Human buffy coats from healthy volunteers were obtained from the Center of Clinical Transfusion Medicine Tuebingen. Samples from hematopoietic cell grafts and peripheral blood mononuclear cells (PBMC) from patients with GvHD were isolated after written informed consent had been obtained. Human leukocyte antigen (HLA) typing was performed by the Center of Clinical Transfusion Medicine Tuebingen or the HLA laboratory of the Department of Medicine II of the University Hospital Tuebingen. The study was approved by our Institutional Review Board to be in accordance with ethical standards and with the Helsinki Declaration of 1975, as revised in 2013 (IRB approvals 483/2015BO2 and 137/2017BO2).

Antibodies and reagents used for flow cytometric analyses are described in the Online Supplementary Appendix.

iNKT cells were expanded from third-party PBMC with some minor modifications as previously described (Online Supplementary Appendix).¹³ Culture-expanded iNKT cells were purified with antiiNKT MicroBeads (Miltenyi Biotech, Bergisch Gladbach, Germany). Alternatively, iNKT cells were stained with DAPI (4',6- diamidino-2-phenylindole, Merck, Darmstadt, Germany), anti- CD3, anti-CD4, anti-CD8 antibodies and PBS57-loaded CD1d tetramer allowing for enrichment of iNKT cells and their different subsets by fluorescence-activated cell sorting (FACS).

Monocyte-derived dentritic cells (Mo-DC) were generated as described previously.¹³ Blood DC from healthy donors and patients were isolated using Blood Dendritic Cell Isolation Kit II (Miltenyi Biotech). Where indicated, HLA-DR⁺ blood DC were further sorted either as CD1c⁺ conventional DC (cDC) or CD303⁺ plasmacytoid DC (pDC).

CD3⁺ T cells were isolated from human PBMC with anti-CD3 MicroBeads (Miltenyi Biotech). For proliferation analysis, T cells were marked with CFSE (carboxyfluorescein succinimidyl ester, Biolegend, San Diego, CA, USA) according to the manufacturer’s instructions and tested in a mixed lymphocyte reaction (MLR).

Major mismatched mo-DC or blood DC were plated together with allogeneic CD3⁺ T cells at a 1:1 ratio. Culture-expanded iNKT cells were added to the MLR at different doses, either directly or separated from the MLR by 0.4 μm TC-Inserts (Sarstedt, Nuembrecht, Germany). Cells were analyzed by flow cytometry for activation markers (CD69 and CD25) and proliferation (CFSE). Alternatively, T cells were incubated with anti-CD3/CD28-coated beads (ThermoFisher Scientific, Waltham, MA, USA) in the presence or absence of iNKT cells. For blocking assays, iNKT cells or DC were pre-treated with the respective antibodies or IgG control (Online Supplementary Appendix).

Apoptosis was assessed with an annexin V-FITC/propidium iodide (PI) Staining Kit (BD Bioscience, Franklin Lakes, NJ, USA), by cell cycle analysis modified according to Nicoletti²¹ or by image stream analysis (Online Supplementary Appendix). The percentage of apoptotic cells was determined by flow cytometry.

Cell culture supernatants from MLR were collected after 4 and 24 hours (h), respectively. In order to analyze cytokine production bead-based immunoassays were performed according to the manufacturer’s instructions. Cytokine release was measured by a LEGENDplex human CD8/NK-cell panel (BioLegend). Data were acquired using the Lyric flow cytometer with autosampler (BD Biosciences).

Student’s t-test and analysis of variance (ANOVA) were used for statistical analysis. P<0.05 was considered statistically significant. Data were analyzed with Prism 8 (GraphPad Software, La Jolla, CA, USA). All experiments were performed at least in technical duplicates and repeated independently at least three times using different iNKT-cell donors.

Human culture-expanded iNKT cells suppress alloreactive T-cell responses when T lymphocytes are stimulated by MHC-mismatched DC.¹³ As iNKT cells exhibit potent immunoregulatory properties through a rapid release of humoral mediators, we wondered whether this effect might be related to the inhibition of T-cell function.²² Therefore, we measured early (CD69 expression) and late activation (CD25 expression) as well as proliferation (CFSE dilution) of T cells co-incubated with DC in presence or absence of iNKT cells. iNKT cells were either added directly to the MLR or separated through a TCinsert (transwell [TW]). Direct addition of iNKT cells at different ratios to the MLR reduced T-cell activation and proliferation in a dose-dependent manner. However, iNKT cells which were separated by a TW did not prevent T-cell activation and proliferation (Figure 1A to C; Online Supplementary Figure S1). We conclude from these experiments that iNKT cells mostly rely on direct cell contact to efficiently suppress alloreactive T-cell responses. In addition, T-cell activation and proliferation initiated by artificial antigen-presenting cells (aAPC, e.g., Dynabeads) was only affected when higher numbers of iNKT cells were added to the culture (Figure 1D and E; Online Supplementary Figure S2). In particular, proliferation speed was decreased with a predominance of early daughter generations (Online Supplementary Figure S2). Our findings suggest that the interaction of iNKT cells with DC largely contributes to the control of alloreactive T cells although a minor direct impact of iNKT cells on T cells could be observed.

