A potency and toxicology study on tolerability of cryopreserved mesenchymal stem cell product with DMSO in septic mice and nude rats

Abstract

Mesenchymal stem/stromal cell (MSC)-based therapies have emerged as a promising treatment for sepsis, supported by encouraging preclinical data. DMSO has been a widely used cryoprotectant in cell-based therapeutic products; however, whether DMSO may cause adverse effects when used for acute critical illness is unclear. This study aimed to assess the impact of DMSO on MSCs and on outcomes in animal models. Cryopreserved MSCs containing 10% DMSO were thawed, washed (Washed MSCs) to remove DMSO, or diluted to 5% DMSO (Diluted MSCs) and evaluated for key quality parameters. Diluted MSCs had significantly higher cell recovery, while viabilities for both MSCs were similar up to 24 h. At 6 h, a higher proportion of early apoptotic cells was observed in Washed MSCs. The potencies of both Washed MSCs and Diluted MSCs were equivalent in rescuing LPS-induced suppression of monocytic phagocytosis. The toxicology study showed that when 5% DMSO-containing MSCs were administered in mice with polymicrobial sepsis, no DMSO-related effects were observed on mortality, body weight loss, body temperature, or organ injury markers. Moreover, no toxicity was detected in nude rats after administration of 5% DMSO-containing MSCs. Altogether, this study demonstrated that cryopreserved MSCs with DMSO did not cause any detectable impairment in animals.

Graphical abstract

Cryopreservation represents a practical approach in manufacturing and productization of cell-based therapies. Mei and colleagues' study shows that the administration of cryopreserved MSC product containing DMSO, at a level that is equivalent to a concentration of 0.98 g/L in blood volume, is well tolerated in acutely ill, septic animals and in immunocompromised rats.

Introduction

Sepsis is a life-threatening condition, characterized by organ dysfunction resulting from a dysregulated host immune response to infection.¹ Earlier studies investigating annual sepsis incidences and sepsis-related mortalities worldwide, estimated the occurrence of 31.5 million cases, potentially leading to 5.3 million deaths each year between 1979 and 2015.² In 2017, the estimated cases increased to 48.9 million, with a reported 11.0 million annual sepsis-related deaths, which accounted for an astounding ∼20% of all global deaths that year.³ Early onset of sepsis pathology is dominated by the activation of cascades of cytokines, as well as components of the innate immune system. The end state is characterized by extensive tissue damage, organ failure, generalized immune paralysis, and potentially death. Unfortunately, there is currently no specific therapy for sepsis, and the standard of care involves prompt use of broad-spectrum antibiotics, fluid resuscitation, and management of the complications that arise due to organ failure.⁴ The rising toll of sepsis, therefore, necessitates the development of novel therapeutics that may offer greater treatment option for patients.

Mesenchymal stem/stromal cells (MSCs) represent a potentially promising new therapy for severe dysregulated inflammatory diseases, such as sepsis or acute respiratory distress syndrome (ARDS).^5,6 Our group and others have shown the robust ability of MSCs to improve outcomes in preclinical models of sepsis, including reducing pro-inflammatory mediator levels, organ dysfunction markers, vascular permeability, and the influx of inflammatory cells, while increasing pathogen clearance and overall survival.^{7,8,9,10,11,12,13,14} This putative immunomodulatory property of MSCs, along with their low immunogenicity, has catapulted them to clinical testing in sepsis patients. A phase 1 dose escalation trial of MSCs in septic shock, the most severe form of sepsis with hypotension, was previously conducted to examine the safety and tolerability of freshly cultured allogeneic bone marrow-derived MSCs (NCT02421484). The results showed freshly cultured MSCs, which did not contain a cryoprotectant such as DMSO that is typically required for a frozen/thawed MSC product, appeared safe after infusion to these vulnerable patients. However, the logistical challenges of preparing and delivering just-in-time fresh MSCs preclude the use of freshly cultured MSCs for a broader population of sepsis patients, who require an off-the-shelf product for immediate use during clinical deterioration. In contrast, a cryopreserved cell product can be thawed at bedside and administered promptly when needed.

DMSO is widely utilized as a cryoprotectant to preserve cell integrity during cryopreservation in cell-based therapeutics. Out of an abundance of caution,^15,16 particularly when used in critically ill patients, current clinical practices sometimes involve removing DMSO as a clinical product preparation step before cell administration. However, it may not always be feasible to remove the cryoprotectant reagent (usually DMSO) before treatment delivery in a typical clinical setting. For optimal therapeutic outcomes, it is critical that the MSCs being administered maintain adequate viability and potency. In this study, we assessed the quality parameters and in vitro potency of cryopreserved MSCs that were processed using two post-thaw protocols “Washed” (DMSO removed) and “Diluted” (DMSO reduced by dilution) to simulate methods of MSC preparation that typically take place in clinical trials. We also assessed the safety of DMSO-containing MSC products in an acute sepsis mouse model. As well, toxicity and biodistribution were evaluated in an immunodeficient rat (RNU nude) model.

Results

Reduced MSC viability and recovery after post-thaw washing

We evaluated whether removing or diluting DMSO to 5% volume/volume (v/v) affected MSC viability using the timeline shown in Figure 1A. First, we quantified live and dead cell numbers post-thaw and after the wash or dilution step (where thaw refers to measurements taken immediately after thawing, and 0 h denotes the time after washing or dilution; Figure 1B). Results showed a significant reduction in percent cell recovery, with a 45% drop in total cell counted in the Washed MSCs compared to 5% reduction in the Diluted MSCs (p < 0.005). This result was likely due to the loss of stressed cells (post-thaw) during the wash and centrifugation step, compared to a dilution step that may present a less disruptive method of reducing DMSO concentration compared to washing. There was no significant difference in cell viability on the Washed MSCs vs. Diluted MSCs at post-thaw and up to 24 h, measured by NucleoCounter NC-200 (Figure 1C), with negligible change in cell viabilities when both MSCs were kept at room temperature over time.

