Long-acting PGE₂ and Lisinopril Mitigate H-ARS

Abstract

Thrombocytopenia is a major complication in hematopoietic-acute radiation syndrome (H-ARS) that increases the risk of mortality from uncontrolled hemorrhage. There is a great demand for new therapies to improve survival and mitigate bleeding in H-ARS. Thrombopoiesis requires interactions between megakaryocytes (MKs) and endothelial cells. 16, 16-dimethyl prostaglandin E₂ (dmPGE₂), a longer-acting analogue of PGE₂, promotes hematopoietic recovery after total-body irradiation (TBI), and various angiotensin-converting enzyme (ACE) inhibitors mitigate endothelial injury after radiation exposure. Here, we tested a combination therapy of dmPGE₂ and lisinopril to mitigate thrombocytopenia in murine models of H-ARS following TBI. After 7.75 Gy TBI, dmPGE₂ and lisinopril each increased survival relative to vehicle controls. Importantly, combined dmPGE₂ and lisinopril therapy enhanced survival greater than either individual agent. Studies performed after 4 Gy TBI revealed reduced numbers of marrow MKs and circulating platelets. In addition, sublethal TBI induced abnormalities both in MK maturation and in in vitro and in vivo platelet function. dmPGE₂, alone and in combination with lisinopril, improved recovery of marrow MKs and peripheral platelets. Finally, sublethal TBI transiently reduced the number of marrow Lin⁻CD45⁻CD31⁺Sca-1⁻ sinusoidal endothelial cells, while combined dmPGE₂ and lisinopril treatment, but not single-agent treatment, accelerated their recovery. Taken together, these data support the concept that combined dmPGE₂ and lisinopril therapy improves thrombocytopenia and survival by promoting recovery of the MK lineage, as well as the MK niche, in the setting of H-ARS.

INTRODUCTION

The prodigious proliferative capacity of the human hematopoietic system, responsible for the steady-state daily production of 200 billion erythrocytes, 100 billion platelets and 10 billion leukocytes (1–3) makes the hematopoietic system particularly vulnerable to radiation injury (4). H-ARS occurs after exposure of the bone marrow during total-body irradiation (TBI), resulting in pancytopenia (5). This may occur under controlled conditions, such as in the setting of radiation therapy for various cancers (6) and in preparation for hematopoietic stem cell transplants (7), or uncontrolled conditions, such as in nuclear accidents or terrorist attacks (8). Thrombocytopenia increases the risk of death from uncontrolled bleeding and related sequelae (9–10). In humans, H-ARS is induced by 2–8 Gy TBI, with the lethal dose occurring in 50% of those exposed being assessed over 60 days postirradiation (LD_50/60) (11); a dose of;4 Gy is considered the LD_50/60 for humans in the absence of supportive care (12). This compares to the LD_50/30 in mice that received 6.5–9 Gy TBI, with the considerable variation in radiosensitivity being dependent on mouse strain (13). Despite recent advancements, platelet transfusions and hematopoietic stem cell transplants remain the current standards of care for patients with production-related thrombocytopenia (14). While often effective, these interventions are fraught with risks and limitations, including the short shelf-life of platelets, infections from blood-borne pathogens compounded by the risk of sepsis in leukopenic patients, the complete reliance on donors as a platelet source, and access to medical facilities equipped to provide these therapies (15–17). Furthermore, elevated concerns regarding nuclear terrorism and increased consumption of nuclear energy worldwide necessitate new, more practical therapies for thrombocytopenia, particularly in the setting of mass radiation exposure (18).

