HEK293T and HeLa were provided by Bioresource collection—Collection of SPF-Laboratory Rodents for Fundamental, Biomedical and Pharmacological Studies supported by the Ministry of Science and Higher Education of the Russian Federation (Contract No. 075-15-2021-1067). Cell lines were cultured in media (DMEM or RPMI 1640) supplemented with 10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin, 0.25 μg/mL amphotericin B, and 2 mM GlutaMAX (Gibco). The HEK293T and HeLa cell lines were cultured in DMEM. The Jurkat 76 TPR cell line was cultured in RPMI 1640. Human PBMCs were isolated from the blood of healthy donors by gradient density centrifugation on a Ficoll-Paque (GE Healthcare) according to the standard protocol. Healthy donors provided informed consent. The cell lines were tested for the presence of mycoplasma contamination (Evrogen).

Synthetic genes were amplified and cloned into a suitable vector (Supplementary Table S1). TCR amplicons of alpha and beta chains were produced by overlapping PCR. CD80 and TCR amplicons were restricted with XbaI/BamhI and cloned into the pLV2 lentiviral vector (Clontech) under the control of the EF1a promoter (Supplementary Figure S1). Each sequence of the peptide linked to MHC-II was produced by PCR, restricted with NheI/BamHI, and ligated into the pLV2 vector carrying the MHC-II leader sequence. The CD4 amplicon was restricted with NheI/BamHI and ligated into the lentiviral pLX301 vector under the control of the CMV promoter.

The following fluorophore-conjugated antibodies were used: anti-human CD80-PE (clone W17149D, BioLegend), anti-human HLA-DR-APC (clone L243, BioLegend), anti-human CD4-APC-Alexa Fluor 750 (clone S3.5, Thermo Fisher Scientific), anti-human CD3-APC or CD3-FITC (clone OKT3, BioLegend), anti-human TCR α/β-APC (clone IP26, BioLegend), anti-human CD69-APC (clone FN50, BioLegend), anti-human IFNγ-PE (clone 4S.B3, BioLegend), and anti-human CD11c-FITC (clone 3.9, BioLegend).

Lentiviral particles were produced by PEI co-transfection of HEK293T cells with plasmids encoding genes of interest and packaging plasmids. The medium was changed after 6 h post-transfection to OptiMEM, supplemented with GlutaMAX, sodium pyruvate, and 5% FBS (Gibco). The supernatant was collected after 48 h post-transfection and filtered. The medium containing lentiviral particles was added to the cell line in a 6-well plate and centrifuged at 1200 g at 30°C for 90 min with the addition of polybrene (Merck) at 10 μg/mL. The following day, the culture medium was changed to the usual medium. After 3 days post-transduction, CD4⁺ Jurkat 76 TPR cells were selected with 1 μg/mL puromycin (InvivoGen) for 7 days.

EVs were obtained following a published protocol with some modifications (Ukrainskaya et al., 2021). The HeLa cell line was detached from the flask using 5 mM PBS-EDTA for 10 min at 37°C and 5% CO₂. Cells were centrifuged and re-suspended in DMEM, 10% FBS, and 1% pluronic F-127 (Merck) supplemented with cytochalasin B (Merck) at a concentration of 10 μg/mL. Cells were incubated for 30 min at 37°C with 5% CO₂ at a density of 1 × 10⁶ cells/mL in a T-25 flask (Corning). Then, cells were shaken intensively for 30 s for the detachment of EVs and centrifuged at 100 g for 10 min at 4°C. The supernatant was collected and centrifuged at 300 g for 10 min. The resulting supernatant was centrifuged at 3500 g for 30 min for EV precipitation. EVs were counted using the NovoCyte flow cytometer (ACEA Biosciences) and frozen at −80°C.

Cells were washed with PBS and re-suspended in PBS with fluorescent antibodies. HeLa cells were stained with anti-human CD80 and anti-human HLA-DR. Jurkat 76 TPR cells were stained with anti-human CD4, anti-human CD3-APC, and anti-human TCR. The staining was performed for 30 min at 4°C. Then, cells were washed with PBS and analyzed using the NovoCyte flow cytometer (ACEA Biosciences). Centrifuging steps were performed at 300 g at 4°C.

EVs were washed with DMEM, 10% FBS, and 1% pluronic F-127 and re-suspended in the same medium with fluorescent antibodies anti-human CD80 and anti-human HLA-DR. The staining was performed for 30 min at 4°C. Then, EVs were washed with the same medium and analyzed using the NovoCyte flow cytometer (ACEA Biosciences). Centrifuging steps were performed at 5000 g at 4°C.

