Different In Vitro-Generated MUTZ-3-Derived Dendritic Cell Types Secrete Dexosomes with Distinct Phenotypes and Antigen Presentation Potencies

Human dendritic cell (DC) dexosomes were evaluated for their function and preclinical validation for vaccines. Dexosomes are small DC-secreted vesicles that contain absorbing immune signals. Vaccine manufacturing requires a significant number of monocyte-derived DCs (Mo-DCs) from donor blood; thus, Mo-DC dexosomes are expected to serve as novel materials for cancer vaccination. In this study, we characterized a potential dexosome model using immature and mature MUTZ3-derived DCs (M-imIL-4-DC, M-imIFN-DC, M-mIL-4-DC, and M-mIFN-DC) and their dexosomes (M-imIL-4-Dex, M-imIFN-Dex, M-mIL4-Dex, and M-mIFN-Dex). Despite the lack of significant differences in viability, M-mIFN-DC showed a significantly higher level of yield and higher levels of maturation surface markers, such as CD86 and HLA-ABC, than M-mIL-4-DC. In addition, M-mIFN-Dex expressed a higher level of markers, such as HLA-ABC, than M-mIL-4-Dex. Furthermore, M-mIFN-Dex exhibited a higher level of antigen presentation potency, as evaluated using a MART-1 system, than either M-imIFN-Dex or M-mIL-4-Dex. We found that M-mIFN-Dex is one of the four types of MUTZ3-derived DCs that harbor potential immunogenicity, suggesting that DC dexosomes could be useful resources in cancer immunotherapy.


Introduction
Dendritic cells (DCs) are antigen-presenting cells (APCs) that play a central role in the immune response targeting tumor-associated antigens (TAAs) [1]. DCs take up antigens and migrate into the lymph nodes, where the antigens are presented to naive T cells through major histocompatibility complexes on the DCs. DC-primed naive CD8 + T cells become cytotoxic T lymphocytes (CTLs) that attack tumors carrying TAAs [2,3]. DCs detected at lower frequencies, i.e., less than 2%, in the human peripheral blood mononuclear cells cannot be expanded in vitro [4,5]. Harvesting a large population of mononuclear cell-rich fractions is required either from patients or donors for manufacturing DC vaccines. Therefore, cancer vaccines using DCs are usually prepared by ex vivo monocyte differentiation isolated via apheresis [6][7][8]. However, DC vaccines are more susceptible to immunosuppressive molecule effects and signals in the tumor microenvironment, making long-term DC vaccine storage difficult while maintaining their expected efficacy as cell-based products [9,10]. Therefore, vaccine research has recently focused on cell-free DC exosomes (dexosomes) [11]. Dexosomes are small DC-secreted vesicles that contain absorbing immune signals. Dexosomes have been applied as cell-free anticancer vaccine materials

MUTZ-3-Derived IL-4-DC and IFN-DC Generation Using a Serum-Free Medium
Fetal bovine serum (FBS) is required for the generation of MUTZ-3-derived DCs, and the vesicles in FBS might affect the dexosome-related experimental results. Hence, we established a protocol for MUTZ3-derived DC generation using a serum-free medium. To isolate the exosomes secreted by immature and mature MUTZ-3-derived DCs (dexosomes), MUTZ-3 progenitor cells were cultured in a serum-free medium supplemented with a cytokine cocktail for 4 days to generate immature MUTZ

DC Generation by IFN-α Increased Maturation Marker Expression on DCs
Next, we analyzed the expression levels of maturation markers in M-imIL-4-DC, Mim-IFN-DC, M-mIL-4-DC, and M-mIFN-DC. Obtaining the expression levels of surface molecules, such as CD80, CD86, CD83, HLA-ABC, HLA-DR, and CD40, was necessary to reveal the antigen presentation ability, which was higher in mature DCs and stimulated CD8 + T cells in vitro [21,22]. Both M-mIFN-DC and M-mIL-4-DC exhibited significantly higher levels of maturation markers than primary MUTZ-3, indicated by the percentage of positive cells and ∆MFI of CD80, CD86, CD83, HLA-ABC, and CD40 in Figure 3. These results indicated that MUTZ-3 could differentiate into DCs either with IL-4 or IFN-α stimulation (Figures 3 and S2). The ∆MFI of CD80, CD86, CD83, HLA-ABC, and CD40 in M-mIFN-DC was significantly higher due to the process of maturation. The

