washed and fixed in PBS supplemented with 1% formaldehyde. Fixed cells were stored at 4 °C protected from light and acquired on the BD FACSymphony™ A5 within 7 days. BD^™ CompBeads (BD Biosciences) were used for compensation according to manufacturer’s instruction. Data was analysed using FlowJo software version 10 (Treestar Inc, OR, USA).

Statistical analysis

Statistical analysis for flow cytometry and cytokine data was performed using Prism version 10 (GraphPad Software, San Diego, CA, USA). Statistical significance was determined by performing Kruskal-Wallis and Dunn’s multiple comparisons tests or a one-way ANOVA with Tukey’s paired post hoc analysis on ln-transformed data. Results were considered significant when p < 0.05.

Results

Phenotype and cytokine profile of DexVD₃DC and VD₃DC

The function of different DC subtypes is linked to the profile of molecules expressed at the cell surface. Aligning to protocols for tolDC generation in research and clinical trial studies (1–4), we compared the phenotypes of tolDC generated with vitamin D₃ alone (VD₃DC) and in combination with dexamethasone (DexVD₃DC) by measuring expression of a set of DC maturation and regulatory -associated markers using flow cytometry (Figure 1). Characteristic markers of DC maturation (CD83, CD86, HLA-DR) were lowly expressed in immature DC (ImDC) and increased upon maturation with LPS (MatDC). Both DexVD₃DC and VD₃DC maintained low levels of CD83 upon LPS maturation with intermediate expression of CD86, whilst HLA-DR was highly expressed by DexVD₃DC but not by VD₃DC. In terms of expression of regulatory cell surface molecules, a different pattern was observed between the two tolDC types. DexVD₃DC but not VD₃DC showed a significant reduction in expression of programmed-death ligand 1 (PD-L1) compared to MatDC. Furthermore, DexVD₃DC uniquely showed high expression of the inflammation resolution molecule MERTK, whilst VD₃DC expressed the highest level of the inhibitory receptor ILT3. Differences were also found for the cytokine secretion profiles of the two tolDC types (Figure 2). DexVD₃DC produced significantly lower levels of pro-inflammatory IL-6, IL-12p70, IL-23 and TNF, and higher levels of anti-inflammatory IL-10 compared to MatDC. Interestingly, although VD₃DC also showed lower levels of IL-12p70 production compared to MatDC, they did not produce high levels of IL-10 after LPS stimulation but produced high levels of TNF. These data show that the two tolerogenic treatments have contrasting effects on the resulting tolDC in terms of the expression of regulatory molecules expressed or secreted, suggesting that these cells may exert their tolerogenic effect through different mechanisms.

Figure 1

Figure 1. Phenotypic analysis of maturation and regulatory- associated markers in DexVD3DC and VD3DC compared to ImDC and MatDC controls. (A) Fluorescence Intensity, (B) Geometric Mean Fluorescence Intensity (GMFI), and (C) Percentage positive of CD83, CD86, HLA-DR, PD-L1, MERTK and ILT3, respectively, were measured using flow cytometry. Histograms (A) are representative of eight independent experiments. Data shown as individual values where n = 8 with a horizontal line representing the median value. Statistical significance was determined by performing Kruskal-Wallis and Dunn’s multiple comparisons tests. Significance is represented as *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.

Figure 2

Figure 2. Cytokine secretion of tolDC compared to MatDC controls. Concentration (pg/ml) of (A) IL-12p70, (B) IL-10, (C) TNF, (D) IL-23 and (E) IL-6 in culture supernatants of MatDC (green), DexVD3DC (red) and VD3DC (blue) was measured using either Luminex® technology (A–D) or the V-plex Plus Human Proinflammatory Panel 1 Kit (Meso Scale Discovery) (E). Data shown as individual values where n = 8 (A–D) or n = 13 (E) with a horizontal line representing the median value. Statistical significance was determined by performing Kruskal-Wallis and Dunn’s multiple comparisons tests. Significance is represented as *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.

DexVD₃DC and VD₃DC exert differential effects on CD4⁺ T cell proliferation and IFN-γ production

The T cell stimulatory capacity of tolDC types was determined by co-culture with allogeneic naïve T cells obtained from healthy donors (Figure 3). VD₃DC-primed T cells (T-VD₃DC) showed reduced proliferation compared to T-MatDC, whereas priming of T cells with DexVD₃DC (T-DexVD₃DC) yielded comparably high levels of proliferation to T cells primed with MatDC (T-MatDC) (Figure 3A). Both T-VD₃DC and T-DexVD₃DC, however, showed a significantly reduced production of pro-inflammatory IFN-γ compared to T-MatDC, suggesting that both tolDC types exerted an immunomodulatory response (Figure 3B).

Figure 3

Figure 3. Effect of tolDC on the CD4+ T cell response compared to MatDC controls. (A) Percentage proliferation of the CD4+ T cells was measured using flow cytometry analysis of the CellTrace™ proliferation tracking dye and (B) concentration of IFN-γ (pg/ml) in supernatants was determined using the V-plex Plus human proinflammatory panel 1 kit (Meso Scale Discovery). Data shown as individual values where n = 6 with a horizontal line representing the median value. Statistical significance was determined by performing Kruskal-Wallis and Dunn’s multiple comparisons tests. Significance is represented as **p < 0.01.

Transcriptomic characterisation of tolDC primed-CD4⁺ T cells from RA patients

To compare the functional implications of different tolDC types in-depth and in a disease-relevant setting, we investigated their effect on the gene expression profile of CD4⁺ T cells that had been primed in co-culture with allogeneic DexVD₃DC, VD₃DC or MatDC. We established allogeneic DC/PBMC co-cultures with PBMC from four treatment-naïve RA patients and after 6 days of culture we extracted the CD4⁺ T cells that had expanded by fluorescence-activated cell sorting (FACS). NanoString technology was used to analyse expression of 785 genes involved in immune activation, immune suppression, immune status, immune checkpoints, epigenetics, metabolism and microenvironment.