Megakaryocyte (MK) lineage maturation is a crucial prerequisite for thrombopoiesis (19). Unlike most hematopoietic cell types, MKs are only replenished by expansion of upstream progenitors as opposed to mitotic division (20). As MKs mature, they complete repeated cycles of endomitosis, resulting in polyploidization and accumulation of cytoplasm (21). In the bone marrow, MKs interact with the sinusoidal endothelial cells, which serve as the microenvironmental niche for thrombopoiesis (22, 23). Cytoplasmic projections of MKs are extended across sinusoidal endothelium into the circulation, where they detach and remodel into circulating platelets (24). It is estimated that a single MK may produce thousands of platelets (25–27). Circulating platelets are also generated from MKs in the lung, which have lower ploidy than their marrow counterparts (28, 29). Platelet α-granules contain P-selectin, which is a cell adhesion molecule that facilitates binding between platelets, endothelial cells, and leukocytes (30). Platelet factor 4 (PF4) and von Willebrand factor (vWF), two key mediators of platelet activity (31), are also stored within platelet granules. Development of α-granules begins in the cytoplasm of MKs before they are subsequently transferred into nascent platelets (32). Taken together, maturation of the MK lineage is important both for the production of functional platelets and for maintaining normal numbers of circulating platelets.

While hematopoietic cells are primarily afflicted in H-ARS, vascular injury also contributes to pathology after radiation exposure (33). Both patients receiving radiotherapy and populations exposed during nuclear incidents are at risk of developing multiple organ dysfunction syndrome, often involving injured endothelium (34). Normally, platelets coalesce at sites of injured endothelium to stop bleeding and to facilitate endothelial cell recovery (35). This repair process is deficient in H-ARS due, in part, to thrombocytopenia (36). In addition, loss of endothelial cells may increase vascular permeability and the risk of hemorrhage, although heterogeneity in radiosensitivity has been observed between endothelium within different vascular beds (37). Despite the dependence of these cell types on one another for thrombopoiesis and hemostasis, few therapies for bleeding-related disorders consider both the MK lineage and the endothelium, which serves as the niche for MKs in the bone marrow.

Various treatments have been investigated for their ability to limit damage to healthy tissues after both medical and pathological radiation exposures. dmPGE₂ has been demonstrated to affect hematopoiesis, promoting hematopoietic stem cell survival and accelerating platelet recovery in mice after sublethal TBI (38). dmPGE₂ also improves platelet recovery and survival in mice after lethal irradiation (39). Interestingly, several ACE inhibitors have been found to improve survival and enhance recovery of endothelium after high-dose irradiation (40–42). Recently published studies have demonstrated the ability of the ACE inhibitor, lisinopril, to mitigate delayed radiation effects to multiple organs in rats when started one week postirradiation, offering a resolution to the practical challenges arising with respect to the provision of care after mass radiological exposures (43). A therapeutic effect of lisinopril was also achieved in rats when given in combination with granulocyte-colony stimulating factor, which is an FDA-approved medical countermeasure for radiation-induced leukopenia (43). Here, we investigate a dual therapy regimen using two agents with FDA approval for other indications to enhance survival and thrombopoiesis after TBI by mitigating injury both to the MK lineage and to endothelial cells.

MATERIALS AND METHODS

Mice and Irradiation

Female C57BL/6J mice (Jackson Laboratory, Bar Harbor, ME) between the ages of 8–10 weeks were housed 3–5 per cage and cared for by trained personnel in accordance with the regulations and standards established by the Federal Animal Welfare Act, the Public Health Service, the New York State Department of Health and the University of Rochester (Rochester, NY). IACUC approval was obtained for all experimental procedures. Unanesthetized mice were placed in a Plexiglas® restrainer to deliver total-body gamma irradiation using a Shephard irradiator with a ¹³⁷Cs source (San Fernando, CA). The cesium source Irradiator staff are responsible for dosimetry and yearly quality assurance. For each animal/jig/collimator combination, Gafchromic™, self-developing, dosimetry film is exposed and scanned using an Epson® Expression 10000 XL flatbed scanner (Long Beach, CA). From these films, intensity maps are acquired at prescribed distances from the collimator surface and, based on film exposures performed on the clinical 6-MV Novalis™, film red intensity is fitted to a radiation dose using a rational function of the form: y = (a + x)/(b + cx) to obtain dose maps.

