Characteristics and Immunomodulating Functions of Adipose-Derived and Bone Marrow-Derived Mesenchymal Stem Cells Across Defined Human Leukocyte Antigen Barriers

Background Vascularized composite allotransplantation opens new possibilities in reconstructive transplantation such as hand or face transplants. Lifelong immunosuppression and its side-effects are the main drawbacks of this procedure. Mesenchymal stem cells (MSCs) have clinically useful immunomodulatory effects and may be able to reduce the burden of chronic immunosuppression. Herein, we assess and compare characteristics and immunomodulatory capacities of bone marrow- and adipose tissue-derived MSCs isolated from the same human individual across defined human leukocyte antigen (HLA) barriers. Materials and methods Samples of omental (o.) adipose tissue, subcutaneous (s.c.) adipose tissue, and bone marrow aspirate from 10 human organ donors were retrieved and MSCs isolated. Cells were characterized by flow cytometry and differentiated in three lineages: adipogenic, osteogenic, and chondrogenic. In mixed lymphocyte reactions, the ability of adipose-derived mesenchymal stem cells (ASCs) and bone marrow-derived mesenchymal stem cells (BMSCs) to suppress the immune response was assessed and compared within individual donors. HLA mismatched or mitogen stimulations were analyzed in co-culture with different MSC concentrations. Supernatants were analyzed for cytokine contents. Results All cell types, s.c.ASC, o.ASC, and BMSC demonstrated individual differentiation potential and cell surface markers. Immunomodulating effects were dependent on dose and cell passage. Proliferation of responder cells was most effectively suppressed by s.c.ASCs and combination with BMSC resulted in highly efficient immunomodulation. Immunomodulation was not cell contact-dependent and cells demonstrated a specific cytokine secretion. Conclusion When human ASCs and BMSCs are isolated from the same individual, both show effective immunomodulation across defined HLA barriers in vitro. We demonstrate a synergistic effect when cells from the same biologic system were combined. This cell contact-independent function underlines the potential of clinical systemic application of MSCs.