Initial principal component analysis (PCA) revealed clustering of tolDC-primed T cells (T-DexVD₃DC and T-VD₃DC) and clear separation from T-MatDC at the mRNA level (Figure 4A). We identified 113 differentially expressed genes (DEGs) between T-DexVD₃DC, T-VD₃DC and T-MatDC (Supplementary Table 1). The data revealed upregulation of several key regulatory T cell-specific genes including FOXP3, IL10RA, TIGIT, SESN3 and TNFRSF9, and downregulation of T cell stimulation-linked genes IFNG and MAP3K8 and Th1 transcription factor TBX21 in T cells co-cultured with both tolDC types compared to the T-MatDC (Figure 4B). Transcription factor inference analysis further indicated a profile of reduced transcription factor activity in tolDC-primed T cells when compared to those primed with MatDC, except for FOXO3, which showed increased activity in T-DexVD₃DC only (Supplementary Table 2). Additionally, LAG3 was upregulated in T-VD₃DC, whilst SOCS1 and IL21R were downregulated compared to T-DexVD₃DC and T-MatDC, and BTLA, TNFSF4 and PDCD1 were upregulated in T-DexVD₃DC compared to T-VD₃DC and T-MatDC, once again highlighting distinct regulatory signatures in T cells primed by tolDC generated by different protocols. Gene set enrichment analysis (GSEA) showed chemokine signalling as the pathway with the greatest significance between groups (Supplementary Table 3), with expression of CXCR6, CCR2, CXCR3, CCL2, XCL1/2 and CCL5 increased and CCL3/L1 and CXCL3 decreased in T cells co-cultured with tolDC compared to T-MatDC. Indeed, CXCR6 and CCR2 were very highly expressed in T-VD₃DC when compared with T-DexVD₃DC and T-MatDC, suggestive of a more inflammatory phenotype induced by VD₃DC (Supplementary Table 2). Interestingly, GSEA additionally revealed cytotoxic activity as one of the key pathways with the largest effect size between groups, showing that 9 cytotoxicity-related genes were differentially expressed (Supplementary Tables 1, 3). Of these, GZMA, GNLY, GZMK were upregulated in T-MatDC and T-VD₃DC, NCR3 and GZMB were upregulated in T-VD₃DC only, and PRF1 was upregulated in T-DexVD₃DC but to a greater extent in T-VD₃DC (Figure 4C). A signature of 11 transcription factors was inferred with reduced activity in T-VD₃DC when compared to T-DexVD₃DC, of which only TFAP2A was predicted to regulate GZMB expression. Transcription factors JUNB and HLX were inferred to have significantly increased activity in T-VD₃DC. Transcription factors known to activate cytotoxicity genes (e.g., EOMES and TBX21) were among those with reduced activity in both T-tolDC when compared to T-MatDC, indicating a complex and non-canonical induction of cytotoxicity by VD₃DC that requires future exploration (Supplementary Table 2). Notably, the relative expression of cytotoxicity associated genes was the lowest in T-DexVD₃DC, potentially revealing a major difference in the regulatory activities of DexVD₃DC and VD₃DC in addition to indicating a possible role for dexamethasone – treated DC in suppressing cytotoxic gene expression in T cells.

Figure 4

Figure 4. Differential gene expression analysis of CD4+ T cells primed with DexVD3DC, VD3DC or MatDC. CellTrace™ proliferation tracking dye-labelled PBMC from RA patients were cultured with DexVD3DC, VD3DC or MatDC for 6 days. CellTrace™ negative CD4+ T cells were isolated by Fluorescence-activated cell sorting (FACS). Cells were then lysed and gene expression was determined using the NanoString nCounter® platform. (A) Principal component analysis (PCA) showing clustering of the different groups. PC1: principal component 1; PC2 principal component 2. Each smaller circle represents an individual sample, coloured as follows: T-MatDC (Green), T-DexVD3DC (Red) and T-VD3DC (Blue). The larger circle in the centre of each ellipses represents the mean value of each group and ellipses represent a 95% confidence interval. (B) 3-way comparison polar plot of differentially expressed genes (Benjamini-Hochberg adjusted p value <0.05) between groups. Genes that were significantly upregulated in one group compared to the other two were coloured as follows: T-MatDC (Green), T-DexVD3DC (Red) and T-VD3DC (Blue). Genes significantly upregulated in two groups compared to the reference group are coloured as follows: genes upregulated in T-MatDC and T-VD3DC compared to T-DexVD3DC are cyan, genes upregulated in T-MatDC and T-DexVD3DC compared to T-VD3DC are yellow, and genes upregulated in T-VD3DC and T-DexVD3DC compared to T-MatDC are purple. Non-significant genes (Benjamini-Hochberg adjusted p value > 0.05) are coloured grey. The x, y position of each gene denotes the degree in which it is associated with one or more groups. (C) Heatmap of cytotoxicity-related gene expression levels that were significantly differentially expressed (p < 0.05) between T-VD3DC and T-DexVD3DC. N = 4.

VD₃DC, but not DexVD₃DC, induce granzyme B expression and cytotoxic activity in naïve CD4⁺ T cells

As an indicative marker and key component of the cytotoxic killing apparatus which was markedly differentially expressed at the gene expression level, we looked to confirm the expression of granzyme B at the protein level in CD4⁺ T cells co-cultured with the different DC types. To this aim, we used naïve CD4⁺ T cells from healthy donors, which lack expression of granzyme B (Figure 5A), thus enabling us to investigate the induction of this protein. In alignment with the mRNA data (Figure 4C), there was a notable and significant difference in the expression of granzyme B protein in T cells that had been primed by VD₃DC as compared to DexVD₃DC, with VD₃DC inducing a higher proportion of cells expressing this protein, as well as a higher level of expression in those cells that expressed granzyme B (Figures 5B, C).

Figure 5

Figure 5. Analysis of intracellular granzyme B expression in DC-primed T cells by flow cytometry. Naïve healthy donor CD4+ T cells were co-cultured with allogeneic MatDC (green), DexVD3DC (red) or VD3DC (blue) for 6 days. (A) One representative experiment of granzyme B expression in live naïve T cells, T-MatDC, T-DexVD3DC and T-VD3DC. (B) Percentage positive of live cells, (C) Geometric Mean Fluorescence Intensity (GMFI) of granzyme B Data shown as individual values where n = 5 with a horizontal line representing the median value. Statistical significance was determined by performing Kruskal-Wallis and Dunn’s multiple comparisons tests. Significance is represented as **p<0.01.

To determine whether the increased expression of granzyme B in T-VD₃DC translated to a functional increase in cytotoxicity, we co-cultured T-VD₃DC, T-DexVD₃DC and T-MatDC with PBMC derived from the allogeneic donor from which the DC types were generated (to facilitate secondary activation of the primed T cells via the same allogeneic HLA) for 24 hours and measured the percentage of cells undergoing apoptosis. Apoptosis was determined using the gating strategy shown (Figure 6A). Indeed, co-culture with T-VD₃DC resulted in an increased percentage of apoptotic PBMC compared to co-culture with T-DexVD₃DC (Figure 6B). In addition, we determined which cell types were most susceptible to killing by T-VD₃DC. Our results revealed T-VD₃DC induced a higher percentage of apoptosis of CD11c⁺ cells compared to T-DexVD₃DC, but not CD3⁺ cells. We noted that, compared to the PBMC-only control group, CD19⁺ cells showed reduced apoptosis when co-cultured with any of the allogeneic DC-activated T cell types, indicative that survival signals (T cell help) may partially offset cytotoxic CD4⁺ T cell activity. However, apoptosis was significantly higher in CD19⁺ cells cultured with T-VD₃DC compared to T-DexVD₃DC, thus suggesting that B cells and myeloid cells, but not T cells are the preferential target of cytotoxic killing in our model (Figures 6C–E). Overall, these data indicate VD₃DC are inducing a cytotoxic phenotype in CD4⁺ T cells at the mRNA and protein level, which translates to an increase in functional cytotoxicity in these cells.

Figure 6

Figure 6. Assessment of functional cytotoxicity of DC-primed T cells. MoDC (1x104/well) were co-cultured with allogeneic naïve T cells at a 1:10 DC: T cell ratio for 6 days. On day 6, cells were labelled with Tag-it Violet™(Biolegend) to differentiate them from the target PBMC and 1x105/well were co-cultured with thawed PBMC from the original moDC donor at a 2:1 T cell:PBMC ratio for 24 hours. (A) Example gating strategy used to determine percentage of apoptosis in (B) total PBMC, (C) CD3+ cells, (D) CD19+ cells and (E) CD11c+ cells. Doublets and cells that were stained positive for Tag-it Violet™ were excluded. Apoptosis was then determined either in the total PBMC population (A top right) or in the individual populations specified using Zombie Aqua™(ZA) and Annexin V (AV) (Biolegend). Cells that were AV+ (Q1:ZA-, AV+ and Q2: ZA+, AV+) were considered apoptotic. Data shown as individual values with mean ± SEM where n = 6. Statistical significance was determined by performing a one-way ANOVA with Tukey’s paired post hoc analysis on ln-transformed data. Significance is represented as *p<0.05.