Both lethal and sublethal models were used, with all mice being irradiated between 9:00–11:00 a.m. to minimize any differential responses due to circadian rhythms. Irradiated mice in the lethal model received a total dose of 7.75 Gy TBI (4.73 min at 1.64 Gy/min dose rate). All mice in this model received acidified water and antibiotic chow (Test Diet Modified Isopro, 0.124% sulfadiazine/0.025% trimethoprim) to prevent biofilm growth in the water packs and infections in accordance with supportive care protocols established for lethal doses of TBI in the setting of induced H-ARS and hematopoietic transplantation (39, 44). Irradiated mice were observed daily for morbidity and mortality, and moribund mice were humanely euthanized and the date of death recorded. Irradiated mice in the sublethal model received a total dose of 4 Gy TBI (2.44 min at 1.64 Gy/min dose rate).

In vivo Treatments

Mice in the dmPGE₂-treated groups received single intraperitoneal (IP) injections of either 2 mg/kg or 6 mg/kg body weight dmPGE₂ (Cayman Chemical, Ann Arbor, MI) dissolved in methyl acetate at 24 h after TBI in the lethal and sublethal models, respectively. Irradiated control mice received a single IP injection of methyl acetate (Sigma-Aldrich® LLC, St. Louis, MO) vehicle at the same dosage and timing as the dmPGE₂-treated groups. All mice also received 500 μl (60 mg/l in the drinking water; Cayman Chemicals; estimated to provide a dose of 28.5 mg/m²/day), was initiated at day 7 after TBI in the lethal model. The dose and schedule of lisinopril are equivalent and consistent with previously published studies in a rat model (43); however, water consumption was not quantitated to account for variation in lisinopril dosage, particularly after irradiation. Lisinopril was started at day 3 postirradiation in the sublethal model to begin therapy more than 24 h after TBI but prior to the decline in peripheral platelet counts (45). All lisinopril-treated groups continued therapy for the duration of the study and water packs containing lisinopril were replaced weekly.

Phlebotomy and Marrow Collection

Mice were euthanized by CO₂ narcosis. Peripheral blood was accessed via inferior vena cava immediately after euthanasia and transferred into K₂EDTA-coated collection tubes (BD Biosciences, Franklin Lakes, NJ) for analysis of peripheral blood counts using the HemaTrue™ Veterinary Hematology Analyzer (Heska, Loveland, CO). Femoral marrow was flushed using PB2 (DPBS, Invitrogen™, Carlsbad, CA; 0.3% BSA, Gemini Bio-Products, West Sacramento, CA; 0.68 mM CaCl₂, Sigma-Aldrich; and 0.1% glucose) with 25 μg/ml heparin. Cells were counted using a hemocytometer.

MK Lineage Analysis

To quantitate MKs and analyze their maturational features including ploidy, 10⁷ marrow cells were analyzed via imaging flow cytometry (ImageStreamX® and IDEAS® software version 6, Amnis/EMD Millipore, Seattle, WA), as described elsewhere (46, 47). The use of imaging flow cytometry facilitates distinguishing MKs from cell clumps with attached platelets. For determination of ploidy, 2N and 4N single hematopoietic cells are first defined by Draq5 staining, and these values are then applied to single CD41⁺ MKs containing Draq5 signal within the CD41 mask (46, 47).

Isolation of Platelets

Peripheral blood was collected via inferior vena cava, transferred into FACS tubes containing Tyrode’s buffer (48), treated with prostacyclin (0.05 μM) and apyrase (0.2 U/ml) to prevent activation, and centrifuged at 280g for 5 min at room temperature. After collecting platelet-rich plasma, platelets were washed with HEN buffer (10 mM HEPES, pH 6.5, 1 mM ethylenediaminetetraacetic acid, 150 mM NaCl) (49) and centrifuged at 600g for 10 min. The supernatant was aspirated, and the remaining platelets were resuspended as needed for assay preparation.