Characterization of MUTZ-3-Derived IL-4-and IFN-DCs That Secrete Dexosomes
The dexosomes from the culture supernatant of immature and mature MUTZ-3-derived DCs were definitively isolated ( Figure 1). Hereafter, the immature and mature M-

MUTZ-3-Derived IFN-DC Dexosomes Exhibit Higher Antigen-Presenting Ability
To investigate the antigen presentation ability to CD8 + T cells in isolated dexosomes, MUTZ-3-derived IL-4-and IFN-DC dexosomes were loaded directly with the MART-1 peptide. MART-1 specific CD8 + T cells could be detected using MART-1 tetramer analysis on days 14 and 21. The M-mIFN-Dex showed the number of MART-1 tetramer + /CD8 + positive cells with a time-dependent increase on days 14 and 21 compared with M-mIL-4-Dex in six healthy donors (Figure 6a,b). Furthermore, the levels of MART-1 tetramer + /CD8 + positive cells in M-mIFN-Dex were significantly higher through the process of maturation than those in M-im-IFN-Dex. This quantitative analysis indicated that M-mIFN-Dex induced higher numbers to present antigens to CD8 + T cells than M-mIL-4-Dex.
peptide. MART-1 specific CD8 T cells could be detected using MART-1 tetramer analysis on days 14 and 21. The M-mIFN-Dex showed the number of MART-1 tetramer + /CD8 + positive cells with a time-dependent increase on days 14 and 21 compared with M-mIL-4-Dex in six healthy donors (Figure 6a,b). Furthermore, the levels of MART-1 tetramer + /CD8 + positive cells in M-mIFN-Dex were significantly higher through the process of maturation than those in M-im-IFN-Dex. This quantitative analysis indicated that M-mIFN-Dex induced higher numbers to present antigens to CD8 + T cells than M-mIL-4-Dex.