Discussion

This study identified functional differences between DexVD₃DC and VD₃DC, indicating divergent regulatory activity between these tolDC cell types, and highlighting non-canonical induction of cytotoxic CD4⁺ T cells.

Comparison of the phenotype and the T cell stimulatory capacity of these tolDC types showed that they conformed to the cell characteristics previously reported by us and others (8, 10, 19, 26, 33). Both DexVD₃DC and VD₃DC displayed a semi-mature phenotype and expressed their typical regulatory cell surface markers. Thus, DexVD₃DC expressed the inflammation resolution marker MerTK, which is upregulated by dexamethasone in monocyte-derived DC (34, 35) and is currently being used as a quality control marker for clinical-grade DexVD₃DC in our AuToDeCRA-2 trial. VD₃DC, on the other hand, did not express high levels of MerTK but did display a marked increase in the regulatory markers PD-L1 and ILT3, which have been implicated in the induction of Tregs and T cell anergy (36, 37). We also noted significant differences in the cytokine secretion profiles of these tolDC types (in response to LPS), with DexVD₃DC being superior in terms of IL-10 production, and VD₃DC (but not DexVD₃DC) producing IL-6, IL-23 and TNF levels similar to mature DC. Whilst IL-10 undoubtedly plays an essential role in controlling pathological inflammation and the induction of Tr1 cells (38, 39), TNF can exert both pro-inflammatory and anti-inflammatory roles, depending on the context. For example, TNF acts synergistically with IL-1β to drive the maturation of immunogenic DC (40, 41), but it can also promote the expansion and function of FoxP3⁺ Tregs through TNFR2 signaling (42). One cytokine feature that was highly similar between the two tolDC types was their very low or undetectable production of IL-12p70, likely explaining their poor ability to induce IFN-γ production in allogeneic naïve CD4⁺ T cells. The observation that DexVD₃DC have a relatively strong ability to induce T cell proliferation but not IFN-γ secretion has been noted by us previously in both allogeneic and autologous settings (8, 12). It is probable that our proliferation findings are linked to the pronounced differences in HLA-DR expression level between tolDC types, whereby DexVD₃DC express HLA-DR levels equivalent to matDC and VD₃DC express HLA-DR levels equivalent to immature DC. HLA expression has implications for signal avidity through the TCR, both in our allogeneic experiments and in in vivo antigen specific systems. HLA is likely to be an important factor in determining downstream T cell effector functions in a therapeutic context. Taken together, these results showed some similarities but also clear differences in phenotypic and functional characteristics of these two tolDC types.

It should be noted that variations in tolDC phenotype may not only depend on the pro-tolerogenic agents used, but also on the patient population from which the tolDC have been generated. For example, elegant work (43) has shown that VD₃DC from MS patients exhibited more pro-inflammatory features than VD₃DC from healthy donors, due to alterations in the aryl hydrocarbon receptor (AhR) and NF-κB pathways, impairing their tolerogenic function. In contrast, this issue does not seem to affect DexVD₃DC to the same extent, at least not when generated from the peripheral blood of RA patients, which are highly similar to DexVD₃DC from healthy donors in terms of their cell surface phenotype, cytokine production profile and T cell suppressive action (10). It could be speculated that Dex’s broad and potent immunosuppressive actions are less sensitive than VD3 to the pro-inflammatory milieu from which the monocytes were obtained. Nevertheless, patient background should be considered when designing tolDC therapy. In the case of the impaired tolerogenic function of VD₃DC from MS patients (43), co-treatment with the drug dimethyl fumarate to enhance the therapeutic action of VD₃DC was proposed.

Because these drug-induced tolDC have been developed for therapeutic application in autoimmune conditions, we investigated their modulatory effects on CD4⁺ T cells derived from four drug-naïve early RA patients. Although the small size of this cohort constitutes a limitation of this study, its strength lies in the fact that the use of these patient samples precluded interference from disease-modifying anti-rheumatic drugs. We found similarities and differences in the regulatory gene profile induced by these tolDC. Most notably, we show a prominent signature of FOXO3 transcriptional regulation in T cells primed by DexVD₃DC and not in VD₃DC which may govern differences in regulatory phenotype and functions (44). We also show LAG3, a regulatory gene that we had previously reported to be upregulated in T cells primed by DexVD₃DC (12), was a more prominent feature of T cells primed by VD₃DC. This apparent discrepancy is most likely explained by the different comparator groups in our previous study, where we compared the T cell priming effects of DexVD₃DC to those of immature DC. Other differences between the two studies are the mode of T cell activation (allogeneic versus antigen-specific) and the donors (RA patients versus healthy controls). Nevertheless, the current study confirms that both tolDC types induce the expression of regulatory genes in CD4⁺ T cells from RA patients, but the pattern of expression differs and the continued exploration of tolDC responses in different contexts is both necessary and valuable.

Interestingly, our analyses revealed that VD₃DC were superior at inducing a cytotoxic gene signature in CD4⁺ T cells that is usually considered characteristic of CD8⁺ T cells. Indeed, VD₃DC induced expression of granzyme B protein, which discriminated most highly between the T-DexVD₃DC and T-VD₃DC cells at the transcriptional level. In line with our observed granzyme B expression, VD₃DC-primed T cells exhibited the highest cytotoxic activity, with their killing mainly directed at myeloid cells and B cells. Rather than being clearly linked with activity of transcription factors previously described to activate cytotoxicity genes such as Eomes (45) Blimp-1, T-bet (46), and Runx3 (47), the superior induction of T cell cytotoxicity by VD3DC appears to be complex. Our findings suggest that transcription factors including JUNB and HLX may influence granzyme expression and cytotoxic function through their role in T-helper cell lineage commitment (48).

The roles of individual components of cytotoxicity apparatus such as granzyme B are complex (49). In addition to its canonical, perforin-dependent role in NK and CD8⁺ T cell cytotoxicity, granzyme B is known to have extracellular activity (50). It has been identified in RA serum, plasma and synovial fluid and is thought to be associated with inflammation and degradation of cartilage proteoglycan (51–53). Conversely, granzyme B has complex perforin-independent and receptor-mediated functionalities and its expression in B cells and plasmacytoid DC has been associated with T cell regulation in inflammatory diseases and cancer (54, 55). The wide-ranging activities of cytotoxicity-related proteins induced in T cells by VD₃DC may be favourable or detrimental in disease settings, and further exploration is required to better understand the implications of this novel function of tolDC.