Platelet P-selectin Surface Expression

A total of 10⁶ isolated platelets were incubated with CD41-EF450, CD62P/P-selectin-FITC and Ter119-APC antibodies. After the addition of 1 mM CaCl₂, platelets were activated with 0.05 U/ml thrombin for 15 min at 37°C and subsequently inactivated with 1 U/ml hirudin (Sigma-Aldrich). Samples were fixed with 2% formaldehyde and analyzed for cell surface marker expression levels using an LSRII flow cytometer (BD Biosciences; FlowJo software, Ashland, OR). To determine if calcein green vs. calcein red-orange staining differentially affected platelet activation, isolated platelets from nonirradiated 10-week-old C57BL/6J mice (Jackson Laboratory) were stained with either 1 μM calcein green or with 1 μM calcein red-orange (Molecular Probes®, Eugene, OR) for 20 min at room temperature. After activation with thrombin, stained platelets, as well as unstained control platelets, were analyzed for P-selectin surface expression.

PF4 and vWf Protein Quantitation

A total of 10⁶ isolated platelets were activated with 0.05 U/ml thrombin for 15 min at 37°C, inactivated with 1 U/ml hirudin, and centrifuged at 2,800g for 7 min at room temperature. After collecting the supernatant containing the platelet releasate, PF4 and vWf levels were measured via ELISAs (R&D Systems™, Minneapolis, MN).

In vivo Microscopy of Platelet Aggregation

Platelets from nonirradiated and 4 Gy TBI vehicle-treated mice were isolated and washed as described above, and pooled (n = 3/group). Platelets from nonirradiated mice were incubated with 1 μM calcein red-orange for 20 min at room temperature and platelets from irradiated mice were incubated with 1 μM calcein green (Molecular Probes). Wild-type recipient mice (3–4-week-old male C57BL/6J, n= 3/group) were anesthetized and injected with 5 × 10⁷ platelets from each group. Mesenteric arterioles were visualized and incorporation of labeled platelets into host thrombi was analyzed after vascular injury with ferric chloride, as described elsewhere (50). Briefly, time-lapse imaging was collected over alternating, 30-s intervals. Still frames from the intermediate time point within each interval were selected for comparison between nonirradiated and irradiated groups. A region of interest (ROI) was applied to both the vessel containing platelets and to the surrounding tissue without platelets. The corrected background signal intensity was calculated as the area of the vessel ROI divided by the area of the background ROI multiplied by the sum intensity of the background ROI. The corrected vessel intensity was calculated as the sum intensity of the vessel ROI minus the corrected background signal intensity.

Endothelial Cell Quantitation

After euthanasia, marrow from bilateral tibiae, femura and pelvic bones was flushed with PB2, filtered and passed through a 100-micron filter. Granulocytes, B lymphocytes, T lymphocytes and erythroid cells were depleted by incubation with 0.5 μg anti-mouse biotinylated Gr-1 (clone RB6–8C5), B220 (clone RA3–6B2), CD3e (clone eBio500A2) and Ter-119 (clone TER-119) antibodies (all from eBioscience™ Inc., San Diego, CA) according to manufacturer’s protocol, respectively. A total of 10⁷ marrow cells were blocked with 10% normal rat serum for 15 min on ice. Cells were incubated with CD45-APC, CD31-FITC, Sca-1-PE and DAPI for 20 min on ice, and analyzed using an LSRII flow cytometer (BD Biosciences, FlowJo software) (51).

Platelet-Leukocyte Aggregate (PLA) Quantitation

Phlebotomy was performed and 250 μl of blood was incubated with 4 ml of 1× one-step lysis/fix solution (eBioscience) at room temperature. Samples were centrifuged at 600g for 10 min. The supernatant was collected, washed and labeled with CD41-EF450, Ly6G/C-AF488, CD45-PE, Ter119-APC and DRAQ5 antibodies. A total of 10⁶ cells were analyzed using imaging flow cytometry (ImageStream and IDEAS software, Amnis/EMD Millipore), as described elsewhere (52).

Plasma C-Reactive Protein (CRP) Quantitation

Phlebotomy was performed and whole blood was centrifuged at 2,000g for 20 min. Plasma was collected and diluted according to manufacturer’s protocol. CRP levels were measured via ELISA (R&D Systems).