Figure 1.

Live cell recovery, cell viability, and levels of apoptotic cells in Washed MSCs (no DMSO) or Diluted MSCs (5% DMSO) over 24 h

(A) Scheme showing experimental design. (B) Percent live cells recovered at thaw, post-wash, or dilution. (C) NucleoCounter NC-200 was used to assess cell viability immediately post-thaw (Thaw) or at 0, 2, 4, 6, or 24 h after wash or dilution. (D) Representative flow cytometry plots of AV/PI-stained cells demonstrating the percentage of late apoptotic cells (AV+/PI+, Dead); early apoptotic cells (AV+, APO), and live cells (AV−/PI−, Live). (E) Summary data of live cells (AV−/PI−), (F) early apoptotic cells (AV+/PI−), and (G) late apoptotic cells (AV+/PI+). n = 3 independent experiments, with data representing mean ± SEM. Group comparisons were analyzed by two-way ANOVA followed by Sidak’s multiple comparisons test across time points. ∗p < 0.05 and ∗∗∗p < 0.005. Panel A was created in BioRender.

To examine the proportion of cells that had undergone apoptosis over the 24 h period post-wash/dilution step, flow cytometry-based annexin V (AV)/propidium iodide (PI) analysis was performed, and the representative flow cytometric plots are shown in Figure 1D. Analysis revealed that the live cell levels (AV−/PI− population) were significantly higher in the Diluted MSCs compared to the Washed MSCs (p < 0.05, Figure 1E), while Washed MSCs displayed a significantly higher population of early apoptotic cells (AV+/PI−) at the 24-h time point compared to the Diluted MSC group (p < 0.05, Figure 1F). Although Washed MSCs showed slightly higher number of cells in late apoptosis (AV+/PI+) at 24 h post the wash step, there was no statistical difference between the two groups at any time point (Figure 1G). Taken together, these data showed that the post-thaw procedure of diluting DMSO concentration can result in higher live cell recovery compared to a washing procedure. Additionally, the dilution procedure resulted in fewer MSCs undergoing apoptosis even when kept at room temperature for a prolonged period of time (i.e., 24 h), implicating better clinical utility.

Diluted MSCs and Washed MSCs demonstrated similar functionality and potency

We next evaluated whether the wash or dilution step post-thaw could affect the characteristics of MSCs, including cell morphology, proliferative capacity, metabolic activity, and potency. After washing or dilution, MSCs were seeded into cell culture flasks either immediately (termed 0 h) or after being stored at room temperature for 4 h (termed 4 h) to mimic bedside conditions. The Washed MSCs and Diluted MSCs showed similar elongated spindle-shaped morphology after 6 days of culture (Figure 2A). Confluency analysis demonstrated similar proliferative capacity between Washed MSCs and Diluted MSCs in both 0-h MSCs (Figure 2B) and 4-h MSCs (Figure 2C), suggesting that even prolonged exposure to 5% DMSO for up to 4 h did not impair the proliferation of Diluted MSCs (Figure 2C). There was no significant difference in growth fold expansion of MSCs that were plated at 0 h (23- and 25-fold, Washed MSCs vs. Diluted MSCs, respectively) or 4 h (24-fold and 24-fold, Washed MSCs vs. Diluted MSCs, respectively) after wash/dilution step (Figure 2D), with similar population doublings time for MSCs plated at 0 h (Washed MSCs: 32.0 h, Diluted MSCs: 31.0 h) or at 4 h (Washed MSCs: 31.3 h, Diluted MSCs: 31.6 h). Lactate levels measured from the collected culture media showed no significant difference in metabolic activity between Washed MSCs and Diluted MSCs in both 0-h MSCs (Figure 2E) and 4-h MSCs (Figure 2F). Finally, MSC functionality assays were conducted to evaluate whether the washing or dilution procedures may impact the potency. Based on the pathogenesis of the disease of interest (i.e., sepsis), the ability of Washed MSCs or Diluted MSCs to rescue the impaired ability of monocytes to phagocytosis bacteria was measured (Figure 3A). Lipopolysaccharide (LPS) treatment significantly reduced the ability of CD14⁺ monocytes, identified from peripheral blood mononuclear cells (PBMCs), to phagocytose bacteria (p < 0.05); Washed MSCs and Diluted MSCs were equally effective in rescuing monocytes’ phagocytic capacity, with no detectable differences (Figure 3B). Overall, the Diluted MSCs had equivalent morphology, proliferative capacity, metabolic activity, and potency to the Washed MSCs.

Figure 2.

Evaluating growth and proliferative potential of Washed MSCs (no DMSO) or Diluted MSCs (5% DMSO)

At 0 or 4 h post-wash/dilution, MSCs were seeded onto T175 culture flasks, cultured for 6 days, and monitored by Incucyte S3 live cell imaging. (A) Representative images of cell morphology from day 5 post-seeding. Scale bars, 400 μm. (B) Cell proliferation assessed as phase objective confluence captured in real time throughout culture period from 0 h after wash or dilution condition. (C) Cell proliferation assessed as phase objective confluence captured in real time throughout culture period from 4 h after wash or dilution condition. (D) Cells were harvested and counted on day 6, and fold expansion was determined by total numbers harvested over initial number seeded. (E) Daily lactate levels measured from culture medium from 0 h post-wash/dilution conditions. (F) Daily lactate levels measured from culture medium from 4 h post-wash/dilution conditions. n = 3 independent experiments, with data representing mean ± SEM. Group comparisons were analyzed by two-way ANOVA followed by Sidak’s multiple comparisons test across time points. ns = not significant between groups compared.