Despite key commonalities such as identification of a prominent role for JunB in regulatory function of T-VD₃DC, our data are to some extent in contrast to a recent study by Navarro-Barriuso et al. (56), who showed that VD₃DC slightly reduced the secretion of granzyme B in autologous CD4⁺ T cells in an antigen-specific setting. Because they used a recall antigen (tetanus toxin), the responding T cells will have been memory rather than naïve, which may explain this contrast. Another possible explanation is the different protocols used for the generation of VD₃DC, with the dose of VD₃ (1 vs 10 nM) and the mode of DC maturation (cytokine cocktail vs LPS) used as the most likely considerations. As VD₃DC are being explored and trialled as a therapeutic for the treatment of autoimmune disease, it will be important to address the basis of these contrasting findings in future studies.

Our observation that VD₃DC-induced cytotoxic CD4⁺ T cells were able to kill antigen-presenting cells (APC) including B cells may be interpreted as a mechanism by which these cells could exert therapeutic effects in autoimmune conditions, as the killing of professional APC that present autoantigen is likely to lead to a reduction in the activation of autoantigen-specific T cells. Furthermore, we show that while CD19⁺ B cell survival is supported by the presence of T cells, T-VD₃DC uniquely subvert this T cell help to induce significant B cell death. Our findings are of interest in the context of the antibody-dependent and independent roles of B cells, which are already key targets for therapeutic intervention in autoimmune conditions. Indeed, cytotoxic function has been described as a feature of certain Tregs (57). However, it should be noted that the large majority of granzyme B-expressing CD4⁺ T cells did not express FoxP3 protein (data not shown), indicating that these cells are not necessarily ‘classical’ Tregs. Whether they are more representative of induced Tr1 cells or terminally differentiated effector cytotoxic CD4⁺ T cells will require further investigation through extensive single-cell profiling by scRNAseq and/or multiparametric flow cytometry methods. In addition, such investigations may also shed further light on the apparent heterogeneity of the VD3DC-primed CD4⁺ T cells.

The potent cytotoxic activity of these VD₃DC-primed CD4⁺ T cells is a striking feature, which is unlikely to be a consequence of the allogeneic in vitro stimulation of these cells. It has previously been shown that induction of cytotoxicity in CD4⁺ T cells upon allogeneic stimulation occurs after long-term in vitro culture (58, 59), but not short-term culture (58). The comparison between T cells primed with the different DC types under identical short-term in vitro culture conditions supports the notion that the enhanced induction of cytotoxic T cells by VD₃DC does reflect a true mechanism by which these cells act. Moreover, VD₃DC express significantly lower levels of HLA-DR (Figure 1) and have impaired capacity to induce allogeneic T cell proliferation (Figure 3), thus making it highly unlikely that cytotoxicity was caused by artifactual overstimulation of T cells. In this context it is interesting to note that low doses of antigen tend to drive more potent cytolytic activity in CD4⁺ T cells (60), thus aligning with our observation that VD₃DC, with the lowest potency in terms of inducing allogeneic MHCII-dependent T cell activation, displayed the strongest ability to induce cytotoxicity. While the same study demonstrated that inflammatory cytokines were not required for cytolytic activity of CD4⁺ T cells (60), the induction of cytotoxicity by DC does correlate with expression of IL-6, IL-23 and TNF in our experiment. These cytokines are potently inhibited by dexamethasone (8, 61) and are not positively associated with inducing cytotoxicity in T cell literature to date. As IL-6 is a hallmark of tolerogenic VD₃DC (62), a possible mechanistic role in supporting a cytotoxic T cell phenotype requires further investigation.

It is now accepted that cytotoxic CD4⁺ T cells are not simply an in vitro phenomenon. Expansion of cytotoxic CD4⁺ T cells has been observed in patients with chronic viral infections, and these cells are now considered to play a critical role in anti-viral defence through the killing of virus-infected MHCII-expressing cells (63–66). However, they have also been implicated in the pathogenesis of autoimmune diseases, with expansion of these cells being positively correlated with disease severity and progression in MS patients (67) and driving excess cardiovascular mortality in RA patients (68). In addition to the expression of granzyme B alongside other cytotoxic molecules, these CD4⁺ T cells produce high levels of IFN-γ and IL-17, accumulate in inflammatory autoimmune lesions, are autoreactive and have cytolytic capacity towards target tissues; for example, in MS they have been shown to kill oligodendrocytes (69–73). It should be noted that the high expression of granzyme B in VD₃DC-primed CD4⁺ T cells was not accompanied by high levels of secreted IFN-γ in our study; on the contrary, levels of this cytokine were significantly reduced as compared to MatDC-primed T cells. Moreover, research by other groups has shown that VD₃DC polarise T cells towards high IL-10 and low IFN-γ and IL-17 cytokine production (20), thus it could be speculated that the cytotoxic activity observed in these cells is more related to their regulatory, rather than pathogenic, function. Nonetheless, further work is needed to establish whether autoantigen-loaded VD₃DC induce cytotoxic autoantigen-specific CD4⁺ T cells in vivo and if so, whether the beneficial effects these cells may have through the killing of MHCII⁺ professional APC, autoantibody-producing B cells and possibly other resident cells such as fibroblasts, outweigh any potential pathogenic activity.

In conclusion, our data show considerable divergence between the regulatory effects of DexD₃DC and VD₃DC, with only the latter displaying potent ability to induce cytotoxic CD4⁺ T cells. Further work is needed to establish to what extent the induction of cytotoxic CD4⁺ T cells by VD₃DC takes place in vivo, and whether it contributes to the therapeutic action of these tolDC.

Data availability statement

The original contributions presented in the study are publicly available. This data can be found here: NCBI’s Gene Expression Omnibus https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE313751.

Ethics statement

Ethical approval for the use of leukocyte reduction system (LRS) cones from platelet donations to the National Health Service Blood and Transplant (NHSBT) was provided by the Faculty of Medical Sciences Ethics Committee. For work carried out at GlaxoSmithKline, research use was in accord with the terms of the informed consents under an IRB/EC approved protocol. Ethical approval for the use of peripheral blood donations obtained from healthy volunteers with informed consent was provided by The Animal Welfare and Ethical Review Body (AWERB), Newcastle University. RA blood samples were obtained with informed consent from the Northeast Early Arthritis Clinic (NEAC) (Musculoskeletal Unit, Freeman Hospital, Newcastle-upon-Tyne Hospital Trust). Ethical approval was given by the Newcastle & North Tyneside 2 Research Ethics Committee under the project titled ‘Prognostic and therapeutic biomarkers in an observational inception cohort: the Northeast Early Arthritis Cohort’ (REC reference 12/NE/0251). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

GB: Data curation, Writing – review & editing, Investigation, Formal analysis, Writing – original draft, Visualization. NN: Writing – review & editing, Formal analysis. AB: Writing – review & editing, Supervision. AA: Supervision, Writing – review & editing. JF: Writing – original draft, Conceptualization, Supervision, Funding acquisition, Writing – review & editing. CH: Writing – original draft, Funding acquisition, Conceptualization, Writing – review & editing, Supervision.

Funding

The author(s) declared financial support was received for this work and/or its publication. This project was funded by the BBSRC/GSK Case studentship award BB/S507039/1, a University of Sunderland project grant and the NIHR Newcastle Biomedical Research Centre (BRC), a partnership between Newcastle Hospitals NHS Foundation Trust and Newcastle University, funded by the National Institute for Health and Care Research (NIHR).

Acknowledgments

We thank the Newcastle University Flow Cytometry Core Facility (FCCF) for assistance, Dr Lee Booty for running samples on Luminex^®, Jamie MacDonald for running samples on the NanoString nCounter^® platform and Dr. Arthur Pratt for providing RA patient samples.