Background: Vascularized composite allotransplantation opens new possibilities in reconstructive transplantation such as hand or face transplants. Lifelong immunosuppression and its side-effects are the main drawbacks of this procedure. Mesenchymal stem cells (MSCs) have clinically useful immunomodulatory effects and may be able to reduce the burden of chronic immunosuppression. Herein, we assess and compare characteristics and immunomodulatory capacities of bone marrow-and adipose tissue-derived MSCs isolated from the same human individual across defined human leukocyte antigen (HLA) barriers.
Materials and methods: Samples of omental (o.) adipose tissue, subcutaneous (s.c.) adipose tissue, and bone marrow aspirate from 10 human organ donors were retrieved and MSCs isolated. Cells were characterized by flow cytometry and differentiated in three lineages: adipogenic, osteogenic, and chondrogenic. In mixed lymphocyte reactions, the ability of adipose-derived mesenchymal stem cells (ASCs) and bone marrow-derived mesenchymal stem cells (BMSCs) to suppress the immune response was assessed and compared within individual donors. HLA mismatched or mitogen stimulations were analyzed in co-culture with different MSC concentrations. Supernatants were analyzed for cytokine contents.
results: All cell types, s.c.ASC, o.ASC, and BMSC demonstrated individual differentiation potential and cell surface markers. Immunomodulating effects were dependent on dose and cell passage. Proliferation of responder cells was most effectively suppressed by s.c.ASCs and combination with BMSC resulted in highly efficient immunomodulation. Immunomodulation was not cell contact-dependent and cells demonstrated a specific cytokine secretion.
conclusion: When human ASCs and BMSCs are isolated from the same individual, both show effective immunomodulation across defined HLA barriers in vitro. We demonstrate a synergistic effect when cells from the same biologic system were combined. This cell contact-independent function underlines the potential of clinical systemic application of MSCs.
Keywords: vascularized composite allotransplantation, mesenchymal stem cell, immunomodulation, human leukocyte antigen, adipose-derived mesenchymal stem cells, bone marrow-derived mesenchymal stem cells, mixed lymphocyte reaction inTrODUcTiOn Vascularized composite allotransplantation (VCA) is an emerging field, expanding the armamentarium for reconstructive surgery after severe trauma, illness, or other causes of extended tissue loss. Next to the reconstitution of anatomical and psychosocial integrity (1), VCA allows functional recovery of extremities including fine motor skills, strength, and sensibility (2). So far, more than 200 VCA transplantations have been performed worldwide, and immunosuppressive strategies improved over time. The principle to replace "same with same" implies multiple important criteria to define suitable grafts for VCA recipients and their individual needs. In addition to age, sex, and skin color (3), general solid organ matching criteria such as cytomegalovirus status, patient sensitization (4), and human leukocyte antigen (HLA) typing play an important role in graft allocation and patient selection. The grade of HLA mismatch has been shown to increase the risk of acute rejection (5) in hand transplant patients and represents one of the biggest hurdles in VCA due to the shortage of donors. Acute rejection in VCA is unique in its char acteristics (6) and occurs in a much higher frequency compared with solid organ transplantation (7). Due to high immunogenic ity of the transplanted skin, acute rejection and chronic rejection remain the key immunologic challenges to solve (8). Currently, intensive immunosuppression is necessary to reduce the risk of graft rejection, but is associated with several complication inclu ding increased risks of infection (9), development of malignan cies (10), kidney impairment, and cardiovascular risks.
A novel approach to modulate the immunologic response to allografts and to potentially decrease the need of chronic systemic immunosuppression is the application of mesenchymal stem cells (MSCs) (11)(12)(13)(14). The capacity of MSCs to interact with the innate and adaptive immune response to inhibit Tcell proliferation and to upregulate regulatory Tcells (15,16) makes this cell population strong candidate for cellular therapy in VCA. MSCs have shown to express immunomodulating capacities by cell contact and paracrine effects (17,18). Clinically, these cells find increasing application to improve engraftment of hematopoietic stem cell transplantation and to decrease the risk of graftvs.host disease (19). Most promising candidates for cytotherapy in VCA are bone marrowderived mesenchymal stem cells (BMSCs) and adiposederived mesenchymal stem cells (ASCs). There are first reports of a successful application of donorderived ASCs (20) and ongoing trials with donorderived bone marrow stem cells in kidney transplant patients (21). ASCs can by isolated from subcutaneous (s.c.), adipose tissue, or from intraabdominal fat of the omentum or perirenal fat pad. Due to high yields and a simple cell isolation process (22), ASCs rep resent an appealing source of stem cells. In a transplant setting, donorderived MSC are plentiful, as a large amount of adipose tissue and bone marrow aspirate is available, in part, because all donor longbones can be used for cell isolation. Isolation of recipientderived MSCs for immunomodulation represents an alternative approach that potentially could be repeated through out the posttransplant period. In this setting, lipoaspiration, or liposuction, represents a much less invasive procedure com pared with bone marrow aspiration. Liposuction and local anes thesia for ASC isolation allows a minimally invasive procedure to make repeated cytotherapy feasible.
Multiple studies have compared suppressive capacities of ASC and BMSC (23)(24)(25), using cells from multiple tissue donors. Systemic application of MSCs showed good results in animal models (13) and promising results in human VCA (14). Recent findings from our group demonstrated a benefit of repetitive applications of ASC in a rodent VCA model (26).
The overall goal of this study was to analyze and character ize MSCs isolated from s.c.ASCs, o.ASCs, and bone marrow aspirate retrieved form the same biological system and com pare between 10 human tissue donors. This is the first study to compare immunomodulating capacities of ASCs and BMSCs derived from the same individual across defined major histo compatibility cluster (MHC) barriers from HLAtyped organ/ tissue donors. Specific aims of this study were to identify and characterize MSCs from the different compartments, to assess immunomodulating functions excluding interindividual differences, and to test these functions across defined HLA barriers.

