Extracellular Vesicles Secreted by Bone Marrow- and Adipose Tissue-Derived Mesenchymal Stromal Cells Fail to Suppress Lymphocyte Proliferation

Introduction

The immunosuppressive effect of mesenchymal stromal cells (MSCs) is not ruled by one single factor, but by a sum of various soluble and membrane-bound factors. These factors contribute to MSC-mediated immunosuppression in several inflammatory diseases as well as in the setting of transplant-related graft-versus-host disease (GvHD).

Recently, it was hypothesized that MSC-derived extracellular vesicles (MSC-EVs) may contribute to the immunosuppressive action observed in MSC-based therapies. This finding would allow the development of sterile biological therapeutics without the many risks and drawbacks associated with cell-based therapies. However, the immunosuppressive effect of MSC-EVs has been actively discussed and conflicting results have been observed in vitro and in vivo. While a case report provided evidences that MSC-EVs may be used as a tool to treat therapy-refractory GvHD [1], a recent study using a mitogen-induced T-cell proliferation setting showed that MSC-EVs were significantly inferior in terms of inhibition of lymphocyte proliferation (LP) when compared to parental cells in vitro [2].

Differences between these studies are that the patient was treated with EVs isolated by polyethylene glycol precipitation from MSCs that were cultured in a medium supplemented with human platelet lysate, known to be replete with EVs, while the in vitro analysis employed EVs isolated by differential centrifugation and ultracentrifugation from cells cultivated in a medium supplemented with EV-depleted fetal bovine serum.

To further evaluate the immunosuppressive potential of purified EVs, we aimed to use a standardized LP assay (LPA) that yields high LP rates upon CD3/CD28 stimulation. To examine if the source of MSCs may have an impact on the immunosuppressive capacity, we analyzed EVs obtained from bone marrow-derived MSCs (BM-MSC-EVs) and adipose tissue-derived MSCs (AT-MSCs-EVs), isolated and purified by ultracentrifugation on a 30% sucrose cushion.

Materials and Methods

BM- and AT-MSCs (n=3) were isolated from healthy donors and characterized according to previously published protocols [3,4]. All work was approved by the ethics committee of the Medical Faculty of the Technical University Dresden (bone marrow and peripheral blood) and Medical Faculty Mannheim, Heidelberg University (adipose tissue). Subsequently, 80%–90% confluent cultures were washed thrice to remove supplement-derived EVs and an EV-free medium was added. After 3 days, EV isolation was performed and MSCs were harvested. For the EV isolation, supernatant was collected and submitted to differential centrifugation (5 min at 300 g, 20 min at 1,200 g, and 30 min at 10,000 g) and ultracentrifugation on a 30% sucrose cushion (75 min at 100,000 g) followed by a washing step (70 min at 100,000 g). Total protein of EVs was quantified using the Pierce™ BCA Protein Assay Kit (Thermo Scientific, Schwerte, Germany).

EVs were characterized by flow cytometry and shown to be positive for exosome markers (CD9, CD63, and CD81, data not shown). In addition, EVs were analyzed using the Nanoparticle Tracking Analysis technique (NTA, NS300; NanoSight Ltd., London, United Kingdom).

The immunosuppressive capacity of EVs and MSCs was assessed using a standardized LPA. LP in response to Dynabeads Human T-Activator CD3/CD28 (1 μL/well; Life Technologies, Darmstadt, Germany) was evaluated in triplicates in round-bottomed 96-well plates containing the supplemented RPMI medium, in the presence or absence of MSCs, EVs (10, 50, and 100 μg/mL), and appropriate controls (supernatant of the 10,000 g centrifugation step and ultracentrifuged supernatant). Briefly, MSCs (5×10³ cells/well) were seeded and allowed to adhere for 1 h before adding peripheral blood mononuclear cells (PBMCs) (10⁵ cells/well). EVs and controls were added at the same volume according to the EV total protein quantification. The read-out was performed using ³H-thymidine incorporation, as previously described [3].

Results and Discussion

Addition of BM- or AT-MSCs to the PBMCs stimulated with anti-CD3/CD28 beads resulted in a vigorous and significant inhibition of LP by BM- and AT-MSCs (88.1%±1.5%, 75.5%±1.5%, respectively; mean±SEM). In contrast, EVs derived from BM- or AT-MSCs failed to inhibit LP at concentrations up to 100 μg/mL (Fig. 1).

OPEN IN VIEWER

FIG. 1.

BM-MSC- and AT-MSC-derived EVs fail to inhibit LP. PBMCs were stimulated with anti-CD3/CD28-coated beads alone [stimulated PBMCs (Stim. PBMCs), used as a positive control] or in the presence of BM-MSCs or AT-MSCs (n=3, P<0.0001 and P<0.001, respectively), BM-MSC-EVs or AT-MSC-EVs. The supernatant of the 10,000 g centrifugation step and ultracentrifuged supernatant were also added to the Stim. PBMCs, but they did not exhibit any immunosuppressive effect (data not shown). The MSC:PBMC ratio used was 1:20. In the presence of MSC-EVs, dosages of 10, 50, or 100 μg/mL were used as indicated. Proliferation was assessed at day 6 of culture by measurement of ³H-thymidine incorporation. Graph depicts the level of radioactivity, given as “Corrected Counts Per Minute” (CCPM). Symbols indicate the mean of each sample and lines indicate the mean of all samples. Statistical analysis of the data was calculated using Student's t-test: P<0.001, P<0.0001, and GraphPad Prism 5.04 (GraphPad Software, Inc., San Diego, CA). AT, adipose tissue; EV, extracellular vesicle; LP, lymphocyte proliferation; MSC, mesenchymal stromal cell; PBMC, peripheral blood mononuclear cell.

Our findings using an LPA based on CD3/CD28 stimulation support the results from Conforti et al., showing impaired inhibition of LP by EVs derived from BM-MSCs in a mitogen-stimulated LPA in comparison to their parental cells [2]. In addition, we show here for the first time that AT-MSC-EVs also failed to suppress LP, although AT-MSCs exhibited a significant immunosuppressive effect. The increasing body of evidence from in vitro studies suggests that the MSC-EV immunosuppressive effect is minor when compared to their parental cells, suggesting that cell–cell contact plays an important role on the immunosuppressive potential mediated by MSCs.

It remains unclear, how the EVs exhibited contrasting results in a first case report [1]. Technically, it may hint to the fact that LPAs may not reflect the mode of action of MSC-EVs and may therefore fail to serve as a potency assay. On the other hand, it cannot be excluded that the method of isolation and, in its consequence, the purity of EV preparation may impact the potency. In fact, although most of the isolated EVs exhibit a diameter of≈90 nm, consistent with the size of exosomes, larger EVs (>100 nm) that were observed in some of our experiments, might exhibit an immunosuppressive effect when enriched to higher purities. Further studies should therefore aim to delineate the immunosuppressive capacity of EV subpopulations to validate EVs for clinical applications in the context of GvHD.

In any case, in view of patient safety, we must call attention to the fact that not only cell-based therapies but also therapies with cell-derived products, such as MSC-EVs, require stringent standardized procedures and valid potency assays for release testing before clinical use. By doing this, it will hopefully be possible to harness the advantages of a cell-free compared to a cell-based medicinal product, in terms of manufacture, sterilization, and ease of clinical application.