Figure 3.

Evaluating functional characteristics of Washed MSCs (no DMSO) or Diluted MSCs (5% DMSO)

At 0 or 4 h post-wash/dilution, MSCs were seeded in fresh growth media onto T75 culture flasks. Naive PBMCs were pre-treated with LPS, followed by co-culture with Washed MSCs (no DMSO) or Diluted MSCs (5% DMSO). PBMCs were then harvested and incubated with fluorescent-tagged E. coli and analyzed by flow cytometry to assess the ability of CD14⁺ monocyte cells to phagocytose bacteria. (A) Representative flow cytometric plot of naive PBMC with or without E. coli and LPS-treated PBMC without or with MSCs (washed or diluted). (B) Summary data to show the percentage of CD14+/E. coli-positive cells as a measure of monocytes’ phagocytic capacity. n = 3 independent experiments, with data representing mean ± SEM. Group comparisons were analyzed by one-way ANOVA with Tukey’s multiple comparisons test. ns = not significant between groups compared.

MSC product containing DMSO (Diluted MSCs) did not result in adverse outcomes in septic mice

By utilizing post-thaw processing procedures that mimic clinical scenario, we further evaluated whether the presence of DMSO in MSC product (i.e., Diluted MSCs) elicited any acute adverse effect in a polymicrobial-induced sepsis mouse model (cecal ligation and puncture [CLP]).⁷ MSC number and DMSO concentration were adjusted to result in a target DMSO blood concentration of ∼0.98 g/L, approximating the predicted blood level for patients receiving a cell dose of 3 million cells/kg (5% DMSO v/v, 5 mL of DMSO in 100 mL of final cell product) in clinical trials (i.e., NCT02421484). Mice were randomly assigned to undergo either sham or CLP procedure, followed by administration of vehicle without DSMO, vehicle with DMSO, Washed (no DMSO) MSCs, or Diluted (5% DMSO final product volume) MSCs 6 h after sham or CLP procedure (Figure 4A). As expected, CLP animals exhibited a noticeable decrease in body temperature (Figure 4B) and weight (Figure 4C) compared to sham animals at 24 h post-surgery. DMSO administration had no effect on body temperature or weight, in the presence or absence of MSCs, in septic or sham operated animals. Notably, septic animals receiving Washed MSCs or Diluted MSCs exhibited similar body temperature (Figure 4B) and weight loss (Figure 4C) at 24 h post-surgery. There were three deaths in the entire study period, two were in the eight animals in the CLP/vehicle (DMSO-free) group (25%) and one in the eight CLP/Diluted MSC-treated mice (12.5%), consistent with no effect of DMSO on survival (Table 1).

Figure 4.

Assessment of Washed MSCs (no DMSO) or Diluted MSCs (5% DMSO) in an in vivo model of polymicrobial sepsis

Mice were subjected to a CLP procedure to induce sepsis. Sham-operated mice underwent the same procedure without ligation and puncture of the cecum. Six hours following the CLP or sham operation, the sham-operated mice received either Washed MSCs (no DMSO) or Diluted MSCs (5% DMSO). The CLP mice received a vehicle solution with or without DMSO, Washed MSCs (no DMSO), or Diluted MSCs (5% DMSO), via infusion into the jugular vein. (A) Scheme showing the experimental design. (B) Body temperature after 24 h. (C) After 24 h, body weights were measured. n = 3–8 per group, with data representing mean ± SEM. Group comparisons were analyzed by one-way ANOVA with Sidak’s multiple comparisons test, ns = not significant between groups compared.

Table 1.

24-h Mortality in CLP mice treated with Washed MSCs (no DMSO) or Diluted MSCs (5% DMSO)

Injury	Sham	Sham	CLP	CLP	CLP	CLP
Treatment	Washed MSCs	Diluted MSCs	Vehicle	Vehicle + DMSO	Washed MSCs	Diluted MSCs
Survived	4	3	6	8	8	7
Dead	0	0	2	0	0	1
Total	4	3	8	8	8	8
% Mortality	0	0	25	0	0	12.5

Sham, animals received sham surgery.

∗p = 0.4667 compared between CLP/vehicle vs. CLP/vehicle + DMSO groups, p > 0.99 compared between CLP/Washed MSCs vs. CLP/Diluted MSCs groups, (chi-square and Fisher’s exact test).