Tissue Procurement
This study was approved by the Committee for Oversight of Research and Clinical Training Involving Descents (CORID No. 475). Tissue donors were referred after informed consent by the Center for Organ Recovery and Education (CORE). All donors were braindead cadaveric solid organ donors, HLA typed by CORE and deidentified. Inclusion criteria were 18-65 years of age male and female subjects. Exclusion criteria were the pres ence of hepatitis B, C, or HIV, sepsis/positive serology results. Adipose tissue from abdominal subcutaneous fat and omental fat (300-500 g) was excised under sterile conditions after solid

cell isolation: asc
Adiposederived mesenchymal stem cells were isolated according to a protocol previously published by Minteer et al. (27). Briefly, adipose tissue was meticulously minced and digested in a type II collagenase solution in a water bath at 37°C with gentle agitation for approximately 30 min. The digested tissue was then filtered through sterile gauze and centrifuged at 1,500 rpm for 10 min. The cell pellet was suspended in erythrocyte lysis buffer and centrifuged at 1,500 rpm for 10 min. The pellet was then sus pended in plating medium [EGM™2 BulletKit™ (Lonza), 10% fetal bovine serum, 1% penicillin/streptomycin, 1% Fungizone, and 0.001% dexamethasone] and filtered through sterile gauze to eliminate any cellular debris.

Bone Marrow-Derived Mesenchymal stem cells
Bone marrowderived mesenchymal stem cells were isolated according to a protocol previously published by Wolfe et al. (28). Briefly bone marrow aspirate was diluted in Hanks bal anced salt solution (HBSS), gently overlaid with Ficoll Paque Plus (GEHealthcare) and centrifuged at 1,800 g for 30 min.
After collection of the "buffy coat," cells were rediluted with Hanks Balanced Salt Solution (HBSS) and centrifuged again at 1,000 g for 10 min. The cell pellet was suspended in EGM™2 medium (Lonza), and plated in 175cm 2 tissue culturetreated flasks overnight. Medium was changed 24h after plating and cells were expanded up to passage 5 and partially cryopreserved at each passage.

Peripheral Blood Mononuclear cells
Briefly, whole anticoagulated blood was diluted in HBSS, gently overlaid with Ficoll Paque Plus (GEHealthcare) and centrifuged at 400 g for 40 min. After collection of the buffy coat, cells were suspended in RPMI complete medium and centrifuged at 200 g for 10 min twice. Cells were then counted manually and cryopreserved.
splenocytes Briefly, splenic tissue was minced under sterile conditions and gently squeezed through a 22 µM filter into sterile phosphate buffered saline (PBS) and centrifuged at 1,600 rpm. Erythrocyte lysis buffer was added for 2 min, 30 mL of PBS added, and cells centrifuged over Ficoll Paque Plus (GEHealthcare) at 1,600 rpm for 5 min. Cells were resuspended in RPMI, counted, and cryopreserved.

cell characterization
After isolation, cells were allowed to adhere to plastic culture dishes overnight and washed 24 h later. Media was changed every 48 h until a confluency of 70% was reached and differentiation protocols and flow cytometric analysis were initiated.

adipogenic Differentiation
Mesenchymal stem cells (s.c.ASC, o.ASC, and BMSC) derived from the same individual were plated at passage 3 at a density of 40,000 cells/cm 2 in 6well plates using EGM2 medium [EGM 2MV BulletKit (Lonza)]. After 24 h, medium was replaced with adipogenic differentiation medium [STEMPRO ® Adipogenesis Differentiation Kit (Invitrogen)] that was changed every 3-4 days over the course of 2 weeks. Control cells were cultured in regular EGM 2 medium for 2 weeks that was changed every 3-4 days.

Osteogenic Differentiation
Mesenchymal stem cells (s.c.ASC, o.ASC, and BMSC) derived from the same individual were plated at passage 3 at a density of 5,000 cells/cm 2 in 6well plates using EGM2 medium [EGM 2MV BulletKit (Lonza)]. After 24 h, medium was replaced with osteogenic differentiation media [STEMPRO ® Osteogenesis Differentiation Kit (Invitrogen)] that was changed every 3-4 days over the course of 3 weeks. Control cells were cultured in regular EGM2 medium for 3 weeks that was changed every 3-4 days respectively.

chondrogenic Differentiation
Briefly, 250,000 cells at passage 3 were suspended in 500 mL EGM2 medium aliquoted into 10 mL sterile tubes, centrifuged at 300 g for 5 min to form pellets, and incubated overnight. Medium was replaced by chondrogenic differentiation medium (Invitrogen) while control cells were cultured in incomplete differentiation medium. Tops were attached loose to allow gas exchange. Culture medium was exchanged every 3-4 days over 4 weeks.