Statistical Analyses

Statistical analyses were performed using GraphPad Prism version 8.1.1 (La Jolla, CA). All groups had a minimum sample size of 3. Log-rank test with pairwise comparisons was used to compare survival curves. Two-tailed t tests were used to compare two groups. One-way ANOVA was used to compare one variable among three or more groups. Two-way ANOVA was used to compare two variables among three or more groups. For all ANOVAs, Tukey’s post hoc test was used to make multiple comparisons between each group. Matched or paired measures were indicated prior to all tests. Platelet aggregation data were compared using non-linear regression and the extra-sum-of-squares F test. P < 0.05 was considered statistically significant in all experiments. All data represent the mean ± the standard error of the mean.

RESULTS

Combined dmPGE₂ and Lisinopril Treatment Enhances Survival after Lethal TBI

To determine whether the combination of two radio-mitigating agents would improve survival after TBI to a greater extent than either individual agent, we induced H-ARS in adult female C57BL/6J mice using 7.75 Gy TBI (8) (Fig. 1A). A total of 90% of irradiated, vehicle-treated mice died within 30 days of TBI (Fig. 1B), consistent with mortality occurring 14 to 30 days postirradiation in murine H-ARS due to pancytopenia-related pathology (13), as evidenced by the multifocal intracranial hemorrhages found on necropsy in 3 of 3 moribund mice at 18 days after TBI (Supplementary Fig. S1; https://doi.org/10.1667/RADE-20–00113.1.S1). dmPGE₂, given at 24 h, and lisinopril, instituted at day 7 after TBI, each mitigated mortality within the first 30 days postirradiation (Fig. 1B). Combined dmPGE₂ and lisinopril treatment improved survival to a greater extent than either individual therapy regimen (Fig. 1B). These findings persisted at 60 days (Fig. 1B). A small reduction in survival was observed in the dmPGE₂ only, lisinopril only, and combined therapy groups at this time point compared to the 30-day time point, suggesting that this late occurrence likely was not due to hematopoietic failure (8) (Fig. 1B). Taken together, these data indicate that combination therapy with dmPGE₂ and lisinopril significantly improved survival after radiation-induced hematopoietic syndrome.

FIG. 1.

dmPGE₂ and lisinopril improve survival after lethal TBI. Panel A: Diagram of experimental approach to test the ability of dmPGE₂ and lisinopril to mitigate mortality in a lethal murine model of H-ARS. Panel B: Treatment with dmPGE₂ and lisinopril each improved survival for up to 60 days after 7.75 Gy TBI compared to irradiated, vehicle controls. Combined treatment enhanced survival to a greater extent than either agent alone. n = 20–25/group. Statistical analyses were performed using log-rank test with pairwise comparisons. *P < 0.05, **P < 0.005, ***P < 0.001 and ****P < 0.0001.

Sublethal TBI Decreases MK Lineage Numbers and Alters MK Maturation

Given the evidence of hemorrhage on necropsy and the literature regarding the ability of dmPGE₂ to mitigate thrombocytopenia (38, 39), we specifically examined the MK lineage using a sublethal 4 Gy TBI model to circumvent practical limitations of lethal irradiation experiments, such as large sample size requirements and LD variability (13). Peripheral platelet counts remained comparable to nonirradiated control levels up to 5 days after 4 Gy TBI, subsequently dropping significantly by day 7 and remaining low at day 10 (Fig. 2A). These kinetics are markedly different from the rapid and severe drop in leukocytes after sublethal irradiation. In contrast, the kinetics of marrow MKs in response to 4 Gy TBI mirrored that of circulating platelets, consistent with the relative radioresistance of both cell types (Fig. 2A and B).

FIG. 2.

