Human PBMC were isolated from six buffy coats generated from citrate-phosphate-dextrose treated blood (Department of Clinical Immunology, Rigshospitalet, Denmark) donated by consenting, anonymized, healthy donors according to Danish Transfusion Medical Standards, on the same day blood was donated. The buffy coats were diluted to a volume of 180 ml in phosphate-buffered saline (PBS,Gibco) and layered on LymphoPrep density gradient medium (Alere Technologies) in LeucoSep tubes with an inert porous membrane (50 ml, Greiner Bio-One) and centrifuged at 800g for 15 minutes (room temperature; RT). Subsequent centrifugations were carried out at 300g for 5 minutes (RT). The supernatant was discarded, the interphase consisting of PBMC transferred to new centrifuge tubes and washed four times with PBS. Prior to the final centrifugation the PBMC were filtered through a 100 µm cell strainer. The cells were counted (Nucleocounter NC100, ChemoMetec) and 2×10⁸ viable cells of the PBMC suspension were put aside for experimental setups exploring formulation of cryomedium, monocyte depletion, and CFSE-labelling (described below). The remaining PBMC were resuspended in 90% fetal bovine serum (FBS,Gibco) supplemented with 10% dimethyl sulfoxide (DMSO, WAK – Chemie Medical GmbH), henceforth referred to as standard cryomedium, at a final cell concentration of 1×10⁷ cells/ml. The combination of FBS and DMSO has previously been optimized (29). 1 ml cell suspension was aliquoted into cryotubes (Nunc) and subsequently placed in CoolCell (Biocision). The CoolCell was cooled for 10 minutes at 4°C before transfer to -80°C overnight for controlled freezing (-1°C/min). The following day the cryotubes were transferred to liquid nitrogen for long term storage. These six PBMC donors were applied for all experimental setups in this study.

Aside from standard cryomedium, two alternatives were tested to eliminate xenogeneic components: human serum albumin (HSA, Sigma-Aldrich), and CryoStor CS10 (Biolife Solutions). HSA was prepared at a 13.9% (w/v) concentration in RPMI 1640 (Sigma-Aldrich), sterile-filtered, aliquoted, and stored at -20°C until use. Upon thawing, DMSO was added to generate a cryomedium consisting of 12.5% HSA and 10% DMSO in RPMI 1640 based on literature (30–32). CS10 is a GMP compliant cryomedium containing 10% DMSO, which is supplied ready-to-use. The cells were cryopreserved rate-controlled in CoolCell containers as previously described.

All centrifugations in this procedure were carried out at 300g for 10 minutes at RT. From the original PBMC, 1×10⁸ cells were transferred to a new 50 ml centrifuge tube, washed in PBS, centrifuged, and resuspended in cold (4°C) degassed Magnetic-activated Cell Sorting (MACS) buffer consisting of PBS supplemented with 2 mM EDTA (Gibco) and 0.5% FBS. CD14-negative PBMC were separated from monocytes (CD14+) by MACS according to the manufacturers instruction (Miltenyi Biotec). In brief, 200 µl CD14 magnetic microbeads were added to the cell suspension, incubated for 15 minutes at 4°C, washed in MACS buffer, and loaded on a LS column in a MidiMACS separator. The column was washed thrice with MACS buffer. The CD14 negative cells were washed in PBS, and 1×10⁷ viable monocyte-depleted PBMC were cryopreserved in standard cryomedium as previously described.

PBMC were thawed in water bath (37°C) for 10 minutes and diluted 1:10 in complete pre-warmed medium, consisting of 10% FBS in RPMI-1640 with 100 units/ml penicillin and 100 µg/ml streptomycin (Gibco). The cells were incubated 1 hour (37°C, 5% CO₂) for recovery followed by centrifugation at 500g for 5 minutes at RT and labelled with 5 µM CFSE (BD Biosciences) in 2 ml PBS supplemented with of 2.5% FBS. For CFSE titration, see Supplementary Materials Figure 1 . In agreement with Quah et al. (33), we have experienced that addition of FBS during staining reduces toxicity substantially and allows for longer incubations, increasing reproducibility by reducing inter-assay and operator variation. The cells were vortexed briefly and incubated in the water bath for 10 minutes (37°C) followed by three washes with 5% FBS in PBS. The cells were resuspended in complete medium and counted. 1×10⁵ CFSE-labelled PBMC were seeded per well in a 96-well plate (round bottom, tissue culture treated, Corning) in absence or presence of 5 µg/ml PHA-L (Sigma-Aldrich) in a final volume of 200 µl.