Safranin O/Fast Green Staining
Briefly, sections were deparaffinized, hydrated with distilled water, and stained with Weigert's iron hematoxylin solution. After rinsing, samples were stained with fast green (FCF) solution for 5 min, rinsed with acetic acid and then stained with safranin O for further 5 min. After dehydrating with alcohol series and xylene, slides were mounted and coverslipped.

Alizarin Red Staining
Briefly, cells in 6well plates were fixed with 4% paraformaldehyde and stained with Mayer's hematoxylin. Alizarin red was then added (0.5 mL of 40 mM solution) and incubated for 20 min. Excessive dye was washed off and cells coverslipped and imaged with an Olympus Provis 1 microscope (Olympus America, Center Valley, PA, USA) at 20× magnification.

Adipored™ Staining
Briefly, culture medium was removed from MSCs in 96well plates and cells were washed with PBS. Each well was filled with 200 µL PBS. 5 µL Adipored was added and cells were incubated for 10 min. The readout was performed using a microplate reader (Infinite ® 200 PRO NanoQuant, Tecan). After readout, cells were imaged with brightfield microscopy.

RT-qPCR
RNA was isolated from ASC cell cultures using the RNAeasy Mini kit (Qiagen) including proteinase K digestion. cDNA was then synthesized from 500 ng RNA using the High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). Realtime quantitative PCR was then performed using the SYBR Green PCR Master Mix (Thermo Fisher Scientific) with primers listed in Table 2 including housekeeping gene YWHAZ. The relative mRNA levels were calculated using the 2 −ΔΔCt method (29). The Cts were obtained from Ct normalized to YWHAZ. The markers used for differentiation analysis are listed in Table 2.

Mixed Lymphocyte Reactions (MLRs)
Responder cells were PBMC isolated according to the protocol mentioned above. Stimulator cells were PBMCs or splenocytes irradiated with 3,000 rads. For flow cytometric analysis, responder cells were labeled prior to the assay using carboxyfluorescein succinimidyl ester (CFSE), CD3 and CD4 antibody (all BD Bioscience).
In the allogenic suppressor assays, PBMCs (2 × 10 5 cells/well) were cocultured for 7 days with either ASC or BMSC (responder cell to MSC ratio of 2:1, 4:1, 8:1, and 16:1) in triplicates in round bottom 96well plates, in the presence of irradiated stimulator cells (1 × 10 5 cells/well). Allogenic assays were analyzed using a phytohemagglutinin (PHA) based and CFSE based technique. . MSCs were positive for CD105, CD73, and CD90 and clearly negative for CD 45, CD31, CD145, CD114, and CD 33. Slight differences were seen in the markers CD34 and CD146 which did not reach significance. Slight differences were noted between individual donors but showed a similar surface pattern for all samples.
Alternatively, to assess Tcell proliferation, cells in both assays were pulsed with [3H] thymidine (1 mCi/well) for the final 8 h and [ 3 HTdR] thymidine incorporation was measured as counts per minute in a liquid scintillation counter (Perkin Elmer).
For the transwell experiment, responder cells were added to 96well plates and MSC were added in 0.1 µm transwell system (SigmaAldrich). After 5 days of stimulation, transwell baskets were removed and cells collected for flow cytometric analysis.