HeLa cells or HeLa-derived EVs were applied to a poly-L-lysine-covered 96-well imaging glass plate (Eppendorf) and centrifuged at 100 g for 10 min at room temperature. The attached cells and EVs were fixed in 4% paraformaldehyde in PBS for 1 h at room temperature and stained with Hoechst 33342 (Invitrogen), anti-human CD80-PE, and anti-human HLA-DR-APC antibodies. Confocal images were captured using a laser scanning microscope LSM 980 (ZEISS) with a ×63 oil Plan-Apochromat objective (numerical aperture 1.4). Hoechst 33342 was exited at 405 nm, and the emission was detected in the range of 410–605 nm; PE fluorescence was exited at 543 nm, and the emission was registered in the range of 543–623 nm; APC fluorescence was excited at 639 nm, and the emission was detected in the range of 569–694 nm.

Jurkat 76 TPR cells (100,000 cells per well) were incubated with EVs or MoDCs at different T cell:EV or MoDC ratios (2:1, 1:1, 1:2, and 1:5) for 16 h in a 96-well flat bottom plate (Corning). Positive control cells were incubated with 50 ng/mL PMA (Merck) and 1 μg/mL ionomycin (Merck). The volume of each well was 200 μL. Following incubation, cells were washed and analyzed for GFP expression using the NovoCyte flow cytometer (ACEA Biosciences). Alternatively, Jurkat 76 TPR cells were analyzed for the CD69 activation marker. Jurkat 76 TPR cells were incubated with EVs under the same conditions, stained with anti-human CD69, washed, and analyzed using the NovoCyte flow cytometer (ACEA Biosciences).

CD4⁺ T-cell isolation from PBMCs was performed following the manufacturer’s protocol (STEMCELL Technologies). CD4⁺ T cells were incubated with EVs at a 1:1 ratio in AIM-V medium (Gibco) supplemented with AlbuMAX (Gibco) in a 96-well round bottom plate. After 3 h of incubation, brefeldin A (BioLegend) was added at a concentration of 10 μg/mL. For positive control, CD4⁺ T cells were stimulated with 50 ng/mL PMA (Merck) and 1 μg/mL ionomycin (Merck) for 2 h, and then brefeldin A (BioLegend) was added at a concentration of 10 μg/mL. CD4⁺ T cells were incubated for 14 h after brefeldin A addition before intracellular cytokine staining (Molodtsov et al., 2022). CD4⁺ T cells were washed with PBS and centrifuged. Cells were stained with anti-human CD3-FITC and CD4 antibodies for 15 min at 4°C. Then, cells were washed with PBS and fixed with 2% paraformaldehyde in PBS for 20 min at 4°C. Cells were washed and stained with an anti-human IFNγ antibody for 40 min at 4°C. Cells were analyzed using the NovoCyte flow cytometer (ACEA Biosciences).

Alternatively, CD4⁺ T cells were incubated with EVs at a 1:1 ratio in a 24-well plate (Corning) for 3 days in RPMI 1640 medium. After 3 days, the medium was changed to RPMI 1640 supplemented with 12 IU/mL recombinant IL-2 (STEMCELL Technologies). The cell culture medium supplemented with 12 IU/mL recombinant IL-2 was replenished every 2 days. The last medium change was absent for IL-2. At day 14, cell CD4⁺ T cells were counted and restimulated with EVs at a 1:1 ratio with culture medium changed to AIM-V medium supplemented with AlbuMAX in a 96-well round bottom plate. The intracellular IFNγ staining was performed under the same conditions as described.

Isolated PBMCs were re-suspended in RPMI 1640 supplemented with 10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin, 0.25 μg/mL amphotericin B, and 2 mM GlutaMAX (Gibco) and seeded in 25-cm² cultural flasks (Corning) at a concentration of 6∗10⁶ cell/mL. After 2 h, the unbound cells were removed, and the media were changed to fresh, supplemented with growth factors—IL-4 (100 ng/mL) and GM-CSF (50 ng/mL) (STEMCELL Technologies), and then cultivated for 6 days with a change of half of the media volume every 2 days. After 6 days, the full medium volume was changed to a fresh portion with bacterial lipopolysaccharide (10 μg/mL) and cultivated for 24 h for DC maturation. MoDCs were analyzed for the expression of CD11c, CD80, and HLA-DR on the membrane surface using the NovoCyte flow cytometer (ACEA Biosciences).

Statistical analysis was performed using GraphPad Prism 8.0 (GraphPad). The statistical test employed is denoted in each figure legend: * indicates p < 0.05, **indicates p < 0.01, *** indicates p < 0.001, and **** indicates p < 0.0001.