To determine ideal CFSE-labelling of PBMC prior to cryopreservation, staining was performed as described above at three different CFSE concentrations: 2.5, 5, and 10 µM. The cells were washed once in 5% FBS in PBS followed by resuspension in complete medium and allowed to recover in incubator (37°C, 5% CO₂). The cells were cryopreserved in standard cryomedium as previously described.

ASC were manufactured as a therapeutic cell product intended for clinical use as previously described (5, 12, 14). In brief, liposuction was performed on healthy consenting donors to obtain lipoaspirate from which the stromal vascular fraction (SVF) was isolated. The use of lipoaspirate has been approved by the Regional Research Ethics Committee (Region Hovedstaden), Denmark. Cardiology Stem Cell Centre holds an Authorization of tissue establishment for the handling of human tissues and cells licensed by The Danish Patient Safety Authority and a Manufacturing and Importation Authorization issued by the Danish Medicines Agency. The SVF was loaded into a bioreactor, Quantum Cell Expansion System (Terumo BCT), with medium consisting of Minimum Essential Medium alpha (α-MEM, Gibco), 100 U/ml penicillin, 100 µg/ml streptomycin (both Gibco) and 5% human platelet lysate (hPL, Sexton Biotechnologies). After expansion the cells were harvested with TrypLE (Gibco) and cryopreserved in CS10.

For this study, the therapeutic cell product was expanded for 2 additional passages. The cells were thawed in water bath (37°) for 5 minutes and cultured in a culture treated T175 flask (Nunc) with α-MEM supplemented with 5% hPL and 100 U/ml penicillin and 100 µg/ml streptomycin. The medium was changed twice a week. For passaging and harvesting the culture medium was discarded and the flasks were washed with PBS prior to addition of 1x TrypLE (Gibco). The cells were incubated in the incubator with TrypLE for 5-10 minutes until the cells had detached the surface and the cells were collected in a 50 ml centrifuge tube (Falcon) followed by centrifugation at 300g for 5 minutes (RT). Following 2 passages, ASC were cryopreserved in CS10 at a concentration of 1×10⁶ ASC/ml.

Prior to use, one vial of each donor was thawed in water bath (37°, 5 min) and diluted in complete medium followed by centrifugation. The cells were resuspended in complete medium and counted to generate a pool of 5 ASC donors, where each donor was equally represented, at final concentration of 1×10⁵ cells/ml. See Supplementary Materials Figure 2 for comparison of single and pooled ASC. The ASC were seeded one day prior to co-culture with PBMC in a 96-well plate (round bottom, tissue culture treated), at a concentration of 2×10⁴ cells/well corresponding to 1:5 (ASC : PBMC).

PBMC were thawed in water bath (37°C, 10 min) and diluted 1:10 in complete pre-warmed medium. The cells were incubated for 1 hour (37°C, 5% CO₂) for recovery followed by centrifugation at 500g for 5 minutes at RT and labelled with CFSE.

To compare the proliferative response, PBMC were stimulated with PHA-L, concanavalin A (ConA), pokeweed mitogen (PWM) (Sigma-Aldrich), TransAct (Miltenyi Biotec), or Dynabeads (Gibco). Prior to use the Dynabeads were washed in Dynabuffer (0.1% bovine serum albumin (Sigma-Aldrich) and 0.5 M EDTA in PBS) and placed in a magnetic rack (Magnetic Particle Concentrator (MPC-6), ThermoFisher) for 1 minute. The Dynabuffer was discarded, and the beads were resuspended in complete medium to generate a concentration of 5 beads/cell which were further diluted in a 10-fold serial dilution (ranging from 5 beads/ml to 5×10^-4 beads/ml). Lyophilized PHA-L, ConA, and PWM were reconstituted in PBS and stored at -20°C until use. Mitogens were diluted in complete medium to generate concentrations from 25 to 4×10^-2 µg/ml in a 5-fold serial dilution. TransAct is a ready-to-use solution and was diluted in complete medium to generate concentrations from 1:20 to 1:2×10⁵ titer in a 10-fold serial dilution. Each CFSE-labeled PBMC donor was stimulated with the mitogens in absence or presence of ASC in triplicates and the assay elapsed for 5 days.