Luminex Multiplex Assay
Briefly, 20 µL of supernatant from MLRs, stimulated with PHA were collected after 5 days of coculture and frozen at −20°C. Supernatants were analyzed using a Cytokine 30Plex Human Luminex™ Panel according to standard protocol. Standard curves for all cytokines were used to proof data quality.

statistical analysis
Statistical analysis was performed using Graph Pad Prism 6.0 (Graph Pad Software, Inc., San Diego, CA, USA). All the data are presented as mean ± SD. p ≤ 0.05 was considered signifi cant. In MLRs, statistics were calculated using a paired ttest. Results were referred to stimulated controls, representing 100% of stimulation. RTqPCR results were analyzed using the 2 −ΔΔCt method (29). Analysis of cytokine profiling was performed using ANOVA for multiple comparisons and the Holm-Sidak method with alpha = 0.05.

resUlTs
Mscs from s.c.ascs, Omental Fat and Bone Marrow, reveal specific cytometric Markers, and Multilineage Differentiation Potential Using flow cytometry, we confirmed surface markers for adipose and BMSCs (Figure 1). Cryopreserved passaged cells (passage 3) from individual donors (n = 4) were analyzed, demonstrating a typical MSC phenotype. No significant differences between the cell types were found when pooled from individual donors (Figure 1). MSCs from all three compartments were positive for CD 105, CD73, and CD90. Slight differences were identified in the hematopoietic marker CD34 which did not reach statistical significance (s.c.ASC vs. BMSC p = 0.31; s.c.ASC vs. o.ASC p = 0.15). CD146 (mCAM) was higher in omental ASCs com pared with s.c.ASC and bone marrowderived MSCs, but again no statistical significance was detected. To assess multilineage dif ferentiation potential of the isolated cells, differentiation into the osteogenic, chondrogenic, and adipogenic lineage of s.c.ASCs, o.ASCs, and BMSCs was analyzed and compared. All samples were taken from the same tissue donor.
After 3 weeks of osteogenic differentiation culture condition, quantitative RTPCR revealed an upregulation of osteogenic markers osteocalcin and collagen type 1 alpha1 for all MSC types (n/group = 3) after differentiation culture (Figure 2A). Omental ASCs (p < 0.01) and bone marrow MSC (p < 0.05) showed a significantly higher expression of osteogenic markers compared with s.c.ASCs. All three cell types showed extracellular calcium depositions visualized by Alizarin red staining, compared with control cells. The most intense deposition of extracellular cal cium was observed in BMSC ( Figure 2D).
After 2 weeks of adipogenic culture conditions, cells mor phologically demonstrated large amounts of lipid droplets and rounder appearance suggestive for adipose differentiation ( Figure 2B). AdipoRed staining and florescence quantification all three MSC types demonstrated a significant increase of lipid inclusions compared with undifferentiated cells (Figure 2E). Notably, ASCs demonstrated a significantly higher content of lipid droplets if compared with standard cultured cells without differentiation media. Omental ASCs and subcutaneous ASCs exhibited more lipid inclusions compared with BMSCs, a differ ence that was significant RTqPCR reflected these results showing the highest increase of adipogenic markers PPARγ and CEBP/B in omental ASCs and s.c.ASCs, which were both significantly higher for PPARγ than those of BMSC (p < 0.05).
After 4 weeks of chondrogenic differentiation condition, cell conglomerates were stained using a Safranin O/Fast Green stain, showing proteoglycan depositions in all three cell types (Figure 2C), while control cells were negative. Quantitative RTqPCR showed an increase of chondrogenic markers aggrecan and collagen type 2 alpha 1 in BMSCs, ASCs and o.ASCs at similar levels. In this experiment, samples were combined in order obtain sufficient RNA for analysis; therefore, no statistical analysis could be performed (Figure 2A).