Conflict of interest

Author AB was employed by GlaxoSmithKline.

The remaining author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The author CH declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2025.1698413/full#supplementary-material

Supplementary Data Sheet 1 | Raw count data for all genes across all samples. Data was acquired on the NanoString nCounter® platform using the Immune Exhaustion panel (NanoString Technologies, Seattle, WA, USA). Includes gene name and accession number.

Supplementary Table 1 | All differentially expressed genes in CD4+ T cells primed with DexVD3DC, VD3DC or MatDC after normalisation and quality control steps. CellTrace™ proliferation tracking dye-labelled PBMC from RA patients were cultured with DexVD3DC, VD3DC or MatDC for 6 days. CellTrace™ negative CD4+ T cells were isolated by Fluorescence-activated cell sorting (FACS). Cells were then lysed and gene expression was determined using the NanoString nCounter® platform. A negative binomial generalised linear model (Bioconductor package DESeq2) was fitted to the raw count data following normalisation and quality control. A likelihood ratio test was performed to compare variation between groups, followed by pairwise comparisons, using Wald tests, between T-MatDC, T-DexVD3DC and T-VD3DC. Genes with a raw p-value < 0.05 and a Benjamini-Hochberg adjusted p-value < 0.05 for the LRT and Wald test, respectively, were considered to be differentially expressed.

Supplementary Table 2 | Transcription Factor (TF) inference analysis results. TF activities were inferred using the Univariate Linear Model (ULM) method within the decoupleR package using the gene-level Wald statistics and the human CollecTRI regulatory network as input. Signed TF-Gene interaction weights from CollectTRI were regressed against the differential expression statistics to compute activity scores per TF.

Supplementary Table 3 | Gene Set Enrichment Analysis (GSEA) results. GSEA was conducted using the Fast Gene Set Enrichment Analysis (FGSEA) method outlined by Korotkevich et al. (Korotkevich et al., 2021) (31). Gene sets were defined based on the NanoString Immune Exhaustion panel annotations and were considered significant if they had a p-value < 0.05.

Abbreviations

tolDC, tolerogenic dendritic cells; moDC, monocyte-derived DC; imDC, Immature moDC; MatDC mature moDC; Dex, dexamethasone; VD₃, active form of vitamin D₃; DexVD₃DC, tolDC generated with Dex and VD₃; VD₃DC, tolDC generated with VD_3; AuToDeCRA, autologous tolerogenic dendritic cells for rheumatoid arthritis; T-DexVD₃DC, T cells primed by DexVD₃DC; T-VD₃DC, T cells primed by VD₃DC; T-MatDC, T cells primed MatDC.

References

1. Mansilla MJ, Hilkens CMU, and Martínez-Cáceres EM. Challenges in tolerogenic dendritic cell therapy for autoimmune diseases: the route of administration. Immunotherapy Adv. (2023) 3:ltad012. doi: 10.1093/immadv/ltad012

PubMed Abstract | Crossref Full Text | Google Scholar

2. Passeri L, Marta F, Bassi V, and Gregori S. Tolerogenic dendritic cell-based approaches in autoimmunity. Int J Mol Sci. (2021) 22:8415. doi: 10.3390/IJMS22168415

PubMed Abstract | Crossref Full Text | Google Scholar

3. Hilkens CMU, John DI, and Thomson AW. Development of dendritic cell-based immunotherapy for autoimmunity. Int Rev Immunol. (2010) 29:156–83. doi: 10.3109/08830180903281193

PubMed Abstract | Crossref Full Text | Google Scholar

4. Navarro-Barriuso J, Mansilla MJ, Naranjo-Gómez M, Sánchez-Pla A, Quirant-Sánchez B, Teniente-Serra A, et al. Comparative transcriptomic profile of tolerogenic dendritic cells differentiated with vitamin D3, dexamethasone and rapamycin. Sci Rep. (2018) 8:14985. doi: 10.1038/s41598-018-33248-7

PubMed Abstract | Crossref Full Text | Google Scholar

5. Navarro-Barriuso J, Mansilla MJ, and Martínez-Cáceres EM. Searching for the transcriptomic signature of immune tolerance induction—Biomarkers of safety and functionality for tolerogenic dendritic cells and regulatory macrophages. Front Immunol. (2018) 9:2062. doi: 10.3389/fimmu.2018.02062

PubMed Abstract | Crossref Full Text | Google Scholar

6. ten Brinke A, Martinez-Llordella M, Cools N, Hilkens CMU, Van Ham SM, Sawitzki B, et al. Ways forward for tolerance-inducing cellular therapies- an AFACTT perspective. Front Immunol. (2019) 10:181. doi: 10.3389/FIMMU.2019.00181

PubMed Abstract | Crossref Full Text | Google Scholar

7. Willekens B, Presas-Rodríguez S, Mansilla MJ, Derdelinckx J, Lee WP, Nijs G, et al. Tolerogenic dendritic cell-based treatment for multiple sclerosis (MS): A harmonised study protocol for two phase i clinical trials comparing intradermal and intranodal cell administration. BMJ Open. (2019) 9:e030309–e030309. doi: 10.1136/BMJOPEN-2019-030309

PubMed Abstract | Crossref Full Text | Google Scholar

8. Anderson AE, Sayers BL, Haniffa MA, Swan DJ, Diboll J, Wang X-N, et al. Differential regulation of naïve and memory CD4+ T cells by alternatively activated dendritic cells. J Leukoc Biol. (2008) 84:124–33. doi: 10.1189/jlb.1107744

PubMed Abstract | Crossref Full Text | Google Scholar

9. Anderson AE, Swan DJ, Sayers BL, Harry RA, Patterson AM, von Delwig A, et al. LPS activation is required for migratory activity and antigen presentation by tolerogenic dendritic cells. J Leukoc Biol. (2009) 85:243–50. doi: 10.1189/jlb.0608374

PubMed Abstract | Crossref Full Text | Google Scholar

10. Harry RA, Anderson AE, Isaacs JD, and Hilkens CMU. Generation and characterisation of therapeutic tolerogenic dendritic cells for rheumatoid arthritis. Ann Rheum Dis. (2010) 69:2042–50. doi: 10.1136/ard.2009.126383

PubMed Abstract | Crossref Full Text | Google Scholar

11. Bell GM, Anderson AE, Diboll J, Reece R, Eltherington O, Harry RA, et al. Autologous tolerogenic dendritic cells for rheumatoid and inflammatory arthritis. Ann Rheum Dis. (2017) 76:227–34. doi: 10.1136/annrheumdis-2015-208456

PubMed Abstract | Crossref Full Text | Google Scholar

12. Spiering R, Jansen MAA, Wood MJ, Fath AA, Eltherington O, Anderson AE, et al. Targeting of tolerogenic dendritic cells to heat-shock proteins in inflammatory arthritis. J Transl Med. (2019) 17:375. doi: 10.1186/s12967-019-2128-4

PubMed Abstract | Crossref Full Text | Google Scholar

13. Stoop JN, Harry RA, von Delwig A, Isaacs JD, Robinson JH, and Hilkens CMU. Therapeutic effect of tolerogenic dendritic cells in established collagen-induced arthritis is associated with a reduction in Th17 responses. Arthritis Rheum. (2010) 62:3656–65. doi: 10.1002/art.27756