For MLR, the six PBMC donors were CFSE-labelled simultaneously. The cells were resuspended in complete medium and counted. Each donor was seeded in a pool of 2, 3, 4, or 5 PBMC donors leading to a combination of 15, 20, 15 and 6 unique lymphocyte reactions, respectively. The number of cells seeded from each donor depends on the number of pooled donors, however, the combined cell number was 1×10⁵ CFSE-labelled PBMC in 200 µl in absence or presence of ASC. The cells were seeded in quadruplicates, except from the experiment with the pool of 5 PBMC donors, which were seeded in octoplicates, and incubated for 8 days (37°C, 5% CO₂). An 8-day time course were selected based on optimization ( Supplementary Material Figure 3 ). The medium was partly changed 4 days after seeding by removing 100 µl supernatant and replacing with an equal amount of prewarmed complete medium.

The PBMC cryopreserved in HSA and CS10 and the monocyte-depleted PBMC, were thawed and labelled with 5 µM CFSE, and seeded alongside PBMC labelled with CFSE prior to cryopreservation in quadruplicates in 96-well plates, as previously described. The cells were cultured with complete medium in presence or absence of 5 µg/ml PHA-L and incubated for five days. The standard PBMC was included as control.

5 or 8 days after the PBMC were seeded for mitogen and MLR assay, respectively, the assays were terminated. The cells suspension of PBMC stimulated with Dynabeads were harvested and transferred into 15 ml centrifuge tubes and subsequently placed in the MPC-6 for 2 minutes. The supernatant was collected, and diluted in PBS followed by centrifugation at 500g for 5 minutes (RT). The cells were resuspended in PBS and held in non-treated polypropylene 96-well plate (round bottom, Corning). For the other assays, the cells were centrifuged at 500g for 5 minutes at RT, resuspended in PBS and transferred to the 96-well plate followed by centrifugation. In the meantime, fixable viability stain (FVS450) was prepared by dilution in PBS. 100 µl FVS450 was added to each well and incubated for 15 minutes at RT, shielded from light. The staining reaction was stopped by washing in FACS buffer consisting of 10% FBS, 1 mM EDTA, and 0.05% (w/v) sodium azide in PBS (Region Hovedstadens Apotek, Denmark). For phenotypic characterization of the lymphocytes a multicolor panel was applied (see Table 1 ). Each antibody was titrated prior to use and the ideal antibody concentration was determined by Staining Index, analyzed in FlowLogic (Inivai, version 700.1A). Prior to antibody staining, Fc-receptors were blocked with 10 µg/mL human Fc-Block for 10 minutes at RT. The concentration was based on data published by Andersen et al. (34). A mastermix of antibodies was prepared and the stained cells were incubated dark for 30 minutes at RT followed by two washes with FACS buffer. The samples were acquired for 120 seconds on a FACSLyric 3-laser 10-color CE-IVD flow cytometer (BD Bioscience) and analyzed in FlowLogic.

The gating strategy ( Figure 1 ) was common for all experimental setups, where the lymphocytes were selected by Size/Complexity, duplets were removed from the analysis, and viable cells were selected for further analysis. The reasoning for size/complexity is presented in Supplementary Figure 4 . T cells (CD3+, CD19-) were analyzed by a simple gating strategy measuring proliferation of cells beyond the second generation (>G2), or an extended analysis by model fitting. T cells were further divided into CD4 and CD8-positive cells. B cells (CD19+) were neither included for analysis of lymphocyte proliferation for mitogen or MLR assays because the proliferative response was too low to be detected.