Blood samples collected from animals at the study endpoint were used to assess surrogate measures of organ dysfunction, including clinical chemistry parameters such as creatinine and blood urea nitrogen (BUN) as measures of renal function, aspartate aminotransferase (AST) and alanine aminotransferase (ALT) as measures of liver function, and lactate dehydrogenase (LDH) as a measure of generalized tissue damage. There were marked increases in all markers of organ injury in animals that underwent CLP surgery, indicative of expected multiorgan injuries typically seen in sepsis (Figure 5). Septic animals that received vehicle (with or without DMSO) or MSCs with or without DMSO (Diluted MSCs or Washed MSCs, respectively) showed comparable levels of organ injury markers without significant difference (CLP/vehicle vs. CLP/vehicle + DMSO, or CLP/Washed MSCs vs. CLP/Diluted MSCs; Figure 5). Prior to testing, each blood sample was visually assessed for hemolysis (may affect LDH and AST) and icterus (may affect bilirubin) (Table S1) by personnel who were blinded for the identities of the samples. Two mice from the CLP/vehicle, three from the CLP/vehicle containing DMSO, one from the CLP/Washed MSCs, and three from the CLP/Diluted MSCs groups had minor to moderate levels of hemolysis that correlated with potentially falsely elevated AST and LDH levels (note in Table S1; individual animal data point that showed sign of hemolysis was marked with asterisk in Figure 5). Despite this, we did not observe significant difference between Washed MSCs and Diluted MSCs or between vehicle with and without DMSO. In summary, our data showed that there appeared to be no DMSO-related statistically significant differences in animals that had received vehicle or MSCs with DMSO in markers of liver, kidney, or general tissue injury in early sepsis.

Figure 5.

Levels of organ dysfunction markers in animals that had received Washed MSCs (no DMSO) or Diluted MSCs (5% DMSO)

Blood samples were collected at 24 h, and clinical biochemistry was assessed to measure organ dysfunction markers, including creatinine, BUN, AST, ALT, LDH, and total bilirubin. n = 3–8 per group, with data representing mean ± SEM. Group comparisons were analyzed by one-way ANOVA with Sidak’s multiple comparisons test. ns = not significant between groups compared. Asterisk symbol denotes specific animal blood samples that showed sign of hemolysis (determined by clinical pathologist), which artificially caused high AST and LDH levels measured by the instrument (Table S1).

MSC product containing DMSO did not result in adverse outcomes in a toxicology study in nude rats

To further investigate the safety of human MSC product containing DMSO, a toxicology study, which is often used to identify any adverse effect over a moderate period between 28 and 180 days,¹⁷ was conducted in athymic nude RNU rats. The immunodeficient rats are commonly used in a regulatory-enabling toxicology and bio-distribution study for evaluating the safety and persistence of a human cell therapy product.¹⁸ Mimicking clinical scenario, a single dose of thawed MSCs, diluted 1:1 with PLA/5% human albumin (HA) to achieve a final concentration of 2.8 × 10⁶ cells/mL with 5% DMSO (DMSO volume/total infusion volume), was administrated to each rat at a volume of 0.214 mL. This dose was calculated to correspond to a target blood DMSO concentration of 0.98 g/L. A total of six male and six female rats were used in the study. One female rat that had received vehicle (PLA/5% HA/5% DMSO) was euthanized on day 0 due to surgery-related complication and was deemed unrelated to the infused substance by the animal facility + staff (blinded by treatment group). Both female and male animals in the vehicle and cell therapy groups showed consistent weight gain over the 28 days during observation period, suggesting that no adverse changes to body weight were present after infusion (Figure 6A). Organ injury markers were measured from the blood samples collected at the end of the study (day 28) and analyzed by the contract research organization after being shipped on the same day. Visual assessment of the blood samples for hemolysis and icterus indicated that all samples were normal and, therefore, not leading to false elevation in the measurements (Table S2). Parameters of renal function (creatinine, BUN), liver function (AST, ALT, GGT [gamma-glutamyl transferase], bilirubin), and generalized tissue damage (LDH) were all within the normal range across all animals (Figure 6B). There is also no significant difference in these markers between the animals that received vehicle or Diluted MSCs (both containing 5% DMSO), although there are inconsistent, nonsignificant fluctuations in between the two groups. At study conclusion, all major organs (brain, heart, lung, liver, spleen, kidney and gonad [testis from male, or ovaries from female]) were collected and weighed for each animal. The organ weights showed no significant differences between animals that received vehicle and those that received Diluted MSCs (Table S3).

Figure 6.

Assessment of the cryopreserved, DMSO-containing MSCs in healthy athymic nude RNU rats

The MSCs were diluted immediately post-thaw with 5% HA in PLA to dilute the DMSO to 5% in the final product (final HA at 2.5%); 7- to 8-week-old nude RNU rats received cell or vehicle via jugular vein on day 0. (A) Body weight of female or male rat was monitored over 28 days of observation post-cell infusion. (B) Blood samples were collected, and clinical biochemistry was assessed to measure organ dysfunction markers, including creatinine, BUN, AST, ALT, LDH, total bilirubin, and GGT. n = 5–6 per group, with data representing mean ± SEM. Welch’s t test. Hemolysis, icterus, and lipemia visual tests indicated all normal (Table S2). As a comparator, athymic nude rats (n = 14) that received vehicle without DMSO (PLA containing 2.5% human albumin) had the following values from a separate study: creatinine, 27.14 ± 1.03 μmol/L; BUN, 7.6 ± 0.3 mmol/L; ALT, 32.57 ± 1.64 IU/L; AST, 81.29 ± 7.19 IU/L; LDH, 216.8 ± 79.5 IU/L; total bilirubin, 1.85 ± 0.1 μmol/L; and GGT, <3 IU/L. Note that these values were obtained from different contract service organization than the one described in this manuscript.

Furthermore, a formal histopathological examination of all organ sections was conducted by a board-certified pathologist, who was blinded to the identity of the slides. There were no pathological features identified in the vehicle or MSC groups, or any evidence of tumor, from all the examined organs (Figure 7; Table S4).

Figure 7.

Assessment of the cryopreserved DMSO-containing MSC product in healthy athymic nude RNU rats

MSCs were diluted immediately post-thaw with 5% HA in PLA to dilute the DMSO to 5% in the final product; 7- to 8-week-old nude RNU rats received cell or vehicle via jugular vein on day 0. After 28 days’ observation, the lung, brain, right ventricle, left ventricle, liver, kidney, spleen, and gonad were collected for histopathological examination. All images are 10× magnification; scale bars, 200 μm.