T-cell activation relies on two essential signals: one initiated by the interaction of TCR with pMHCs and the other facilitated by the CD28 stimulatory receptor, which binds the B7 costimulatory molecules (CD80/CD86) expressed on APCs. Notably, CD86 is a key ligand for CTLA-4, inducing an overall inhibitory effect, whereas CD80 is responsible for restraining CTLA-4 recycling (Kennedy et al., 2022). Therefore, we used transgenic cell lines engineered to express both CD80 molecules and pMHCs featuring variable peptides as a source of EVs. To ensure optimal presentation of the peptide on each MHC-II molecule of different alleles, we introduced a peptide of interest, tethered with an SG linker, to the beta chain of MHC-II—an approach distinct from exogenous peptide loading (Kozono et al., 1994; Li et al., 2022). The presence of clusters of specific pMHCs has been shown to possess superior potency in the activation of antigen-specific CD4⁺ T cells (Al-Aghbar et al., 2022).

Although MoDCs are commonly employed for the presentation of antigens on MHC-II molecules due to their robust phagocytic activity (Schlitzer et al., 2015), the generation of MoDCs often yields a limited cell population. As an alternative, the immortalization of B cells, another frequently utilized approach for antigen-presentation studies, requires extended and intricate protocols for production (Topalian et al., 1994; Linnemann et al., 2015; Wang et al., 2023). In this study, we opted for HeLa cells for the generation of EVs, a widely used and versatile cell line enabling the fast generation of transgenic cell lines (Figure 1A). We created a panel of genetic constructs with two different MHC-II molecules: HLA-DR1 (HLA-DRB1*01:01) in complex with influenza A HA_306-318 peptide and HLA-DR15 (HLA-DRB1*15:01) in complex with autoantigenic myelin basic protein MBP_85-99 peptide. As a negative control, we developed a construct with a fragment of the invariant chain—CLIP peptide—for both the DR1 and DR15 complexes that does not elicit a T-cell response. A stepwise lentiviral transduction of HeLa cells with constructs encoding CD80 and specific pMHC molecules was performed (Figure 1B). All transduced HeLa cells exhibited more than 75% positivity for the expression of CD80 and pMHC.

Obtained transgenic HeLa CD80⁺DR⁺ cell lines were treated with cytochalasin B, which induces a swift disintegration of the actin cytoskeleton and the establishment of elongated tubular protrusions that could be effectively separated through agitation. The produced CD80⁺DR⁺ EVs were separated from cells with a series of benchtop centrifugations. This procedure resulted in the production of EVs that bear a replica of the membrane surface proteins expressed by the parent cell. Obtained EVs had a mean diameter of 2,400 nm (Ukrainskaya et al., 2021) and were readily detectable with flow cytometry. More than 25% of EVs demonstrated dual positivity for CD80 and pMHC (Figure 1C). Confocal imaging of HeLa cells and HeLa-derived EVs further verified the presence of HLA-DR and CD80 on the membrane surface (Figure 1D; Supplementary Figure S2).

To evaluate the activation potential of the isolated EVs, we established a specialized cell line bearing cognate TCRs specific to tested pMHCs. To minimize potential experimental biases associated with the emergence of chimeric TCRs, we adopted the Jurkat 76 TPR (triple parameter reporter) cell line with green fluorescent protein (GFP) under transcription nuclear factor of activated T cell (NFAT) promoter, which lacks endogenous TCR alpha and beta chains (Rosskopf et al., 2018).

To augment the low endogenous expression of the CD4 molecule in Jurkat 76 TPR cells, we transduced DNA coding for this essential co-receptor (Figure 2). Subsequently, this transgenic cell line was further transduced by DNA coding for HA1.7 or Ob.1A12 TCRs, specific for HA_306-318 peptide in the context of HLA-DR1 (CD4⁺HA1.7 TCR⁺ Jurkat 76 TPR) and MBP_85-99 in the context of HLA-DR15 (CD4⁺Ob.1A12 TCR⁺ Jurkat 76 TPR), respectively, with cysteine substitutions in the constant domains to enhance chain pairing (Kuball et al., 2006). Notably, it was previously reported that HA1.7 exhibits a relatively high affinity for its cognate pMHC, whereas Ob.1A12 exhibits a lower affinity (K_D>200 μM) attributable to an alternative positioning topology of the TCR (Sundberg et al., 2007).