The model fitting of proliferating cells enables calculation of six metrics according to Roederer: Fraction diluted/Proliferating (P), Precursor Frequency (PF), Expansion Index (EI), Replication Index (RI), Division Index (DI) and Proliferation Index (PI). A detailed description of each metric is reviewed by Roederer (35). In brief, P denotes the fraction of cells in the entire culture which have divided at least once where PF describes the fraction of responder cells, based on modelling of the original population. EI and RI quantify the fold expansion of entire cell culture and responder cells, respectively. DI is the average number of cell division for all cells in the culture, where PI describes the average number of cell division that responder cells have undergone.

Normality of data was confirmed by Shapiro-Wilk normality test. For pair-wise comparisons, paired or unpaired T tests were performed as appropriate, and for comparisons of more groups, one-way or two-way ANOVAs were applied. Analysis of the effect of dose-response to ASC was performed using a repeated-measures one-way ANOVA with Geisser-Greenhouse correction. For relative half maximal effective concentration (EC₅₀), minimum-maximum normalization and non-linear regression by variable slope dose-response curve was fitted. Cryomedia and monocyte-depleted conditions were compared by mixed-effects model. For multiple comparisons, p values were adjusted for false discovery rate (Q=0.05) by the two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli (36), except CFSE-labeling prior to cryopreservation, for which Dunnett’s test was used to obtain 95% confidence intervals. Significance is reported at adjusted levels of *) p < 0.05; **) p < 0.01; and ***) p < 0.001. All statistical analysis was performed in GraphPad Prism 9.3.1 (GraphPad Software, LLC).

Five mitogens were titrated over a broad range of concentrations. The observed proliferation was restricted to T cells, and no discernible cell divisions were detected on B cells. Thus, the quantification was carried out on T cells (CD3+ CD19-). Instead of focusing solely on the proliferative response elicited, a fixed dose of ASC (at a ratio of 1 ASC for every 5 PBMC; 1:5) was included to determine the effect ASC at the given concentration. To minimize donor variation, a pool of five ASC donors was used. The pool of donors was compared to the included ASC separately to rule out potential inhibiting, additive, or synergistic effects. Two doses of ASC (1:5 and 1:10) were tested, and 1:5 subsequently selected due to the stronger inhibition displayed ( Supplementary Materials Figure 2 ).

For the unspecific mitogens, PHA and ConA induced a strong proliferative response, while the pattern was less discernible for PWM which did not exceed 50% at any concentration ( Figure 2A ). The specific TransAct and Dynabeads generated a response closely resembling PHA. Addition of ASC reduced the proliferation in PHA, TransAct, and Dynabeads, but not PWM or ConA. The inhibition was especially evident at a titer of 1:200 of TransAct. By focusing on later generations of proliferation beyond two divisions ( Figure 2B ), the effect of ASC is emphasized, particularly on PHA and Dynabeads.

Based on these finding, we determined the ideal concentrations to be 5 µg/ml PHA, 5 µg/ml ConA, 1:200 TransAct, and 0.5 beads/cell Dynabeads. PWM was dropped from the study as the data did not allow for analysis, i.e., absence of proliferation and distinct CFSE peaks.

Once the concentrations of mitogens had been pinpointed, more elaborate analyses could be performed by curve fitting ( Figure 3 ). The inhibition on PHA, TransAct, and Dynabeads was significant on the parameters detailing the entirety of the culture (i.e., P, EI, and DI) and focusing on proliferating cells alike (RI and PI). ASC reduced PF on PHA-induced proliferation only. Interestingly, ASC increased some parameters on ConA-treated PBMC significantly (P, PF, and DI) whereas others remained unchanged (EI, RI, PI). Common trends were shared between CD4 and CD8 T cells subtypes ( Supplementary Materials Figure 5 ).

In order to find the ideal combination of PBMC donors to induce an allogeneic response, the PBMC were co-cultured in an MLR of 2, 3, 4, or 5 donors. The population of proliferating cells at different late generations does not allow for curve fitting ( Figure 4A ), and therefore, a simple gating strategy was applied. The proliferative response was enhanced by combining multiple PBMC donors, presented by increased number of events ( Figure 4A left). ASC potently inhibited the proliferative response independently of the number of pooled PBMC ( Figure 4A right). A significant increase in the proliferative response was observed when combining 4 or more PBMC donors ( Figure 4B ). The fraction of CD3 positive proliferating cells increased with number of donors included, highly driven by the response of CD8 positive T cells, while the fraction of CD4 positive cells was not affected ( Figure 4C ). Regardless of T cell subset, ASC greatly reduced the fraction of proliferating cells.