Previous preclinical studies have shown that MSCs have short persistence in the host after infusion.^19,20 As part of the regulatory safety examination, any prolonged persistence of cell therapy product could suggest potential tumorigenicity risk.²¹ In this study, we confirmed the absence of MSCs in all organs collected from the nude rats at 28 days post-infusion. qPCR-based method was utilized to amplify the human-specific DNA sequences to determine the persistence of cells injected into the nude rats. Expression of human genomic sequence was not detected from any of the organs collected, suggesting that there is no persistence of any cell at 28 days after infusion (Table S5).

Discussion

Sepsis and septic shock represent significant challenges to our health care system, and MSCs may represent a promising treatment option, with demonstrated therapeutic benefit in several animal models of sepsis.^{7,8,9,10,11,12,13,14} Multiple clinical trials have investigated their safety and efficacy in critically ill patients.^22,23,24,25 While cryopreserved MSCs provide a convenient, off-the shelf therapeutic product ideally suited for acute illness, it is critical to ensure the quality and potency of the cell product after freeze-thaw procedures. DMSO is routinely used for cryopreservation of cell therapy products.²⁶ In ex vivo studies, varying degrees of cytotoxicity for DMSO has been reported under different exposure concentrations. A 10% v/v DMSO was well tolerated by PBMCs for up to 4 h, but 24- and 48-h exposures significantly decreased cell viability.^27,28 Up to 2% v/v DMSO was well tolerated in PBMC for 48 h, but the same concentration resulted in increased cell apoptosis of mouse macrophages over the same period.^28,29 One study also reported that MSC viability decreased to 10% after 24 h of incubation in medium containing 10% DMSO at 37°C.³⁰ These observations highlight the DMSO dose, exposure time, and cell type dependence of cellular response to DMSO. Studies on the effect of DMSO on the viability and function of MSCs have yielded conflicting results. Giri et al. reported low cell viability and impaired MSC functionality post-cryopreservation and thawing in the presence of DMSO,³¹ although the concentration of DMSO used was not specified. In contrast, other studies showed preserved post-thaw viability and immunoregulatory activity in the presence of DMSO.^32,33 In our previous study, Tan et al. evaluated MSCs cryopreserved with 10% DMSO and observed a slight decline in viability over time, from 93% ± 2.6% immediately post-thaw to 81% ± 2.5% at 6 h. However, no significant difference was found between freshly cultured MSCs (without DMSO) and thawed MSCs (with 10% DMSO). Both groups demonstrated comparable efficacy in improving bacterial clearance and reducing systemic inflammation in a murine model of acute inflammatory injury.³² Similarly, Horiuchi et al. reported comparable viability between MSCs thawed after cryopreservation with 5% DMSO and freshly cultured MSCs without DMSO.³³

We have previously assessed the impact of different cryopreservation solutions on key MSC quality parameters.³⁴ In this study, we examined the effects on MSCs after post-thaw procedures (wash or dilution to remove or reduce cryoprotectant DMSO), which are frequently employed by cell manufacturers or bedside clinical researchers who infuse clinical MSC products to patients. The results revealed that both Washed MSCs (remove DMSO) and Diluted MSCs (cryopreserved MSCs product diluted 1:1 to reduce DMSO level to 5% v/v) maintained viabilities above 90% for up to 24 h at room temperature. These results provided practical solution to ease clinical translation and practice. Moreover, we did notice a significant loss of live cells followed by the thaw and wash procedure, which involved centrifugation to pellet cells, removal of DMSO-containing solution, and resuspension of cell pellets as final cell product. The significant loss of cells is most likely due to the washing procedure. There was no loss of function and potency for either the Washed MSCs or Diluted MSCs, even if left at room temperature for up to 4 h. These findings suggest that diluting cryopreserved MSC product may be a practical alternative method to reduce the DMSO level since this step can more easily be incorporated into bedside clinical practice.

In addition to the effect of DMSO on the cell therapy product itself after thaw, there are questions regarding the possible toxic effects of residual DMSO for patients receiving cryopreserved cell therapies.²⁶ DMSO is classified by the US Food and Drug Administration (FDA) as a class 3 solvent with low toxic potential at levels commonly accepted in pharmaceuticals. DMSO is also the gold standard for cryopreservation of hematopoietic stem cells and is the most commonly used cryoprotectant in hematopoietic stem cell therapy.³⁵ Health Canada has approved a cryopreserved DMSO-containing MSC product (Prochymal) for the treatment of acute graft-versus-host disease.³⁶ Clinical trials have also demonstrated the safety for the treatment of myocardial infarction (NCT00877903), type 1 diabetes (NCT00690066), and Crohn disease (NCT00482092). Additionally, a growing number of chimeric antigen receptor T (CAR-T) cell therapies have already shown promising outcomes in clinical trials. These DMSO-containing T cell products in clinical trials reported mild or no adverse events.³⁷ Among them, Kymriah, Yescarta, Tecartus, Breyanzi, and Abecma have been approved by the FDA for patient treatment.

