Inter- and Intra-donor variability in bone marrow-derived mesenchymal stromal cells: implications for clinical applications.

BACKGROUND AIMS: Mesenchymal stromal cells (MSCs) are attractive as a therapeutic modality in multiple disease conditions characterized by inflammation and vascular compromise. Logistically they are advantageous because they can be isolated from adult tissue sources, such as bone marrow (BM). The phase 2a START clinical trial determined BM-MSCs to be safe in patients with moderate-to-severe acute respiratory distress syndrome (ARDS). Herein, we examine a subset of the clinical doses of MSCs generated for the phase 2a START trial from three unique donors (1-3), where one of the donors' donated BM on two separate occasions (donor 3 and 3W). METHODS: The main objective of this study was to correlate properties of the cells from the four lots with plasma biomarkers from treated patients and relevant to ARDS outcomes. To do this we evaluated MSC donor lots for (i) post-thaw viability, (ii) growth kinetics, (iii) metabolism, (iv) surface marker expression, (v) protein expression, (vi) immunomodulatory ability and (vii) their functional effects on regulating endothelial cell permeability. RESULTS: MSC-specific marker expression and protection of thrombin-challenged endothelial barrier permeability was similar among all four donor lots. Inter and intra-donor variability was observed in all the other in vitro assays. Furthermore, patient plasma ANG-2 and protein C levels at 6 hours post-transfusion were correlated to cell viability in an inter- and intra-donor dependent manner. CONCLUSIONS: These findings highlight the potential of donor dependent (inter-) and collection dependent (intra-) effects in patient biomarker expression.

Histone deacetylase-6 modulates the effects of 4°C platelets on vascular endothelial permeability.

Platelets (PLTs) stored at 4°C exhibit equivalent or superior hemostatic function compared with 22°C PLTs, but have shorter circulation times and a decreased ability to modulate vascular permeability. These differences may be due to morphological changes and storage-induced activation. Using a proteomics-based approach, we found that 4°C-stored PLTs express decreased α-tubulin, a key PLT structural protein. PLT activation is characterized by α-tubulin deacetylation, which is regulated by histone deacetylase-6 (HDAC-6). We hypothesized that inhibition of HDAC-6 in stored PLTs will improve their ability to regulate vascular permeability through reduced activation and α-tubulin deacetylation. In an in vivo model of vascular permeability, treatment of 4°C PLTs with the HDAC-6 inhibitor tubacin enhanced the vasculoprotective properties of untreated 4°C PLTs. 4°C PLT circulation, however, was unchanged by tubacin treatment, suggesting that circulation time may not be a critical factor in determining the vasculoprotective effects of PLTs. Assessing the factor content of stored PLTs revealed that angiopoietin-1 (Ang-1) increased in 4°C PLTs over time, which was further enhanced by tubacin treatment. In addition, angiopoietin-2, an inducer of vascular leak and antagonist of Ang-1, inhibited PLT barrier protection, suggesting involvement of the Tie-2 pathway. This study demonstrates that HDAC-6 inhibition with tubacin attenuates the diminished vasculo-protective properties of 4°C PLTs, and these properties may be independent of PLT circulation time.

Freeze-dried platelets promote clot formation, attenuate endothelial cell permeability, and decrease pulmonary vascular leak in a murine model of hemorrhagic shock.

BACKGROUND: Hemorrhagic shock (HS) and trauma induce endothelial barrier compromise, inflammation, and aberrant clotting. We have shown that fresh human platelets (Plts) and Plt extracellular vesicles mitigate vascular leak in murine models of injury. Here, we investigate the potential of freeze-dried platelets (FDPlts) to attenuate pulmonary vascular permeability, decrease inflammation, and promote clotting in a murine model of HS. METHODS: Human FDPlts were characterized using in vitro assays of Plt marker expression, aggregation, coagulation, and endothelial cell permeability. An intravital model of vascular injury in the mouse cremaster muscle was used to assess the ability of FDPlts to incorporate into clots. Mouse groups subjected to controlled hemorrhage for 90 minutes were (1) lactated Ringer solution (LR), (2) FDPlts, (3) fresh human Plts, (4) murine whole blood (WB), and (5) shams (only instrumented). Hemorrhagic shock mouse endpoints included coagulation, pulmonary vascular permeability, and lung injury. RESULTS: Freeze-dried Plts expressed Plt-specific markers and retained functionality similar to fresh Plts. In in vitro assays of Plt aggregation, differences were noted. In vivo, FDPlts and Plts were found to incorporate into clots in postcapillary venules in the mouse cremaster muscle. Hemorrhagic shock mice resuscitated with LR displayed increased pulmonary vascular permeability compared with sham (sham, 686.6 ± 359.7; shock-LR, 2,637 ± 954.7; p = 0.001), and treatment with FDPlts or WB attenuated permeability compared with shock: shock-FDPlts, 1,328 ± 462.6 (p = 0.05), and shock-WB, 1,024 ± 370.5 (p = 0.0108). However, human Plts (Days 1-3) did not attenuate vascular leak in HS mice compared with shock-LR (shock-Plts, 3,601 ± 1,581; p = 0.33). CONCLUSION: FDPlts contribute to clot formation similar to fresh human Plts. FDPlts also attenuated vascular permeability in vitro and in vivo. Mouse WB resuscitation but not fresh human Plts attenuated vascular permeability after HS. These data suggest that the effect of FDPlts may be a suitable alternative to fresh Plts in modulating hemostasis and the endotheliopathy associated with injury.