ascs and BMscs
In proliferation assays with unspecific mitogen stimulation using PHA for 5 days (n = 3), MSCs isolated from all three different com partments, demonstrated a dosedependent inhibitory effect on PBMCs ( Figure 3A). In this assay, ratios of 1:4 to 1:16 between the modulating MSCs and responding cells (PBMCs) were analyzed using a tritiated thymidine based MLR. MSC in this experiment derived from the same individual as the stimulated PBMCs (n = 4) and were therefore autologous. At all ratios, MSCs demonstrated a significant reduction in responder cell proli feration (p < 0.05), while there was no significant difference between the different MSC groups with equal doses. Subcutaneous ASCs demonstrated a trend toward more potent suppression compared with other cell types that did not reach significance (p = 0.09). Results were confirmed by a CFSE assay with flowcytrometric gating on CD3+ and CD4+ Tcell proliferation ( Figure 3B). This demonstrates an efficient inhibition [51% proliferation (SD ± 18) with addition of ASC and 56% (SD ± 19%) with BMSC] proliferation of the responding Tcells, with a ratio of 1:4 between the added MSCs and the stained responder cells.   Figure 4A). In MLRs (n = 4), allogenic ASCs isolated from s.c.ASCs and o.ASCs and BMSCs demonstrated a dosedependent suppres sion of allogenic stimulated responder cells (Figure 4A). At all tested doses ranging from 1:4 up to 1:16 between MSC and responder cells, the applied stem cells led to a significant sup pression of responder cell proliferation relative to stimulated PBMCs alone (p < 0.05). Data were confirmed using a CFSE assay, demonstrating a significant suppression of CD3+, CD4+ cells by the presence of MSC at a constant ratio of 1:4 between the responding PBMC and the added MSCs ( Figure 4B). Analysis of CD3+, CD8+ showed a suppression by all three types of MSC that did not reach significance (data not shown).

There Was no significant Difference in suppression Potential Between Donor-Derived and recipient-Derived asc
To investigate the differences in immunomodulating potency of MSC deriving from the same individual as the stimulated PBMCs and those from a fully mismatched donor, we compared both scenarios in MLRs (n = 3). In this setting, we compared the immunomodulating effects of "donorderived" (allogenic) and "recipientderived" (autologous) MSCs, were the latter were isolated from the same individual as the responding PBMCs (n = 3 for each group). In this assay, the allogeneic ASCs of a full HLA mismatched donor relative to the responder cells were used. We confirmed a significant (p < 0.05) dosedependent inhibition of PBMC proliferation for all tested ratios without

supernatants of Mlrs reveal Different concentrations of Pro-and antiinflammatory cytokines
To further describe the individual mechanisms of immunomod ulation, supernatants of PHA stimulated MLRs (n = 3) were analyzed for cytokine concentration after 5 days of coculture. Therefore, supernatants from MLR triples were collected, frozen at −20°C and analyzed later using a multiplex assay. While BMSC and s.c.ASCs demonstrated a significant increase of IL7 (each p < 0.001), no statistical significance in levels IL10, TNFα, or IFNγ could be detected (Figure 5A). s.c.ASC, o.ASC, and BMSCs at a ratio of 4:1 between responder cells (PBMC) and MSCs, respectively, showed a significant increase of IL6 (each p < 0.001) and VEGF (each p < 0.001) in the MLR superna tant compared with PBMCs alone (Figure 5B). o.ASC based MLRs had an increased the concentration of GCSF. HGF was increased in all experimental groups but did not reach statistical significance.

asc and BMsc have comparable suppressive Functions on cD3+ cD4+ T cells in a Trans-Well experiment
Recent studies (30) have shown that intravenously applied MSC get trapped in the microvasculature of the lungs and may therefore have a noncell contactdependent effect when used in vivo. To reveal whether the suppressive effects of the tested MSCs are cell contact dependent, we performed a transwell coculture experiment and compared results head to head with a regular coculture. In a CFSE assay with gating on CD3+-CD4+ cells, allogenic MSCs demonstrated comparable dosedependent suppression of responder cells without cell contact following PHA stimulation. The results were consistent for all three types of MSCs, showing almost no alterations in performance by disabling cell contact (Figures 6A,B).

a combination of asc and BMsc Deriving From the same individual results in a strong inhibition of PBMc Proliferation
Each experiment demonstrated suppressive effect with all three types of MSC. Since suppression in all assays was dose depend ent, we evaluated if the combination of bone marrowderived MSCs and ASCs isolated from the same individual would result in a synergistic effect or not. Therefore, we combined s.c.ASC and BMSC from one individual (n = 3) and cocultured them with mitogen stimulated PBMCs from fully mismatched (6/6) donors. Demonstrated is a synergistic effect with significant responder cell suppression (p < 0.05) after combination of both cell types (Figure 7C). The composition of the mix of ASC and BMSC (i.e., more ASCs or BMSCs) did not alter the suppressive potency in this assay.