Discussion
In this study, we evaluated the functions of dexosomes secreted by MUTZ-3-derived IL-4-and IFN-DCs. The viability, yield, and DC phenotypes in MUTZ-3-derived IL-4-and IFN-DCs are considered critical attributes for dexosome quality verification. Despite the lack of significant differences in viability, M-mIFN-DC exhibited a significantly higher level of yield than M-mIL-4-DC. Furthermore, M-mIFN-DC expressed significantly higher levels of maturation markers, such as CD86 and HLA-ABC, than M-mIL-4-DC. In addition, M-imIFN-Dex and M-mIFN-Dex expressed higher levels of HLA class I than M-imIL-4-Dex and M-mIL-4-Dex. Moreover, the M-mIFN-Dex exhibited a potentially higher level of antigen presentation capability based on the MART-1 peptide through HLA class I.
Traditionally, human monocyte-derived DCs have been used to study DC biology and DC vaccination efficacy, in which CD14 + monocytes are relatively abundant in the peripheral blood. To reveal the characteristics of dexosomes in vitro requires harvesting a significant number of monocytes from the peripheral blood. Hence, an experimental model using a cell line is a durable resource for the mechanistic evaluation of DC functions. Santegoets et al. reported that tumor cells of myeloid and lymphoid lineages exhibit the potential to differentiate into DC-like APCs. Furthermore, this DC-differentiating potential has also been shown for leukemia-derived cell lines, especially those that originate from myelogenous or monocytic lineages, such as THP-1, HL-60, KG-1, K562, and MUTZ-3. However, other than MUTZ-3, these cells have low differentiation potential into DCs [19]. The phenotypic and functional characteristics of MUTZ-3 derived DCs that were essential for CTL-mediated immune response generation in vivo [23,24]. Furthermore, Jurjen et al. reported that MUTZ-3-derived IL-4-DCs and IFN-DCs established a model for studying DC functions in vitro [20]. Studies have shown that IFN-α drives the rapid differentiation of monocytes into highly activated DCs exhibiting the characteristics of functional APCs [25,26]. Delayed intracellular proteolysis, efficient routing to recycling vesicles, and long-term antigen presentation ability are specific IFN-DC features [27]. Given that the experiments using MUTZ-3-derived DCs require FBS, there was a concern that the vesicles in FBS would affect the dexosome-related experimental results [28]. Therefore, we used a serum-free medium in our protocol for generating MUTZ3-derived DCs with IL-4 and IFN-α to evaluate the characteristics of dexosomes (Figure 1).
M-mIFN-DC and M-mIL-4-DC exhibited significantly increased expression of maturation markers, such as CD80, CD86, CD83, HLA-ABC, and CD40, compared with MUTZ-3, confirming that MUTZ-3 differentiated into DCs (Figures 3 and S2). Moreover, M-mIFN-DC displayed a significant increase in the expression levels of mature markers (CD86 and HLA-ABC) and yield compared with those of M-mIL-4-DCs (Figures 2 and 3). Mature DCs express cell surface co-molecules necessary for antigen presentation, such as CD80, CD86, CD83, and CD40, as well as HLA-ABC and HLA-DR, with higher expression levels than those of immature DCs commitment with antigens to T cells in vitro [29,30]. We previously reported that maturation markers on DCs were clinically presumed to detect antigen-specific CTLs induced by a DC vaccine with a potential antigen-presenting ability [31]. Therefore, it could be suggested that higher levels of maturation markers on DCs provide superior antigen presentation ability, leading to immune acquisition. As shown in Figure S3, the M-mIFN-DCs tended to show higher antigen-presenting ability induction than M-mIL-4-DCs. The present results suggested that M-mIFN-DCs might have a superior antigen-presenting ability. IFN-α is known to rapidly differentiate human monocytes into DCs known as IFN-DCs, and it is effective in terms of mediating cross-priming and cross-presentation of CD8 + T cells [27]. Further in vivo studies will be required to clarify whether M-mIFN-DCs have an antigen-presenting ability superior to that of M-mIL-4-DCs. TLR (Toll-like receptor) signaling that strongly induces maturation and survival of human DCs was reported to have no influence on DC maturation in MUTZ-3 [32]. Therefore, the sensitivity of MUTZ-3-and monocyte-derived DCs would be different. The same might be elucidated for the characteristics of MUTZ-3-derived dexosomes. Further studies are needed to determine whether MUTZ-3-and monocyte-derived DC-dexosomes would be differentially characterized.
DC-derived extracellular vesicles that induce exosomes and microvesicles are part of a rapidly developing research field with applications that include diagnostics and therapeutics. DC exosomes not only carry DC-presented antigens, but also contain functional MHC-peptide complexes [16,33]. The characteristics of dexosomes released from MUTZ-3derived IL-4-and IFN-DCs have never been evaluated. The amount of dexosomes obtained from single cell secretion increased in M-mIL-4-DCs and M-mIFN-DCs, and both M-mIL-4-Dex and M-mIFN-Dex displayed typical exosomal characteristics (Figures 4 and S4a). The mean size of mature M-IFN-Dex is only slightly larger (111.5 nm) than that of M-mIL-4-Dex (100.6 nm), and the same amount of dexosome secretion could be detected from a single cell, suggesting that dexosome size and amount did not determine different functional dexosome profiles. Sara et al. reported that the phenotype of human monocyte-derived DCs is partly reflected in exosomes [34], and their results are consistent with studies in mice conducted by Segura [35]. We evaluated whether this phenomenon is observed in MUTZ-3-derived IL-4-and IFN-DC dexosomes. The ∆MFI of CD86 + or CD40 + cells were higher in M-mIFN-DC than that in