PubMed Abstract | Crossref Full Text | Google Scholar

14. Jansen MAA, Spiering R, Ludwig IS, van Eden W, Hilkens CMU, and Broere F. Matured tolerogenic dendritic cells effectively inhibit autoantigen specific CD4+ T cells in a murine arthritis model. Front Immunol. (2019) 10:2068. doi: 10.3389/fimmu.2019.02068

PubMed Abstract | Crossref Full Text | Google Scholar

15. Stoppelenburg AJ, Schreibelt G, Koeneman B, Welsing P, Breman EJ, Lammers L, et al. Design of TOLERANT: phase I/II safety assessment of intranodal administration of HSP70/mB29a self-peptide antigen-loaded autologous tolerogenic dendritic cells in patients with rheumatoid arthritis. BMJ Open. (2024) 14:e078231. doi: 10.1136/BMJOPEN-2023-078231

PubMed Abstract | Crossref Full Text | Google Scholar

16. Penna G and Adorini L. 1α,25-dihydroxyvitamin D3 inhibits differentiation, maturation, activation, and survival of dendritic cells leading to impaired alloreactive T cell activation. J Immunol. (2000) 164:2405–11. doi: 10.4049/jimmunol.164.5.2405

PubMed Abstract | Crossref Full Text | Google Scholar

17. Naranjo-Gómez M, Raïch-Regué D, Oñate C, Grau-López L, Ramo-Tello C, Pujol-Borrell R, et al. Comparative study of clinical grade human tolerogenic dendritic cells. J Transl Med. (2011) 9:89. doi: 10.1186/1479-5876-9-89

PubMed Abstract | Crossref Full Text | Google Scholar

18. Penna G, Roncari A, Amuchastegui S, Daniel KC, Berti E, Colonna M, et al. Expression of the inhibitory receptor ILT3 on dendritic cells is dispensable for induction of CD4+Foxp3+ regulatory T cells by 1,25-dihydroxyvitamin D3. Blood. (2005) 106:3490–7. doi: 10.1182/blood-2005-05-2044

PubMed Abstract | Crossref Full Text | Google Scholar

19. Unger WWJ, Laban S, Kleijwegt FS, van der Slik AR, and Roep BO. Induction of Treg by monocyte-derived DC modulated by vitamin D3 or dexamethasone: Differential role for PD-L1. Eur J Immunol. (2009) 39:3147–59. doi: 10.1002/eji.200839103

PubMed Abstract | Crossref Full Text | Google Scholar

20. Bakdash G, van Capel TMM, Mason LMK, Kapsenberg ML, and de Jong EC. Vitamin D3 metabolite calcidiol primes human dendritic cells to promote the development of immunomodulatory IL-10-producing T cells. Vaccine. (2014) 32:6294–302. doi: 10.1016/j.vaccine.2014.08.075

PubMed Abstract | Crossref Full Text | Google Scholar

21. Mansilla MJ, Sellès-Moreno C, Fàbregas-Puig S, Amoedo J, Navarro-Barriuso J, Teniente-Serra A, et al. Beneficial effect of tolerogenic dendritic cells pulsed with MOG autoantigen in experimental autoimmune encephalomyelitis. CNS Neurosci Ther. (2015) 21:222–30. doi: 10.1111/cns.12342

PubMed Abstract | Crossref Full Text | Google Scholar

22. Mansilla MJ, Contreras-Cardone R, Navarro-Barriuso J, Cools N, Berneman Z, Ramo-Tello C, et al. Cryopreserved vitamin D3-tolerogenic dendritic cells pulsed with autoantigens as a potential therapy for multiple sclerosis patients. J Neuroinflamm. (2016) 13:113. doi: 10.1186/s12974-016-0584-9

PubMed Abstract | Crossref Full Text | Google Scholar

23. Xie Z, Chen J, Zheng C, Wu J, Cheng Y, Zhu S, et al. 1,25-dihydroxyvitamin D3-induced dendritic cells suppress experimental autoimmune encephalomyelitis by increasing proportions of the regulatory lymphocytes and reducing T helper type 1 and type 17 cells. Immunology. (2017) 152:414–24. doi: 10.1111/imm.12776

PubMed Abstract | Crossref Full Text | Google Scholar

24. Derdelinckx J, Mansilla MJ, De Laere M, Lee W-P, Navarro-Barriuso J, Wens I, et al. Clinical and immunological control of experimental autoimmune encephalomyelitis by tolerogenic dendritic cells loaded with MOG-encoding mRNA. J Neuroinflamm. (2019) 16:167. doi: 10.1186/s12974-019-1541-1

PubMed Abstract | Crossref Full Text | Google Scholar

25. Lord P, Spiering R, Aguillon JC, Anderson AE, Appel S, Benitez-Ribas D, et al. Minimum information about tolerogenic antigen-presenting cells (MITAP): A first step towards reproducibility and standardisation of cellular therapies. PeerJ. (2016) 2016:e2300–0. doi: 10.7717/PEERJ.2300/SUPP-3

PubMed Abstract | Crossref Full Text | Google Scholar

26. Hilkens CMU, Diboll J, Cooke F, and Anderson AE. In vitro generation of human tolerogenic monocyte-derived dendritic cells. In: Baldari CT and Dustin ML, editors. The Immune Synapse: Methods and Protocols. New York, NY Springer US (2023). p. 477–92. doi: 10.1007/978-1-0716-3135-5_31

Crossref Full Text | Google Scholar

27. Huber W, Carey VJ, Gentleman R, Anders S, Carlson M, Carvalho BS, et al. Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. (2015) 12:115–21. doi: 10.1038/nmeth.3252

PubMed Abstract | Crossref Full Text | Google Scholar

28. Class CA, Lukan CJ, Bristow CA, and Do K-A. Easy NanoString nCounter data analysis with the NanoTube. Bioinformatics. (2023) 39:btac762. doi: 10.1093/bioinformatics/btac762

PubMed Abstract | Crossref Full Text | Google Scholar

29. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. (2002) 3:research0034.1. doi: 10.1186/gb-2002-3-7-research0034

PubMed Abstract | Crossref Full Text | Google Scholar

30. Love MI, Huber W, and Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. (2014) 15:550. doi: 10.1186/s13059-014-0550-8

PubMed Abstract | Crossref Full Text | Google Scholar

31. Korotkevich G, Sukhov V, Budin N, Shpak B, Artyomov MN, and Sergushichev A. Fast gene set enrichment analysis. bioRxiv. (2021), 060012. doi: 10.1101/060012

Crossref Full Text | Google Scholar

32. Lewis MJ, Barnes MR, Blighe K, Goldmann K, Rana S, Hackney JA, et al. Molecular portraits of early rheumatoid arthritis identify clinical and treatment response phenotypes. Cell Rep. (2019) 28:2455–2470.e5. doi: 10.1016/j.celrep.2019.07.091

PubMed Abstract | Crossref Full Text | Google Scholar

33. Lee W-P, Willekens B, Cras P, Goossens H, Martínez-Cáceres E, Berneman ZN, et al. Immunomodulatory effects of 1,25-dihydroxyvitamin D3 on dendritic cells promote induction of T cell hyporesponsiveness to myelin-derived antigens. J Immunol Res. (2016) 2016:5392623. doi: 10.1155/2016/5392623