DiscUssiOn
This study investigates, for the firsttime, properties of MSCs isolated from bone marrow and adipose tissue obtained from the same human individual. This allowed us to perform a head tohead comparison between the isolated cell types, eliminating interindividual variations. The availability of HLA typing for each of the 10 tissue donors provided the unique possibility for a clear definition of mismatch constellations in immunological assays.
Flow cytometric cell analysis of cell surface markers and definitions of cellular subsets is subject of continuous debate (31,32). This is the first study comparing human s.c.ASC, o.ASC, and BMSC isolated from the same individual. In our study, we confirmed findings from the literature for most common surface markers (CD105+, CD73+, CD90+, CD235a low, CD34−, CD 45−, CD31−, CD145−, CD114−) for ASCs and BMSCs. We also confirmed decrease in CD34 expression with ongoing cell culture in ASC, while CD34 expression in BMSCs from the same donor was clearly lower (33,34). Recent studies analyzing ASCs revealed further details of their cell surface proteome aiming to refine cytometric cell characterization and subgroup categori zation (35). The importance of these findings is that human MSCs expressing the discussed surface markers are capable to modify the immune response no matter if they were isolated from FigUre 7 | Head to head comparison between trans-well and standard well proliferation assay stimulated with phytohemagglutinin (PHA). Allogenic mesenchymal stem cell (MSC) deriving from the same donor were added in different ratios to carboxyfluorescein succinimidyl ester stained responder cells, resulting in a dose-dependent suppression of CD3+, CD4+ cells. Absence of cell contact (a) did not result in a loss of suppression performance compared with standard co-culture (B). When adipose-derived mesenchymal stem cells (ASCs) and bone marrow-derived mesenchymal stem cells (BMSCs) were mixed and added to co-cultures (c) at a constant ratio of 4:1 between responder peripheral blood mononuclear cells and the MSCs added, a synergistic effect of s.c.ASC and BMSC is demonstrated resulting in a significant suppression (p < 0.01), independent on the composition of the mixed MSC (more ASC vs. more BMSC). Results for mixed MSCs showed no significant difference if compared with single-cell types. adipose tissue or bone marrow nor if they are donor derived or recipient derived. Ability of MSCs to differentiate into cell lines of the meso dermal lineage is an important proof of stemness and purity of the isolated cells. In our study, we compared MSCs isolated from different tissues but from the same donor and revealed specific differentiation potentials. Adipogenic differentiated ASCs from subcutaneous adipose tissue and omentum showed an increased deposition of lipid droplets and adipogenic markers analyzed by RTqPCR compared with BMSCs (Figures 2A,B,E). These findings confirm data from other comparative studies of adipogenic differentiation potential (23), while others described a nonsignificant increase of adipogenic markers only (36,37). Chondrogenic differentiation potential was increased in BMSCs when compared with ASC and BMSC as it has been shown in previous studies (38,39) (Figures 2C,D). Osteogenic differenti ated BMSCs and o.ASC showed comparable upregulation of the osteogenic markers Col1α1 and Osteocalcin, while extracellular calcium deposition seemed slightly higher in BMSCs. These results support data from multiple groups (22), while others demonstrated nonsignificant differences (40). Interestingly, the osteogenic differentiation potential of o.ASCs was significantly higher than in s.c.ASC.
The ability to adhere to plastic in combination with multiline age differentiation and flow cytometric analysis are important baseline experiments to confirm ASC and BMSC characteristics (31). Conflicting data in the literature about differentiation potentials can be explained by heterogeneous cell isolation tech niques (41), culture conditions, and interindividual differences (25). The results presented add a valuable piece of information to the current expertise, eliminating interindividual differences caused by different donors for each tissue type.