The effects of human prothrombin complex concentrate on hemorrhagic shock-induced lung injury in rats: Implications for testing human blood products in rodents.

BACKGROUND: Hemorrhagic shock (HS) and trauma can result in an endotheliopathy of trauma, characterized by endothelial compromise, inflammation, and aberrant coagulation. Kcentra, a prothrombin concentrate, has been demonstrated to mitigate pulmonary vascular leak in a murine model of HS. We investigated the effects of Kcentra in a rat model of HS, to achieve physiologic endpoints of relevance. METHODS: Rats subjected to a grade intravenous splenic injury and controlled hemorrhage for 60 minutes were resuscitated with shed volumes of (1) Lactated Ringer's (LR) solution, (2) LR + 20 IU/kg Kcentra, (3) LR + 50 IU/kg Kcentra, (4) rat fresh frozen plasma (RFFP), or (5) human fresh frozen plasma (HFFP). Blood was harvested for monitoring metabolic and coagulation function. Rat lungs were evaluated for lung injury and permeability. RESULTS: Animals resuscitated with LR displayed a significant increase in pulmonary vascular permeability (sham, 407.9 ± 122.4; shock + LR, 2040 ± 1462). Resuscitation with RFFP (606.5 ± 169.3) reduced leak; however, treatment with Kcentra (HS + Kcentra [20 IU/kg]: 1792 ± 903.4, HS + Kcentra [50 IU/kg]: 1876 ± 1103), and HFFP (1450 ± 533.2) had no significant effect on permeability. Kcentra modestly altered clotting parameters. Metabolic measures, such as lactate, pH, and base deficit, were restored to baseline levels by both RFFP and HFFP, but not Kcentra or LR. CONCLUSION: Kcentra did not alter pulmonary vascular permeability, but modestly increased clotting potential in injured rats. This suggests that there may be a xenogenic reaction of human products in rats and that the effects of Kcentra on vascular stability may be distinct from its ability to modulate clotting. Our data indicate that the species chosen and utilized for in vivo preclinical testing of human derived blood products is of critical importance in determining their efficacy in animal models and is the primary impetus to communicate these results.

Is all plasma created equal? A pilot study of the effect of interdonor variability.

BACKGROUND: Clinical benefits of plasma as an adjunct for treatment of hemorrhagic shock (HS) have been well established. However, its use is not without risk. Little is understood regarding the clinical implications of plasma variability. We hypothesized there to be interdonor variability in plasma that would impact endothelial and organ function postinjury. METHODS: Pulmonary endothelial cells (ECs) were incubated with plasma from 24 random donors, and transendothelial electrical resistance was measured. Plasma units with a more or less protective effect on reducing EC permeability were selected for testing in vivo. Syndecan-1 and cytokines were measured. Mice underwent laparotomy and then HS followed by resuscitation with the selected plasma units and were compared with mice receiving no resuscitation and shams. Lung tissue was sectioned and stained for myeloperoxidase and pulmonary syndecan-1 and scored for lung histopathologic injury. RESULTS: Plasma from 24 donors revealed variability in the reversal of EC monolayer hyperpermeability; transendothelial electrical resistance for the more protective plasma was significantly higher than that for the less protective plasma (0.801 ± 0.022 vs. 0.744 ± 0.035; p = 0.002). Syndecan-1 was also markedly increased in the less protective compared with the more protective plasma (38427 ± 1257 vs. 231 ± 172 pg/mL, p < 0.001), while cytokines varied. In vivo, the more protective plasma mitigated lung histopathologic injury compared with the less protective plasma (1.56 ± 0.27 vs. 2.33 ± 0.47, respectively; p = 0.005). Similarly, myeloperoxidase was significantly reduced in the more protective compared with the less protective plasma group (2.590 ± 0.559 vs. 6.045 ± 1.885; p = 0.02). Lastly, pulmonary syndecan-1 immunostaining was significantly increased in the more protective compared with the less protective plasma group (20.909 ± 8.202 vs. 9.325 ± 3.412; p = 0.018). CONCLUSION: These data demonstrate significant interdonor variability in plasma that can adversely influence the protective effects of plasma-based resuscitation on HS-induced lung injury. This may have important implications for patient safety and clinical outcomes.