PubMed Abstract | Crossref Full Text | Google Scholar

34. Cabezón R, Carrera-Silva EA, Flórez-Grau G, Errasti AE, Calderón-Gómez E, Lozano JJ, et al. MERTK as negative regulator of human T cell activation. J Leukoc Biol. (2015) 97:751–60. doi: 10.1189/jlb.3A0714-334R

PubMed Abstract | Crossref Full Text | Google Scholar

35. Li V, Binder MD, and Kilpatrick TJ. The tolerogenic influence of dexamethasone on dendritic cells is accompanied by the induction of efferocytosis, promoted by MERTK. Int J Mol Sci. (2023) 24:15903. doi: 10.3390/IJMS242115903/S1

PubMed Abstract | Crossref Full Text | Google Scholar

36. Francisco LM, Salinas VH, Brown KE, Vanguri VK, Freeman GJ, Kuchroo VK, et al. PD-L1 regulates the development, maintenance, and function of induced regulatory T cells. J Exp Med. (2009) 206:3015–29. doi: 10.1084/jem.20090847

PubMed Abstract | Crossref Full Text | Google Scholar

37. Kim-Schulze S, Scotto L, Vlad G, Piazza F, Lin H, Liu Z, et al. Recombinant ig-like transcript 3-fc modulates T cell responses via induction of th anergy and differentiation of CD8+ T suppressor cells1. J Immunol. (2006) 176:2790–8. doi: 10.4049/jimmunol.176.5.2790

PubMed Abstract | Crossref Full Text | Google Scholar

38. Groux H, O’Garra A, Bigler M, Rouleau M, Antonenko S, de Vries JE, et al. A CD4+T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature. (1997) 389:737–42. doi: 10.1038/39614

PubMed Abstract | Crossref Full Text | Google Scholar

39. Saraiva M, Vieira P, and O’Garra A. Biology and therapeutic potential of interleukin-10. J Exp Med. (2019) 217:e20190418. doi: 10.1084/jem.20190418

PubMed Abstract | Crossref Full Text | Google Scholar

40. Jonuleit H, Knop J, and Enk AH. Cytokines and their effects on maturation, differentiation and migration of dendritic cells. Arch Dermatol Res. (1996) 289:1–8. doi: 10.1007/s004030050144

PubMed Abstract | Crossref Full Text | Google Scholar

41. Kaliníski P, Schuitemaker JHN, Hilkens CMU, and Kapsenberg ML. Prostaglandin E2 induces the final maturation of IL-12-deficient CD1a+CD83+ Dendritic cells: the levels of IL-12 are determined during the final dendritic cell maturation and are resistant to further modulation1. J Immunol. (1998) 161:2804–9. doi: 10.4049/jimmunol.161.6.2804

PubMed Abstract | Crossref Full Text | Google Scholar

42. Chen X, Ba&öumel M, Ma&önnel DN, Howard OMZ, and Oppenheim JJ. Interaction of TNF with TNF receptor type 2 promotes expansion and function of mouse CD4+CD25+ T regulatory cells1. J Immunol. (2007) 179:154–61. doi: 10.4049/jimmunol.179.1.154

PubMed Abstract | Crossref Full Text | Google Scholar

43. Fondelli F, Willemyns J, Domenech-Garcia R, Mansilla MJ, Godoy-Tena G, Ferreté-Bonastre AG, et al. Targeting aryl hydrocarbon receptor functionally restores tolerogenic dendritic cells derived from patients with multiple sclerosis. J Clin Invest. (2024) 134:e178949. doi: 10.1172/JCI178949

PubMed Abstract | Crossref Full Text | Google Scholar

44. Kerdiles YM, Stone EL, Beisner DL, McGargill MA, Ch’en IL, Stockmann C, et al. Foxo transcription factors control regulatory T cell development and function. Immunity. (2010) 33:890–904. doi: 10.1016/j.immuni.2010.12.002

PubMed Abstract | Crossref Full Text | Google Scholar

45. Cruz-Guilloty F, Pipkin ME, Djuretic IM, Levanon D, Lotem J, Lichtenheld MG, et al. Runx3 and T-box proteins cooperate to establish the transcriptional program of effector CTLs. J Exp Med. (2009) 206:51–9. doi: 10.1084/jem.20081242

PubMed Abstract | Crossref Full Text | Google Scholar

46. Xin A, Masson F, Liao Y, Preston S, Guan T, Gloury R, et al. A molecular threshold for effector CD8+ T cell differentiation controlled by transcription factors Blimp-1 and T-bet. Nat Immunol. (2016) 17:422–32. doi: 10.1038/ni.3410

PubMed Abstract | Crossref Full Text | Google Scholar

47. Serroukh Y, Gu-Trantien C, Hooshiar Kashani B, DeFrance M, Vu Manh T-P, Azouz A, et al. The transcription factors Runx3 and ThPOK cross-regulate acquisition of cytotoxic function by human Th1 lymphocytes. Elife. (2018) 7:e30496. doi: 10.7554/eLife.30496

PubMed Abstract | Crossref Full Text | Google Scholar

48. Mullen AC, Hutchins AS, High FA, Lee HW, Sykes KJ, Chodosh LA, et al. Hlx is induced by and genetically interacts with T-bet to promote heritable TH1 gene induction. Nat Immunol. (2002) 3:652–8. doi: 10.1038/ni807

PubMed Abstract | Crossref Full Text | Google Scholar

49. Thompson R and Cao X. Reassessing granzyme B: unveiling perforin-independent versatility in immune responses and therapeutic potentials. Front Immunol. (2024) 15:1392535. doi: 10.3389/FIMMU.2024.1392535

PubMed Abstract | Crossref Full Text | Google Scholar

50. Zheng Y, Zhao J, Shan Y, Guo S, Schrodi SJ, and He D. Role of the granzyme family in rheumatoid arthritis: Current Insights and future perspectives. Front Immunol. (2023) 14:1137918. doi: 10.3389/fimmu.2023.1137918

PubMed Abstract | Crossref Full Text | Google Scholar

51. Tak PP, Spaeny-Dekking L, Kraan MC, Breedveld FC, Froelich CJ, and Hack CE. The levels of soluble granzyme A and B are elevated in plasma and synovial fluid of patients with rheumatoid arthritis (RA). Clin Exp Immunol. (2001) 116:366–70. doi: 10.1046/J.1365-2249.1999.00881.X

PubMed Abstract | Crossref Full Text | Google Scholar

52. Buzza MS, Zamurs L, Sun J, Bird CH, Smith AI, Trapani JA, et al. Extracellular matrix remodeling by human granzyme B via cleavage of vitronectin, fibronectin, and laminin. J Biol Chem. (2005) 280:23549–58. doi: 10.1074/jbc.M412001200

PubMed Abstract | Crossref Full Text | Google Scholar

53. Ronday HK, van der Laan WH, Tak PP, De Roos JADM, Bank RA, Tekoppele JM, et al. Human granzyme B mediates cartilage proteoglycan degradation and is expressed at the invasive front of the synovium in rheumatoid arthritis. Rheumatology. (2001) 40:55–61. doi: 10.1093/RHEUMATOLOGY/40.1.55