Clinically, the most frequently reported adverse events related to the use of DMSO include nausea, vomiting, and hypo- or hypertension.³⁸ Headaches, abdominal cramps, diarrhea, bradycardia, fever, and chest tightness have also been reported,^{39,40,41,42,43} though they were typically mild and transient, requiring no intervention. The dose of DMSO also closely associates with the occurrence of adverse effects.^35,43 A clinical study in 2018, comparing cryopreserved autologous peripheral blood stem cells containing 10%, 7.5%, and 5% DMSO for transplantation in 150 patients, found that the lowest adverse reactions occurred in patients who received 5% DMSO-containing peripheral blood stem cells.⁴⁴ Their results suggest that the concentration of 5% DMSO in the infusion product could become the new standard in blood stem cell cryopreservation.^35,44 A systematic review of 109 original studies that administered DMSO to humans found that cardiovascular and respiratory adverse reactions were observed only in patients who received DMSO-containing cells intravenously.³¹ Other factors, such as the volume transfused and the temperature of freshly thawed cells, may also contribute to the occurrence of adverse reactions.³¹ One study demonstrated that acute volume expansion, electrolyte imbalance, and vagal responses to the low temperature of the infused cells were more likely responsible for cardiac arrhythmias during stem cell transfusions than the DMSO infused.⁴⁵

Severe reactions, such as life-threatening organ impairment after exposure to DMSO-containing cell product, have been reported in a few case reports. In 2000, Zenhäusern et al.¹⁵ described a patient with amyloidosis and renal failure who was well after the first fraction of autologous peripheral blood stem cells containing 5% DMSO transfusion with a volume of 455 mL on day 1, but developed arrhythmias at the end of the second autologous peripheral blood stem cell infusion with a volume of 350 mL on day 2. The two fractions of stem cells infused into patients contained a total amount of 48 g DMSO. As the patient had renal insufficiency with poor dialysability, the author suspected that the cumulative dose of 48 g DMSO administrated within 24 h and the volume expansion led to cardiac arrest. In another case report by Benekli group,¹⁶ a patient with AL amyloidosis and cardiac involvement developed acute respiratory arrest seconds after the start of autologous peripheral blood stem cell infusion. Although the exact reason was unclear, the authors suspected a vagal reaction triggered by the low temperature of the infused cell product. They also considered the release of histamine or other mediators of anaphylaxis and mast cell degranulation mediated by DMSO as a contributing factor, because this reaction occurred shortly after the infusion began, independent of the volume of DMSO infused. These results, along with other reports,^32,40,41 indicate that both the concentration and the total amount of DMSO infused into the patient can be critical safety factors in cell transfusion. Furthermore, electrolyte imbalance, concomitant medications, and underlying medical conditions—rather than DMSO—are potential factors contributing to adverse effects.^{42,46,47,48,49,50}

There are studies that have evaluated the effect of removal of DMSO prior to cell transplantation. Syme and colleagues assigned 56 breast cancer patients to receive high-dose chemotherapy, followed with treatment with either DMSO-containing autologous blood stem cells or DMSO-depleted autologous blood stem cells after washing.³⁹ Safety outcomes showed that the occurrence of nausea, vomiting, and diarrhea was less in patients who received the DMSO-depleted cells. However, the washing procedure required up to 4 h for each patient representing a logistical limitation for the treatment of acute diseases like sepsis in a timely fashion. Moreover, the washing procedure resulted in significant loss of cells, as seen in our study and others.⁵¹ Additionally, the requirement of the procedures for the complete removal of DMSO by washing the thawed cell product may be prone to unintentional loss of cells due to the rigorous washing steps. Indeed, a previous ARDS clinical trial (NCT02097641) using cryopreserved MSCs reported a wide range of MSC viabilities (30%–80%), and the authors proposed to eliminate the wash step for DMSO removal for future trials.⁵¹

Toxicology studies are part of the nonclinical study package that examines the safety of a new drug or therapy for regulatory agencies. In this study, we conducted clinical translation-relevant analyses and evaluated the acute safety of a cryopreserved MSC product containing 5% DMSO in an animal model of acute critical illness (i.e., sepsis). Additionally, we assessed the sub-chronic safety and persistence of the MSC product containing DMSO in immunodeficient RNU rats. For the acute safety, we used a CLP polymicrobial sepsis model and demonstrated that no DMSO-associated adverse effects or changes were noted in body temperature, body weight, organ injury, or animal mortality in septic animals. The target DMSO concentration administrated to animals in this study was 0.98 g/L in the blood (corresponding to 0.078 mg/g body weight), across species (mouse and rat), and mimicking the proposed cell dose planned for patients in a sepsis clinical trial (NCT02421484). To further assess the safety, the cryopreserved MSC product containing 5% DMSO was prepared using the intended clinical protocol and infused into athymic nude RNU rats. No DMSO-contained cell infusion-associated adverse effects or changes were observed in selected kidney and liver function clinical chemistry parameters, organ weight, or histopathological examination of relevant organs at 28 days after cell infusion.

In conclusion, cryopreservation represents a practical and widely used approach in MSC therapy manufacturing, which allows ease of distribution and increases clinical feasibility. This approach also makes the off-the-shelf MSC products into reality, as now the products can be promptly prepared and infused to patients with acute and critical illness, such as sepsis and ARDS. Our results add to the growing evidence showing that complete removal of DMSO from MSC products not only poses logistical challenges in clinical practice but also may significantly affect the recovery of final cell product prior to patient infusion. Our study shows that the presence of 5% DMSO in the final thawed MSC product after dilution does not cause any detectable impairment in MSC viability and potency and that cryopreserved MSCs containing 5% DMSO are well tolerated in septic animals and demonstrate safety in immunocompromised rats.