Next, we assessed the specificity and amplitude of the activation potential of CD4⁺TCR⁺ Jurkat 76 TPR cell lines in response to CD80⁺DR⁺ EV exposure. The Jurkat 76 TPR cell line expresses GFP upon activation, facilitating our assessment of TCR specificity (Figure 3A). Our data revealed that Jurkat 76 TPR cell activation occurred exclusively when these cells were subjected to incubation with EVs bearing cognate pMHCs (Figures 3B, C). Specifically, CD4⁺HA1.7 TCR⁺ Jurkat 76 TPR cells exposed to a five-fold excess of CD80⁺DR1-HA⁺ EVs demonstrated a substantial 73% GFP-positive response. In contrast, incubation with EVs carrying either CD80 molecules (CD80⁺ EVs) alone or irrelevant pMHCs (CD80⁺DR1–CLIP⁺ or CD80⁺DR15–MBP7⁺ EVs) resulted in negligible activation, with less than 10% of GFP-positive cells. Similarly, the CD4⁺Ob.1A12 TCR⁺ Jurkat 76 TPR line exhibited activation exclusively in response to incubation with CD80⁺DR15–MBP7⁺ EVs. Moreover, the incubation with antigen-specific EVs resulted in the expression of CD69 activation markers for both CD4⁺HA1.7 TCR⁺ and CD4⁺Ob.1A12 TCR⁺ Jurkat 76 TPR (Supplementary Figure S3).

The observed CD4⁺TCR⁺ Jurkat 76 TPR activation response exhibited dose-dependency (Figures 4A, B), as evidenced by levels of GFP-positive reporters in response to incubation with EVs in different ratios. The number of GFP-positive cells increased with the administration of a greater quantity of EVs. We noted that EVs carrying weakly interacting DR15–MBP7 complexes induced an even more potent activation response in CD4⁺Ob.1A12 TCR⁺ Jurkat 76 TPR cells compared to CD80⁺DR1-HA⁺ EVs interacting with CD4⁺HA1.7 TCR⁺ Jurkat 76 TPR in the same quantity. Additionally, we compared the activation capacity of CD80⁺DR1-HA⁺ EVs and MoDCs to activate CD4⁺HA1.7 TCR⁺ Jurkat 76 TPR cell lines (Supplementary Figure S4). EVs activated cell lines more potently than MoDCs, further confirming their activation ability.

Collectively, both high- and low-affinity TCRs demonstrated activation in response to EVs without inducing any discernible nonspecific background activation. We also demonstrated that pMHCs remained intact throughout the EV generation protocol involving cytochalasin B treatment and retained their capability to engage TCR. Consequently, our data demonstrate the high capacity of EVs to trigger T-cell responses.

Our data suggest that EVs boast a notable advantage as carriers of a substantial load of pMHCs and exhibit relative stability. These characteristics render them particularly suitable for the antigen-specific expansion of CD4⁺ T cells usually required for the study of T-cell responses originating from rare clones. Therefore, we aimed to investigate the feasibility of utilizing EVs to stimulate and expand human CD4⁺ T cells in an antigen-specific manner. To achieve this, we generated EVs bearing the Flu B HA_270-286 epitope presented within the HLA-DR15 complex (CD80⁺DR15–FLU⁺ EVs). This specific epitope has been previously reported as an elicitor of anti-viral CD4⁺ T-cell responses in individuals bearing the HLA-DRB1*15:01 allele (Xiaomin et al., 2020).

We isolated CD4⁺ T cells from four healthy individuals with a confirmed heterozygous HLA-DRB1*15:01-positive haplotype. First, isolated cells were incubated with EVs for 16 h and then fixed and intracellularly stained for the production of IFNγ. The difference in the percentage of IFNγ⁺ cells between CD80⁺DR15–FLU⁺ and CD80⁺DR15–CLIP⁺ EVs was not significant (Figure 5A). Alternatively, isolated cells were subjected to stimulation with CD80⁺DR15–FLU⁺ EVs for 3 days, followed by a subsequent 10-day expansion with a periodical addition of IL-2 to further maintain T-cell growth. On the 10th day, T cells were subjected to secondary stimulation. They were exposed to CD80⁺DR15–FLU⁺ or CD80⁺DR15–CLIP⁺ EVs for an additional 16 h. The difference in the percentage of IFNγ⁺ cells between CD80⁺DR15–FLU⁺ or CD80⁺DR15–CLIP⁺ EVs was statistically significant (p = 0.0286) (Figure 5B; Supplementary Figure S5). Collectively, our findings illustrate the efficacy of EVs in selectively promoting the proliferation of antigen-specific CD4⁺ T cells, enabling the detection of antigen-specific responses from rare T-cell clones.