M-mIL-4-DC (Figure 3), although the percentage of positive exosomes and ∆MFI in CD86 + and CD40 + were not increased in M-mIFN-Dex ( Figure 5). We could not detect the MUTZ3-derived IL-4-and IFN-DC surface molecules CD80 on the dexosomes (Figures 5 and S4b). Furthermore, the expression levels of CD80, CD86, CD83, HLA-ABC, and CD40 increased significantly with the maturation procedure in M-mIFN-DC, whereas these findings were not observed in dexosomes. Our results demonstrated that the phenotypes of MUTZ-3-derived DCs were reflected in exosomes for some but not all molecules. Although the mechanism of exosome biosynthesis is not understood entirely, it is generally accepted that exosomes are formed and developed through endocytic pathways, such as early endosomes and late endosomes, and subsequently released into the extracellular environment through exocytosis [36]. Spadaro et al. analyzed the subcellular distribution of MHC-I in IFN-DC and IL-4-DC [27]. They reported that MHC-I molecules in IFN-DCs were preferentially located in early endosomes involved in exosome formation. These reports suggest that HLA-ABC of HLA class I may be preferentially located in M-mIFN-Dex and not M-mIL-4-Dex during exosome formation. Further studies will be required to clarify whether HLA-ABC in M-mIFN-DC, compared with in M-mIL-4-DC, are preferentially located in the early endosomes involved in exosome formation. CD81, CD86, and CD40 have been reported to interact with MHC molecules in APCs to regulate their antigen-presenting capacities [37,38]. However, the expression of CD81 in mature MUTZ-3-DC dexosomes was lower than that in immature MUTZ-3-DC dexosomes, and did not change between MUTZ-3 derived IL-4-and IFN-DC dexosomes. Furthermore, CD86 and CD40 levels were unchanged between MUTZ-3-IL-4-and IFN-DC-dexosomes. Therefore, these results suggest that CD81, CD86, and CD40 on the dexosomes were not involved in the antigen-presenting ability.
Casper et al. reported that ovalbumin-loaded dexosomes were superior to microvesicles in terms of inducing antigen-specific CD8 + T cells [39]. Zitvogel et al. also demonstrated that dexosomes gave a more effective tumor suppression response than parental DCs, and that self-dexosomes could induce antigen specific CTLs, highlighting the role of dexosomal MHC class I in ex vivo processes [17]. Another study showed that human dexosomes directly loaded with MART-1 peptide possessed functional peptide-MHC complexes focusing on DCs in vitro [40]. We have shown that HLA class I on dexosomes can directly load the MART-1 peptides that stimulate CD8 + T cells ( Figure S5). Compared with M-mIL-4-Dex, M-mIFN-Dex was more efficient in inducing antigen-presenting ability immune responses (Figure 6a,b). The increased levels of HLA-ABC among HLA-associated antigens on M-mIFN-Dex could be partially associated with MART-1-specific induction of antigen presentation observed in all healthy donors. Despite no significant difference in HLA-ABC expression between M-imIFN-Dex and M-mIFN-Dex, there was a significant difference in the MART-1-specific induction of antigen-presenting ability at day 21. The difference in induction of antigen-presenting ability between M-imIFN-Dex and M-mIFN-Dex may be due to changes in the protein composition and priming abilities of dexosomes, which reflect the maturation signals received by DCs. Mature DC dexosomes contain higher levels of intercellular adhesion molecule 1 than are found in immature DC dexosomes, but bear low levels of milk fat globule-epidermal growth factor VIII (MFG-E8) compared with those in immature DC dexosomes, which may allow mature DC dexosomes to expose targeted cells more efficiently than immature DC dexosomes [35]. Therefore, the active protein composition and priming abilities of dexosomes might be reflected in the mature profiles with DCs. Further studies will be required to clarify whether M-mIFN-Dex contains ICAM-1 and MFG-E8.
The differentiated ratio of MUTZ-3-DCs indicated less than approximately 40% of the seeded MUTZ-3 cells, presuming that extensive cell death occurred in these cultures during differentiation (Figure 2b). Apoptotic cells release apoptotic bodies of 500 nm to 2 µm after they are degraded into intracellular fragments [41]. The M-IL-4-and M-IFN-DC-derived dexosome size was observed to be approximately 100 nm (Figure 4a), suggesting that the dexosomes isolated with magnetic beads were unlikely to be contained within apoptotic bodies. Furthermore, Ramesh et al. reported that apoptotic cells released exosomes. However, the underlying mechanism of apoptotic cell-derived exosome release is poorly understood and has only begun to be elucidated [42]. Further studies are warranted to determine how apoptotic cell-derived dexosomes affect the antigen-presenting capacity.
Our results revealed that treatment with IFN-α enhances the functional profiles of MUTZ-3-derived DCs, and that the highly expressed HLA-ABC-positive dexosomes they secrete have high antigen-presenting abilities. Furthermore, dexosome characteristics may enable large-scale and long-term M-IFN-Dex preservation, while maintaining efficacy. Accordingly, M-IFN-DC dexosomes might be efficient vehicles for HLA-ABC-peptide complexes in antigen presentation, enabling potent T cell activation. Based on our results, M-mIFN-DC dexosomes, with their higher antigen-presenting capability, could be useful for improving efficacy as treatment resources. Our data suggest that antigen-presenting molecules of HLA-ABC on M-IFN-Dex exist; thus, it is expected that M-IFN-Dex could function as a cancer vaccine, effectively carrying tumor antigen peptides. In addition, experimental models of MUTZ-3-derived DC dexosomes could be useful tools in the functional analyses of cell-free anticancer vaccines.