TBI at 4 Gy decreases peripheral platelet and marrow MK numbers, and alters marrow MK maturation. Panel A: Circulating platelets, hemoglobin and white blood cells (WBCs) are significantly decreased with unique kinetics over 10 days after 4 Gy TBI. Panel B: Bone marrow MK numbers, like circulating platelets (2A), are comparable to nonirradiated control levels early but were significantly decreased at days 7 and 10 after 4 Gy TBI. Panel C: The ploidy distribution of MKs in marrow was abnormal at day 5 after TBI. Panels D–F: Representative images of MKs by ploidy using imaging flow cytometry. Panel G: The median area of 8N and greater ploidy of bone marrow MKs were significantly decreased in vehicle-treated mice compared to nonirradiated controls at day 5 after 4 Gy TBI. Panel H: Differences in cell area were not observed for MKs of any ploidy at day 10 after TBI. n =3–8/group, five experiments. Statistical analyses were performed using one- and two-way ANOVA with multiple comparisons between each group. Dashed lines with shading represent the average values for nonirradiated controls. *P < 0.05, **P < 0.005, ***P < 0.001 and ****P < 0.0001.

Having observed effects of sublethal TBI on MK numbers, we next investigated whether MK maturation was affected by radiation. Imaging flow cytometry was used to quantitate MK cell numbers (Fig. 2C) and to characterize their morphologic features (Fig. 2D–F), as well as ploidy, of MKs in the marrow, since this analytical approach allows us to distinguish MKs from cell clumps with attached platelets (46, 47). At steady state, the most abundant ploidy group of MKs in the marrow was 16N (Fig. 2C). At 4 Gy, TBI significantly decreased the number of 16N MKs and led to an abnormal ploidy distribution at day 5 after 4 Gy TBI (Fig. 2C). In addition, MKs with a ploidy of 8N and greater in the bone marrow of irradiated mice were significantly smaller than nonirradiated controls at day 5 after 4 Gy TBI (Fig. 2G, Supplementary Fig. S2A and B; https://doi.org/10.1667/RADE-20–00113.1.S1). Differences in cell size were not observed at 10 days after TBI (Fig. 2H, Supplementary Fig. S2C and D), suggesting a transient effect of radiation on MK maturation. Taken together, these data reveal abnormalities not only in platelet and MK numbers, but also in MK maturation after sublethal irradiation.

Sublethal TBI Leads to Acute Alterations in Platelet Function

We next investigated whether TBI altered the function of platelets derived from irradiated MKs. To test platelet function in vivo, we performed intravital imaging, comparing the contribution of transfused platelets from nonirradiated and irradiated mice to host thrombi formation after vascular injury (50). Platelets from mice that received 4 Gy TBI 10 days earlier integrated at a slower rate than platelets isolated from nonirradiated control mice (Fig. 3A and B). Furthermore, platelets from irradiated mice comprised significantly smaller portions of the thrombi at the end of image collection (Fig. 3A and B). The platelets were stained either with calcein green or with calcein red-orange, which does not interfere with platelet activation (Supplementary Fig. S2E) or differentially effect incorporation into thrombi (53).

FIG. 3.

TBI at 4 Gy leads to acute platelet defects. Panel A: Representative images of in vivo platelet aggregation at day 10 after 4 Gy TBI. Panel B: Platelets from nonirradiated mice (red) contributed more to thrombi formation than platelets from irradiated, vehicle-treated mice (green) after vessel injury. Panel C: At day 10 after 4 Gy TBI, in vitro platelet P-selectin expression in response to thrombin was decreased below nonirradiated control levels. Panel D: Platelets from irradiated, vehicle-treated mice released significantly lower levels of PF4 than nonirradiated controls. Panel E: vWF levels in platelet releasate were comparable. n = 3–6/group, eight experiments. Statistical analyses were performed using two-tailed t test and two-way ANOVA with multiple comparisons between each group. Platelet aggregation data were analyzed via non-linear regression. *P < 0.05, **P < 0.005 and ****P < 0.0001.