We performed flow cytometry to determine the phenotype of DC challenged with iNKT cells. Notably, DC numbers were highly reduced (Figure 2A) and we suspected induction of apoptosis through iNKT cells. Annexin V assays showed that culture-expanded iNKT cells rapidly induced apoptosis of allogeneic DC, while co-culture of DC with conventional allogeneic CD3⁺ T cells did not, suggesting that apoptosis induction is not only dosedependent but also specific to iNKT cells (Figure 2B and C). The Nicoletti assay revealed that DC start to defragment their DNA after co-culture with iNKT cells, which represents a further hallmark of apoptosis (Figure 2D). Interestingly, an increase of spontaneous DNA defragmentation could be observed in DC without iNKT cells after 18 h which could be explained by the lack of specific stimuli. Image stream analysis also confirmed morphologic changes in DC after co-culture with iNKT cells. Whereas DC cultured alone presented a healthy and round morphology, DC co-cultured with iNKT cells were small, squashed and with a blobbing membrane. Further, upregulation of Annexin V and loss of nuclear integrity in DC co-incubated with iNKT cells could be observed in image stream assays, confirming our previous assumption (Figure 2E). Next, we investigated whether the induction of apoptosis is specific to certain iNKT-cell subpopulations. Therefore, culture-expanded iNKT cells were sorted into double negative, CD4⁺CD8^– and CD4^–CD8⁺ subsets and co-cultured separately with allogeneic DC: all iNKTcell subsets were able to induce apoptosis of DC with comparable efficiency (Figure 2F).

In order to further elucidate the cellular and molecular mechanisms responsible for iNKT-cell-induced DC apoptosis, we first analyzed image stream data visualizing doublets consisting of DC and iNKT cells. Image stream analysis revealed a direct binding of iNKT cells (PBS57- loaded CD1d tetramer⁺) to the surface of allogeneic DC (HLA-DR⁺), which subsequently revealed positive surface staining for the apoptosis marker annexin V (Figure 3A). In order to test whether this direct cellular interaction is required, iNKT cells and DC were separated by a transwell insert demonstrating that iNKT cells were unable to induce DC apoptosis anymore (Figure 3B). We further performed blocking experiments of common key molecules to identify critical pathways responsible for apoptosis induction through iNKT cells. It was observed that blocking FasL, TRAIL or NKG2D did not significantly reduce apoptosis of DC exposed to iNKT cells. However, blocking the CD1d and invariant T-cell receptor interaction reduced DC apoptosis significantly indicating that T-cell receptor engagement contributes to efficient lysis. Further, blocking apoptosis via caspase inhibitor zVAD-fmk (Nbenzyloxycarbonyl- Val-Ala-Asp(O-Me) fluoromethylketone) or inhibition of the perforin pathway via CMA (concanamycin A) also diminished iNKT-cell-mediated cell death of DC. Moreover, the inhibition of iNKT-cell degranulation by monensin and brefeldin A was shown to impede apoptosis induction most efficiently (Figure 3C; Online Supplementary Figure S3). By adding supernatant of DC-triggered degranulated iNKT cells to viable DC, we could show that iNKT cells released cytotoxic factors during degranulation which further induced apoptosis in DC (Figure 3D). In order to identify these factors, we performed bead-based multiplex assays and thereby revealed the release of interferon-g (IFN-g), granzyme B, perforin and granulysin (Figure 3E).

Our previous observations are based on ex vivo cultured mo-DC. In order to support our findings, we additionally performed MLR and apoptosis assays using blood DC isolated from PBMC of healthy donors and GvHD patients following allogeneic HCT. Blood DC are mainly composed of cDC and pDC with the latter expressing lower levels of CD1d (Online Supplementary Figure S4A). Also, human blood DC of healthy volunteers induce activation and proliferation of MHC-mismatched T cells that can be diminished through the addition of iNKT cells (Figure 4A). Given that iNKT cells interact with DC through CD1d and CD1d engagement contributes to efficient lysis of target cells, we aimed to determine how human blood pDC and cDC are susceptible to iNKT-cell apoptosis induction. For this purpose, we isolated HLA-DR+ pDC (CD303⁺) and cDC (CD1c⁺) by FACS, co-cultured them separately with iNKT cells for 4 h and stained with annexin V and PI. We observed preferential apoptosis induction of cDC, while pDC were less affected by the addition of iNKT cells (Figure 4B). Further, we wondered whether preferential apoptosis of cDC by iNKT cells would also affect Tcell alloreactivity using fresh human blood DC as stimulators. We observed that only allogeneic cDC in contrast to pDC could induce significant T-cell activation and proliferation (Online Supplementary Figure S4B). Consequently, co-culture of these distinct blood DC subsets with allogeneic T cells and iNKT cells revealed that iNKT cells were also able to suppress activation and proliferation of alloreactive T cells induced by cDC (Figure 4C).

Finally, we tested whether our findings also apply to patients with acute GvHD having received grafts from HLA-matched donors. Therefore, we isolated blood DC from PBMC obtained from patients with clinical manifestations of acute GvHD grade ≥2 prior to induction of systemic treatment with steroids. At the time point of blood collection patients had complete donor chimerism in their peripheral blood. cDC from GvHD patients also showed higher expression levels of CD1d (Online Supplementary Figure S4C) and were more susceptible to iNKT-cellinduced apoptosis than pDC (Figure 4D), similarly as demonstrated in our previous experiments with cells from healthy donors. Next, T cells derived from donors prior to transplantation were co-cultured with blood DC from GvHD patients. Importantly, adding iNKT cells from third-party donors also inhibited alloreactive T-cell activation and proliferation (Figure 4E).