Subjects and Ethics Statement
This study was approved by the Ethics Committee of Kanazawa Medical University (approval numbers: G131 and I489). All cellular materials were obtained from healthy donors after they provided written informed consent. Peripheral blood samples were collected from healthy donors using blood collection tubes (BD Biosciences, Franklin Lakes, NJ, USA), and the centrifuged mononuclear cells were washed with PBS. All investigations were performed in accordance with the declaration of Helsinki.

MUTZ-3-Derived IL-4-and IFN-DC-Dexosome Preparation
The culture supernatant from either MUTZ-3-derived IL-4-or IFN-DCs were collected and subjected to three successive centrifugations: 300× g for 5 min to remove the cells; 1200× g for 20 min to remove the debris; and 10,000× g for 30 min to remove the large EVs. The culture supernatant was collected, and dexosomes were isolated using the MagCapture exosome isolation kit (Fujifilm Wako Pure Chemical Co., Osaka, Japan) according to the manufacturer's instructions.

Cell Surface Marker Analysis of MUTZ-3-Derived DCs
To examine the expression of cell surface markers on MUTZ-3-derived IL-4-and IFN-DCs under each condition, aliquots of 1 × 10 5 DCs were prepared in the FACSFlow TM (BD Biosciences). DCs were treated with human FcR Blocking Reagent (Miltenyi Biotec B.V. & Co. KG) for 10 min at room temperature. Each aliquot was incubated with the following mouse IgG anti-human monoclonal antibodies conjugated to fluorescein isothiocyanate (FITC), phycoerythrin (PE), and allophycocyanin (APC): FITC-conjugated anti-CD80 mAbs (BD Biosciences), PE-conjugated anti-CD86 mAbs (eBioscience, Inc., San Diego, CA, USA), APC-conjugated anti-CD83 mAbs (BioLegend, Inc., San Diego, CA, USA), FITCconjugated anti-CD40 mAbs (eBioscience, Inc.), FITC-conjugated anti-HLA-ABC mAbs (BD Biosciences), and PE-conjugated anti-HLA-DR mAbs (eBioscience, Inc.). The antibodies were added to the cells and incubated for 30 min at 4 • C in the dark. After incubation, the cells were washed with FACSFlowTM (BD Biosciences) and centrifuged at 500× g and 4 • C for 5 min. DCs were resuspended in FACSFlowTM containing 7-amino-actinomycin D (7-AAD; BD Biosciences) to detect dead cells. The live cells, defined as negative for 7-AAD, were gated on forward scatter (FSC) and side scatter (SSC) in the main population ( Figure S1). Gated cells were examined for immunophenotyping. All analyses were performed on a flow cytometer (FACS Calibur, Becton Dickinson, San Jose, CA, USA), and the data were analyzed with the Flowjo software (BD Biosciences).

Phenotypic Dexosome Characterization
To examine the surface marker expression on the dexosomes, dexosomes in the culture supernatant from MUTZ-3-derived IL-4-and IFN-DCs were isolated using a PS Capture TM Exosome Flow Cytometry Kit (Fujifilm Wako Pure Chemical Co., Osaka, Japan) according to the company's protocol, followed by a flow cytometric analysis of surface antigens after immunostaining with fluorescence-labeled antibodies. Immature and mature MUTZ-3derived IL-4-and IFN-DC dexosomes are referred to as "M-imIL-4-Dex", "M-imIFN-Dex", "M-mIL-4-Dex", and "M-mIFN-Dex", respectively.

NTA
Dexosome size measurement was conducted using the NanoSight NS300 instrument (Malvern, UK), following the manufacturer's protocol. The instrument uses a laser light source to illuminate nanoscale particles (10-1000 nm), seen as individual point-scatters moving under Brownian motion. Each sample was diluted with PBS (80-fold dilution) and analyzed in five 60-s recordings using camera level 15, detection threshold 1, the auto minimum expected particle size, and auto jump distance in NTA version 3.0, then quantified and measured.

Statistical Analysis
We used a paired two-tailed Student's t-test, one-way analysis of variance followed by Tukey's multiple comparison test, and a Wilcoxon signed-rank test to compare the differences among the groups. All statistical analyses were performed using GraphPad Prism (version 9; GraphPad Software Inc., San Francisco, CA, USA). Differences were considered statistically significant at p-values of p < 0.05.

Conclusions
In conclusion, we showed that dexosomes secreted by various cytokine-induced and MUTZ3 cell line-differentiated DCs harbor functional DC characteristics. M-mIFN-DC showed higher maturation marker expression than M-mIL-4-DC. Furthermore, the HLA-ABC expression of M-mIFN-DC was reflected in the dexosomes. Compared with M-mIL-4-Dex, M-mIFN-Dex functionally exhibited the potency of antigen presentation to CD8 + T cells in vitro. M-mIFN-Dex is expected to be useful for developing dexosome vaccines and providing further insights into dexosomes in vitro and in vivo.