Materials and methods

Human MSC isolation and culturing

Fresh, unprocessed human bone marrow from one donor was purchased from a commercial vendor (AllCells, Alameda), while bone marrow from another donor was obtained from a healthy volunteer enrolled through The Ottawa Hospital (REB ID: 20120929-01 H). Bone marrow aspirates were diluted with PBS (Gibco) and mixed with 3% acetic acid with methylene blue (STEMCELL Technologies) to count the total nucleated cells via a hemocytometer. Cells were plated in T-175 flasks (CellBIND, Corning) pre-coated with fibronectin solution (Roche) and containing complete serum-free, xeno-free media (NutriDtem XF basal medium [Biological Industries], NutriStem XF supplement mix [Biological Industries], gentamicin reagent solution [Gibco]). All cells were maintained at 37°C, 5% (v/v) CO₂. The MSCs were lifted and transferred to HYPER Flasks (Corning), with media changes occurring every 3–4 days. Upon 70%–80% confluence, MSCs were harvested and cryopreserved at 4–6 × 10⁶ cells/mL in NutriFreez D10 (Biological Industry). A working cell bank was established by subculturing MSCs and freezing cells at 6 × 10⁶ cells/mL in PlasmaLyte A [PLA] (pH 7.4, Baxter), 5% HA (Alburex 25, Canadian Blood Services), 10% DMSO (Sigma-Aldrich) in a CryoMed controlled-rate freezer (Thermo Scientific) with long-term storage in liquid nitrogen. All experiments used thawed and cultured MSCs at passage 3.

Human MSC characterization

The capacity of the MSCs to differentiate into adipocytes, osteocytes, and chondrocytes was previously demonstrated and confirmed.³² The immune phenotype of MSCs was analyzed by flow cytometry and previously demonstrated³² detecting surface marker expression of CD73, CD90, and CD105 and lack of CD14, CD19, CD34, CD45, and HLA-DR (BD Pharmingen) expression. The MSCs were trypsinized, washed, and resuspended in cold PBS supplemented with 3% fetal bovine serum (Life Technologies) and stained with 5 or 20 μL antibody against the aforementioned markers at 4°C for 30 min according to manufacturer’s instructions. After washing, MSC surface markers were detected by flow cytometry (Attune acoustic focusing cytometer, Invitrogen) and analyzed by FlowJo 10.0 software (FlowJo, LLC) using isotype control to set gating.

Preparation of MSCs using post-thaw, wash or dilution protocol, and assessing MSC viability by flow cytometry and NucleoCounter analysis

Immediately after thawing, a portion of the cells was used for cell viability analysis (NucleoCounter NC-200, ChemoMetec) or AV (BD Biosciences) and PI (Thermo Fisher Scientific) staining followed by flow cytometry analysis. Cells were either diluted 1:1 with PLA/5% HA to reach a final concentration of 3 × 10⁶ cells/mL and 5% DMSO or washed with NutriStem XF growth medium for the complete removal of DMSO before resuspending in PLA/5% HA to reach a final concentration of 3 × 10⁶ cells/mL with 0% DMSO. Washed, DMSO-free MSCs are termed Washed MSCs, while diluted MSCs still containing DMSO are termed Diluted MSCs. Following wash or dilution, cells were kept at room temperature and analyzed at 0, 2, 4, 6 and 24 h with NC-200 cell viability analysis or flow cytometry. For flow cytometry analysis, cells were stained with AV and PI for 15 min at room temperature according to manufacturer’s instructions to determine the population of live, early, or late apoptotic MSCs via flow cytometer (Attune, Invitrogen).

Measuring proliferative and metabolic capacities of MSCs

At 0 or 4 h after wash or dilution, MSCs were seeded with fresh growth media onto T-75 flasks (Nunc, Thermo Fisher) coated with fibronectin, at a density of 1,000 cells/cm². Cell images were taken with the Incucyte S3 live cell imaging system at 8-h intervals for 6 days. The rate of cell proliferation was assessed as phase objective confluence captured in real time, using Incucyte S3 software. For cellular metabolic status, 0.5–1.0 mL of conditioned media was also collected from each flask, and lactate levels were measured using Nova Biomedical Lactate Plus Meter and test strips.

MSC potency assay

PBMCs were prepared as previously described.³⁵ Briefly PBMCs were pre-treated with LPSs (100 ng/mL) followed by co-culture with MSCs (washed MSCs or Diluted MSCs, collected at 0 or 4 h post-wash/dilution step) at a ratio of 5:1 for 24 h. PBMCs were then harvested, incubated with green fluorescent-tagged E. coli Bioparticles (pHrodo; Invitrogen), and stained with mouse anti-human CD14 conjugated with BV421 antibody (BD Biosciences). PBMC’s ability to phagocytose E. coli Bioparticles was assessed via flow cytometer (Attune, Invitrogen).

Preparation of MSCs for in vivo toxicity studies

Washed MSCs or Diluted MSCs were prepared as described above. The vehicle controls containing either 5% HA in PLA or 5% HA with 10% DMSO in PLA were thawed from frozen stocks and diluted equivalently to mimic the process described above for MSCs. The target treatment dose was calculated based on the direct DMSO concentration and the proposed human cell dose of 3 million cells/kg, with blood volume scale-down considered. Most septic mice studies in the literature showed efficacy using mouse MSCs at a dose of 0.25–1 million cells per animal. For patients, a cell suspension of 100 mL containing 5% DMSO results in 5.5 g DMSO per patient for infiltration, with DMSO density of 1.1 g/mL. Assuming an average patient weight of 70 kg and a blood volume of 80 mL/kg, the target DMSO concentration in the blood was calculated to be 0.98 g/L. This target concentration was standardized across species (human, mouse and rat) for consistency as the absolute amount of DMSO was deemed most important when assessing toxicity. In sepsis mouse model, the cells were prepared to a concentration of 2 million cells/mL, with or without 5% DMSO, administrated at 0.036 mL per mouse, and to a concentration of 2.8 million cells/mL, with 0.214 mL administration for rat, resulting in 1.98 mg of DMSO/mouse and 11.66 mg of DMSO/rat for cell administration. Assuming an average blood volume of 2 mL and a body weight of 25 g for mouse, and 12 mL and 200 g for rat, the amounts of DMSO administered align with the target DMSO concentration of 0.98 g/L in the blood, corresponding to 3 million cells/kg.