Besides experimental read-outs, other considerations are of importance for potency assays. Six relevant terms were identified: 1) Proliferative response: the mitogen should result in a consistent quantifiable proliferative response of lymphocytes; 2) Effect of ASC: for assessment of potency, ASC must have an effect on the proliferative response; 3) Feasibility: to minimize inter-/intra-assay variation and the risk of user error, the assay should be practical and easy to perform; 4) Supply Chain: having multiple vendors or a generic product increases the availability the mitogen; 5) Cost: the price of the mitogen per assay; 6) Specific/biological: for some applications, it may be relevant to mimic specific or biological interactions ( Figure 5 ). Three observers ranked the importance of the terms on a scale from 0-10 (least to most). The sum of the three was used to calculate a score, derived from subtracting the sum of terms “against” from the terms “for”.

Based on the ranking, we decided to focus on PHA. It should be noted we consider all tested assays, apart from PWM, valid for development of potency assays.

Rather than assessing a single concentration of ASC, we sought to elucidate how the effect of escalating cell doses affect the proliferative output. 5 independent PBMC donors were stimulated with PHA. As previously described, a pool of ASC was used to minimize variation and produce results reflecting multiple donors. With increasing concentrations of ASC, a marked decrease in proliferation was observed ( Figure 6 ). The effect was least evident on P and PF, as many lymphocytes will initiate proliferation in response to PHA regardless of ASC, however, the extent was more convincing on the remaining metrics (>G2, EI, DI, RI, and PI).

In the optimized assay, very uniform tendencies were observed for PBMC donors and experiments. It can be challenging to compare potencies, and we therefore sought to assess EC₅₀ as a tool to ease this process. PHA induces a response in a high fraction of lymphocytes and creating conditions which eliminate the response altogether require much higher concentrations of ASC, which is reflected in P and PF (EC₅₀ = 5.281 and 6.160, respectively). The actual concentration is likely higher, as we observe no clear evidence of achieving a maximum effect. To gain this insight, even higher concentrations of ASC are required (e.g., 1:1 or even 2:1). However, the remaining metrics displayed the characteristic sigmoidal curve and EC₅₀ was well within the range of ratios. The effect on EI and DI reached half at ratios of 12.712 and 10.090, respectively. Slightly higher doses were required to obtain a similar effect when focusing solely on proliferating cells in terms of RI and PI (EC₅₀ = 11.770 and 9.328, respectively).

Cryopreservation in FBS in combination with DMSO is common and has been optimized in-house previously (29). For development of potency assays, GMP-grade reagents are required. Some assays require formulations devoid of xenogeneic components. To test formulations of medium for cryopreservation (cryomedium), we included the comparators CS10 as a ready-to-use GMP-grade reagent and HSA (12.5% HSA, 10% DMSO) as a human derivative. No significant differences were observed between standard FBS/DMSO and CS10 ( Figure 7 ), but proliferation was significantly reduced in the HSA cryomedium.

Seeing that the PBMC are derived from blood donations, we sought to separate monocytes from lymphocytes in an attempt to preserve these for other uses. Interestingly, depletion of monocytes significantly impairs the proliferative response of lymphocytes ( Figure 7 ).

We tested the option to label the PBMC with CFSE prior to cryopreservation to increase feasibility. Labelling was performed with three concentrations: 2.5, 5, and 10 µM. For all metrics, there were no difference of the proliferative response compared to PBMC labelled with 5 µM CFSE post thawing (standard conditions). The suppressive effect of ASC was significantly different for PI between standard and 5 µM CFSE, but not for the other metrics ( Figures 8A, B ). In general, the proliferative response of PBMC labelled with 2.5 or 5 µM was slightly higher compared to standard, however, not statistically significant, where 10 µM CFSE prior to cryopreservation was most similar to 5 µM post thawing ( Figure 8B ).