Regulation of endothelial cell permeability by platelet-derived extracellular vesicles.

BACKGROUND: Platelet (Plt)-derived extracellular vesicles (Plt-EVs) have hemostatic properties similar to Plts. In addition to hemostasis, Plts also function to stabilize the vasculature and maintain endothelial cell (EC) barrier integrity. We hypothesized that Plt-EVs would inhibit vascular EC permeability, similar to fresh Plts. To investigate this hypothesis, we used in vitro and in vivo models of vascular endothelial compromise and bleeding. METHODS: In the vitro model, Plt-EVs were isolated by ultracentrifugation and characterized for Plt markers and particle size distribution. Effects of Plts and Plt-EVs on endothelial barrier function were assessed by transendothelial electrical resistance measurements and histological analysis of endothelial junction proteins. Hemostatic potential of Plt-EVs and Plts was assessed by multiple electrode Plt aggregometry. Using an in vivo model, the effects of Plts and Plt-EVs on vascular permeability and bleeding were assessed in non-obese diabetic-severe combined immunodeficient (NOD-SCID) mice by an established Miles assay of vascular permeability and a tail snip bleeding assay. RESULTS: In the in vitro model, Plt-EVs displayed exosomal size distribution and expressed Plt-specific surface markers. Platelets and Plt-EVs decreased EC permeability and restored EC junctions after thrombin challenge. Multiplate aggregometry revealed that Plt-EVs enhanced thrombin receptor-activating peptide-mediated aggregation of whole blood, whereas Plts enhanced thrombin receptor-activating peptide-, arachidonic acid-, collagen-, and adenosine diphosphate-mediated aggregation. In the in vivo model, Plt-EVs are equivalent to Plts in attenuating vascular endothelial growth factor (VEGF)-A-induced vascular permeability and uncontrolled blood loss in a tail snip hemorrhage model. CONCLUSION: Our study is the first to report that Plt-EVs might provide a feasible product for transfusion in trauma patients to attenuate bleeding, inhibit vascular permeability, and mitigate the endotheliopathy of trauma.

Bone marrow donor selection and characterization of MSCs is critical for pre-clinical and clinical cell dose production.

BACKGROUND: Cell based therapies, such as bone marrow derived mesenchymal stem cells (BM-MSCs; also known as mesenchymal stromal cells), are currently under investigation for a number of disease applications. The current challenge facing the field is maintaining the consistency and quality of cells especially for cell dose production for pre-clinical testing and clinical trials. Here we determine how BM-donor variability and thus the derived MSCs factor into selection of the optimal primary cell lineage for cell production and testing in a pre-clinical swine model of trauma induced acute respiratory distress syndrome. METHODS: We harvested bone marrow and generated three different primary BM-MSCs from Yorkshire swine. Cells from these three donors were characterized based on (a) phenotype (morphology, differentiation capacity and flow cytometry), (b) in vitro growth kinetics and metabolic activity, and (c) functional analysis based on inhibition of lung endothelial cell permeability. RESULTS: Cells from each swine donor exhibited varied morphology, growth rate, and doubling times. All expressed the same magnitude of standard MSC cell surface markers by flow cytometry and had similar differentiation potential. Metabolic activity and growth potential at each of the passages varied between the three primary cell cultures. More importantly, the functional potency of the MSCs on inhibition of endothelial permeability was also cell donor dependent. CONCLUSION: This study suggests that for production of MSCs for cell-based therapy, it is imperative to examine donor variability and characterize derived MSCs for marker expression, growth and differentiation characteristics and testing potency in application dependent assays prior to selection of the optimal cell lineage for large scale expansion and dose production.