PubMed Abstract | Crossref Full Text | Google Scholar

54. Lindner S, Dahlke K, Sontheimer K, Hagn M, Kaltenmeier C, Barth TFE, et al. Interleukin 21-induced granzyme b-expressing b cells infiltrate tumors and regulate t cells. Cancer Res. (2013) 73:2468–79. doi: 10.1158/0008-5472.CAN-12-3450

PubMed Abstract | Crossref Full Text | Google Scholar

55. Jahrsdörfer B, Vollmer A, Blackwell SE, Maier J, Sontheimer K, Beyer T, et al. Granzyme B produced by human plasmacytoid dendritic cells suppresses T-cell expansion. Blood. (2010) 115:1156–65. doi: 10.1182/BLOOD-2009-07-235382

PubMed Abstract | Crossref Full Text | Google Scholar

56. Navarro-Barriuso J, Mansilla MJ, Quirant-Sánchez B, Teniente-Serra A, Ramo-Tello C, and Martínez-Cáceres EM. Vitamin D3-induced tolerogenic dendritic cells modulate the transcriptomic profile of T CD4+ Cells towards a functional hyporesponsiveness. Front Immunol. (2021) 11:599623. doi: 10.3389/fimmu.2020.599623

PubMed Abstract | Crossref Full Text | Google Scholar

57. Bolivar-Wagers S, Larson JH, Jin S, and Blazar BR. Cytolytic CD4+ and CD8+ Regulatory T-cells and implications for developing immunotherapies to combat graft-versus-host disease. Front Immunol. (2022) 13:864748. doi: 10.3389/fimmu.2022.864748

PubMed Abstract | Crossref Full Text | Google Scholar

58. Fleischer B. Acquisition of specific cytotoxic activity by human T4+ T lymphocytes in culture. Nature. (1984) 308:365–7. doi: 10.1038/308365a0

PubMed Abstract | Crossref Full Text | Google Scholar

59. Chen L-K, Bensussan A, Burns GF, Tourvieille B, Soulié A, and Sasportes M. Allospecific proliferative human T-cell clones acquire the cytotoxic effector function after three months in culture, in IL-2 conditioned medium. Hum Immunol. (1986) 17:30–6. doi: 10.1016/0198-8859(86)90071-6

PubMed Abstract | Crossref Full Text | Google Scholar

60. Brown DM, Kamperschroer C, Dilzer AM, Roberts DM, and Swain SL. IL-2 and antigen dose differentially regulate perforin- and FasL-mediated cytolytic activity in antigen specific CD4+ T cells. Cell Immunol. (2009) 257:69–79. doi: 10.1016/j.cellimm.2009.03.002

PubMed Abstract | Crossref Full Text | Google Scholar

61. Ray A and Prefontaine KE. Physical association and functional antagonism between the p65 subunit of transcription factor NF-kappa B and the glucocorticoid receptor. Proc Natl Acad Sci. (1994) 91:752–6. doi: 10.1073/pnas.91.2.752

PubMed Abstract | Crossref Full Text | Google Scholar

62. Català-Moll F, Ferreté-Bonastre AG, Godoy-Tena G, Morante-Palacios O, Ciudad L, Barberà L, et al. Vitamin D receptor, STAT3, and TET2 cooperate to establish tolerogenesis. Cell Rep. (2022) 38:110244. doi: 10.1016/j.celrep.2021.110244

PubMed Abstract | Crossref Full Text | Google Scholar

63. Verma S, Weiskopf D, Gupta A, McDonald B, Peters B, and Sette A, Benedict CA. Cytomegalovirus-specific CD4 T cells are cytolytic and mediate vaccine protection. J Virol. (2015) 90:650–8. doi: 10.1128/jvi.02123-15

PubMed Abstract | Crossref Full Text | Google Scholar

64. Weiskopf D, Bangs DJ, Sidney J, Kolla RV, De Silva AD, de Silva AM, et al. Dengue virus infection elicits highly polarized CX3CR1+ cytotoxic CD4+ T cells associated with protective immunity. Proc Natl Acad Sci. (2015) 112:E4256–63. doi: 10.1073/pnas.1505956112

PubMed Abstract | Crossref Full Text | Google Scholar

65. Laiqing H, Shuyu Y, Duy P, Li J, Jeffrey W, Deepali S, et al. Cytokine-Dependent Induction of CD4+ T cells with Cytotoxic Potential during Influenza Virus Infection. J Virol. (2013) 87:11884–93. doi: 10.1128/jvi.01461-13

PubMed Abstract | Crossref Full Text | Google Scholar

66. Hoeks C, Duran G, Hellings N, and Broux B. When helpers go above and beyond: development and characterization of cytotoxic CD4+ T cells. Front Immunol. (2022) 13:951900. doi: 10.3389/fimmu.2022.951900

PubMed Abstract | Crossref Full Text | Google Scholar

67. Peeters LM, Vanheusden M, Somers V, Van Wijmeersch B, Stinissen P, Broux B, et al. Cytotoxic CD4+ T cells drive multiple sclerosis progression. Front Immunol. (2017) 8:1160. doi: 10.3389/FIMMU.2017.01160

PubMed Abstract | Crossref Full Text | Google Scholar

68. Broadley I, Pera A, Morrow G, Davies KA, and Kern F. Expansions of cytotoxic CD4+CD28- T cells drive excess cardiovascular mortality in rheumatoid arthritis and other chronic inflammatory conditions and are triggered by CMV infection. Front Immunol. (2017) 8:195. doi: 10.3389/FIMMU.2017.00195

PubMed Abstract | Crossref Full Text | Google Scholar

69. Broux B, Markovic-Plese S, Stinissen P, and Hellings N. Pathogenic features of CD4+CD28– T cells in immune disorders. Trends Mol Med. (2012) 18:446–53. doi: 10.1016/j.molmed.2012.06.003

PubMed Abstract | Crossref Full Text | Google Scholar

70. Broux B, Pannemans K, Zhang X, Markovic-Plese S, Broekmans T, Eijnde BO, et al. CX3CR1 drives cytotoxic CD4+CD28– T cells into the brain of multiple sclerosis patients. J Autoimmun. (2012) 38:10–9. doi: 10.1016/j.jaut.2011.11.006

PubMed Abstract | Crossref Full Text | Google Scholar

71. Ruck T, Bittner S, Gross CC, Breuer J, Albrecht S, Korr S, et al. CD4+NKG2D+ T cells exhibit enhanced migratory and encephalitogenic properties in neuroinflammation. PLoS One. (2013) 8:e81455. doi: 10.1371/journal.pone.0081455

PubMed Abstract | Crossref Full Text | Google Scholar

72. Zaguia F, Saikali P, Ludwin S, Newcombe J, Beauseigle D, McCrea E, et al. Cytotoxic NKG2C+ CD4 T cells target oligodendrocytes in multiple sclerosis. J Immunol. (2013) 190:2510–8. doi: 10.4049/jimmunol.1202725

PubMed Abstract | Crossref Full Text | Google Scholar

73. Groh V, Brühl A, El-Gabalawy H, Nelson JL, and Spies T. Stimulation of T cell autoreactivity by anomalous expression of NKG2D and its MIC ligands in rheumatoid arthritis. Proc Natl Acad Sci. (2003) 100:9452–7. doi: 10.1073/pnas.1632807100

PubMed Abstract | Crossref Full Text | Google Scholar