Murine model of polymicrobial sepsis

All animal experiments were conducted following ethical approval by the University of Ottawa Animal Care Committee (Ottawa, Canada) and complied with the principles and guidelines of the Canadian Council on Animal Care. Eight-week-old, healthy C57BL/6J male mice were obtained from Jackson Laboratories (Bar Harbor, ME). Animals were randomly assigned (random.org; Randomness and Integrity Services Ltd.) to receive either cecal ligation and puncture (CLP) or sham procedure. On day 0 of the study, animals were anesthetized by intraperitoneal injection of a mixture of ketamine (120 mg/kg; Ketaset, Zoetis) and xylazine (6 mg/kg; Rompun, Bayer Inc.). The average weight of the animals was 25 g (range: 21–27 g). The cecum was isolated and ligated directly below the ileocecal valve followed by puncture of the cecum by an 18G needle. Sham operation was performed by isolation of the cecum without ligation and puncture. Six hours post-CLP or sham operation, 36 μL of either the DMSO-free MSCs or MSCs with 5% DMSO were infused via the cannula inserted into the jugular vein. Animals also received DMSO-free vehicle solution or vehicle solution with 5% DMSO. Animals were randomly assigned to receive one of the treatments. Twenty-four hours after the sham or CLP procedure, the animals' weights and temperatures were taken, and the animals were sacrificed. Whole blood was collected from the inferior vena cava of each mouse using heparinized collection tubes and sent to a central laboratory (IDEXX Laboratories Canadian Corp., Toronto, Canada) for same-day clinical chemistry analysis. Cell infusion, animal wellness, and sample analyses were performed in a blinded fashion to the operator performing the analysis.

Toxicity study in healthy nude rats

Six female (at 8-week-old) and six male (at 7-week-old) healthy athymic nude RNU rats were obtained from Charles River Laboratories for this study. On day 0, the animals were anesthetized per laboratory protocol and approved animal regulations. The jugular vein was cannulated, and a single dose of MSC therapy or vehicle was infused into each animal. The animals were sacrificed after 28 days of observation. Body weight was taken at pre-treatment and weekly until the end of the study. At the conclusion of the 28-day observation period, the animals were euthanized using standard laboratory protocol. Heparinized blood samples were collected and sent to an independent contract laboratory (IDEXX Laboratory) for analysis of specific clinical chemistry parameters. A set of vascularized organ tissue samples, including the brain, heart, right lung, whole liver, spleen, both kidneys, and both gonads (testes or ovaries), were collected and weighed for histopathology examination and bio-distribution analysis. The samples were sent to a contract laboratory (IDEXX Laboratory) for histopathology examination.

qPCR determines biodistribution of MSCs in RNU nude rat

qPCR-based techniques were utilized to amplify the human-specific DNA sequences to determine the persistence of cells injected into the nude rats. Real-time qPCR was carried out using Qiagen QuantiTect SYBR Green PCR mix for 38 amplification cycles on the Bio-Rad CFX384 Real-Time System. The lower limit of detection for human genomic DNA in our assay is 0.01 ng per 50 μg of genomic DNA isolated from rat tissue (equivalent of 0.2 pg human genomic DNA per 1 μg of rat genomic DNA; 1 human cell contains 6.6 pg genomic DNA and 0.01 ng of DNA is ∼1.5 human cells). Anything below 0.01 ng is stated as less than the limit of detection (Table S5). The primer sequence (forward primer: ATGCTGATGTCTGGGTAGGGTG, reverse primer: TGAGTCAGGAGCCAGCGTATG) was designed to amplify a 141-bp fragment of human Down syndrome region of chromosome 21.

Statistical analysis

Statistical analysis was performed using GraphPad PrismV10.0 software (GraphPad Software, San Diego). Numerical data are presented as mean ± SEM unless otherwise stated. Multiple groups were analyzed by one-way ANOVA followed by Tukey’s multiple comparison test unless otherwise stated. Statistical significance was set at p < 0.05.

Data and code availability

All the data from the manuscript will be accessible upon publication.

Acknowledgments

The authors gratefully acknowledge the support of Irene Watpool and Rebecca Porteous from the ICU research team for recruitment of bone marrow volunteers. Financial support for this research work is provided by the Ontario Institute for Regenerative Medicine ([OIRM] to S.H.J.M. and L.M.) and Stem Cell Network ([SCN] to S.H.J.M. and L.M.).

Author contributions

Y.W, G.H, and A.B.P.M., conceptualization, data curation, methodology, investigation, visualization, formal analysis, and writing – original draft; J.-P.W., Y.D., M.S., Y.T., and C.W., investigation, methodology, and validation; M.F. and P.S., resources; D.J.S., supervision and writing – review & editing; L.M., funding acquisition, methodology, and writing – review & editing; L.S.-M., investigation, methodology, and writing-review & editing; S.H.J.M, conceptualization, funding acquisition, project administration, supervision, writing – review & editing.

Declaration of interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.