Due to sparsity of tissue donors and additional matching criteria such as age, sex, or skin color, HLA mismatch is not an exclusion donor criterion in VCA. One way to modulate the immunologic response to allografts is the application of donor or recipientderived MSC. In this study, we show slightly superior immunomodulatory capacities of ASCs if compared with BMSCs that were isolated from the same individual after unspecific stimulation, using a mitogen. This trend was consistent using a specific full HLA mismatched allogeneic stimulation for both donor and recipientderived MSCs. Across defined HLA mismatches BMSC and ASC showed powerful suppressive function after specific stimulation with splenocytes or PBMCs deriving from the same donor as the MSCs. Regarding the immunogenicity of MSCs, we were able to confirm low immunogenic effects in our negative controls (42) of paired MSCs deriving from s.c.ASCs, o.ASCs, and BMSCs across a full HLA mismatch, with only minimal stimulation of responder cells. In a scenario of VCA, donorderived MSC could be isolated from s.c. fat tissue, omental fat, and bone mar row and be processed in a Good Clinical Practice facility. This variety of MSC could be applied systemically and, if necessary in a repetitive fashion.
The feasibility of obtaining and using adipose stromal cells and/or a combination of adipose and bone marrow cells is justi fied by established minimally invasive methods of harvesting fat tissue and marrow from living donors, as well as the already available rapid bulk isolation techniques that can be used in cadaveric donors. For living donors, adipose tissue can be easily obtained via suction assisted liposuction (43) and similar tech niques (44), while bone marrow aspirate can be isolated via iliac crest puncture (45). In cadaveric donors, bone marrow MSCs can be isolated from retrieved vertebral bodies and by long bone flushing (46), resulting in a high cell yield. Five human hand transplant patients have been treated using a cellbased protocol for immunomodulation, using stem cells isolated from vertebral bodies and long bones (14). For adipose tissue from cadaveric donors, rapid excision of subcutaneous or omental fat tissue with bulk isolation by automated machine or by manual techniques are available. Within one human cadaveric donor 500 g-1 kg of adipose tissue can be easily and rapidly col lected. A number of enzymatic and nonenzymatic automated devices are on the market (47) allowing processing under GMP conditions.
Stromal vascular fraction (48) represents an attractive source of ASC that can be obtained within the operating room with out the necessity to be further cultured in the laboratory. The heterogeneity of SVF (49) containing cell debris, endothelial cells, and bloodderived cells next to ASCs, results in a higher immunogenicity (42) and lower suppressive function if compared with cultured ASC of BMSC. Analysis of cytokine production ( Figure 6) revealed a significant upregulation of IL7 by s.c.ASC and BMSC which plays an important role in T cell inhibition, as reported earlier (50). Co cultures with o.ASC did not reach significantly higher levels of IL7. IL6 is a well described to be upregulated in cocultures of MSC and PBMC and to play an immunoregulatory role (51); our findings confirm an upregula tion with all three types of MSC tested. The primary finding was that there were slight differences in the cytokine profile of MLR between the three cell types, but markers IL6, IL10, IFN γ, TNF α, HGF, and VEGF were at comparable levels, pointing out similar mechanisms.
We were able to show no significant difference in transwell cocultures, avoiding cell contact between responder cells and modulating MSCs. This was an important concept to prove, as recent studies demonstrated that MSCs after intravenous injec tion may become trapped in the microvasculature of the lung and were not present anymore after 24 h. The same group postulated that the immunomodulating effect might be triggered through phagocytosis by monocytes (52). However, the complex in vivo mechanisms can only be partially mimicked by a transwell culture.
Limitations of this study were the in vitro design of the experi ments and therefore the relative translatability of the results. Further in vivo studies will be necessary to confirm the described results in a VCA transplant setting.
We demonstrated a synergistic effect of MSCs (Figure 7) from different compartments if added combined coculture systems. In VCA, donor bone marrow MSCs have been used in a clinical scenario for immunomodulation and showed promising results (14). Our findings are of great importance given the fact that suppression in all in vitro assays was dose dependent and the possibility to combine ASC and BMSC for immunomodulation would be a new approach to increase cell yields for cytotherapy. This was only possible since we combined MSC from different compartments, but from the same individual. In addition, recipientderived ASCs retrieved by liposuction, may be used for repetitive cytotherapy to expand the therapeutic effect, as our group proposed recently (26)