Platelets degranulate and express extracellular P-selectin upon stimulation by agonists such as thrombin. A significantly lower percentage of P-selectin-positive platelets was observed in vitro in response to thrombin at day 10 after 4 Gy TBI compared to platelets from nonirradiated control mice (Fig. 3C). This effect was transient, with significant recovery by day 21 after sublethal irradiation. We also found that platelets from mice that were irradiated 10 days earlier released significantly lower levels of PF4, but not vWF, compared to platelets from nonirradiated mice (Fig. 3C and D). Taken together, these data indicate that sublethal TBI alters MK maturation at day 5 after TBI and leads to significant defects in the function of circulating platelets 5 days later, as evidenced by altered P-selectin surface decoration upon stimulation in vitro and reduced contribution to thrombus formation in vivo.

Delivery of dmPGE₂ as a Single Agent or in Combination with Lisinopril Accelerates Recovery of Circulating Platelets and Marrow MKs

Next, we investigated whether dmPGE₂, lisinopril, or combined dmPGE₂ and lisinopril therapy mitigates injury to the MK lineage in a sublethal irradiation model (Fig. 4A). Treatment with dmPGE₂, both alone and in combination with lisinopril, elevated circulating platelet numbers above vehicle control levels and back to nonirradiated control levels at day 10 postirradiation, while lisinopril did not have any effect on platelet numbers at this time point (Fig. 4B). Consistent with these findings, dmPGE₂, both alone and in combination with lisinopril, also significantly enhanced recovery of MKs in the marrow compared to vehicle controls (Fig. 4C). Interestingly, we did not observe an effect of dmPGE₂ or lisinopril on the recovery of hemoglobin levels or white blood cell numbers at day 10 after 4 Gy TBI (Supplementary Fig. S2F and G; https://doi.org/10.1667/RADE-20–00113.1.S1 ).

FIG. 4.

dmPGE₂, alone and in combination with lisinopril, improves MK and platelet recovery at day 10 after 4 Gy TBI. Panel A: Diagram of experimental approach to test the ability of dmPGE₂ and lisinopril to mitigate injury in a sublethal murine model of H-ARS. Panel B: Treatment with dmPGE₂ as a single agent and in combination with lisinopril increased peripheral platelet counts relative to vehicle controls, back to normal levels. Panel C: dmPGE₂, both alone and in combination with lisinopril, significantly increased marrow MK numbers above vehicle control levels. Panel D: The ploidy distribution of MKs in marrow was abnormal at day 10 after 4 Gy TBI. Both regimens significantly increased the number of 2N MKs relative to vehicle. Combined therapy also significantly increased MK numbers at other ploidies. Panel E: Platelet P-selectin surface expression was not affected by any treatment regimen. n = 4–8/group, three experiments. Statistical analyses were performed using one- and two-way ANOVA with multiple comparisons between each group. Dashed lines with shading represent the average values for nonirradiated controls. *P < 0.05, **P < 0.005, ***P < 0.001 and ****P < 0.0001.

To determine whether the effective dmPGE₂-containing treatment regimens improved MK maturation, we analyzed the ploidy distribution of MKs in the marrow 10 days after TBI. All ploidy groups were expanded compared to vehicle-treated controls, although the ploidy distribution was not normalized (Fig. 4D). While both dmPGE₂ alone and combined with lisinopril significantly expanded 2N ploidy MKs, the combined therapy regimen significantly expanded most higher ploidy MKs, as well (Fig. 4D). Since therapy with dmPGE₂ as a single-agent or in combination with lisinopril expanded MKs and altered their maturation, we investigated if this was associated with an improvement not only in platelet numbers, but also in platelet function. However, no significant improvement in platelet P-selectin surface expression was noted upon thrombin stimulation at day 10 after 4 Gy TBI (Fig. 4E). Taken together, these data support the notion that dmPGE₂-based mitigation therapies promote recovery of MK and platelet numbers after TBI, but likely have little effect on the transient alterations observed in platelet function.

Combined dmPGE₂ and Lisinopril Treatment Mitigates Injury to Marrow Endothelium after Sublethal TBI

Given the necessity of sinusoidal endothelium for thrombopoiesis (22, 23), we next examined the effects of 4 Gy TBI on sinusoidal endothelial cell numbers in the marrow. Irradiation led to a severe drop in sinusoidal endothelial cell numbers by day 7, with recovery to normal levels by 21 days (Fig. 5A). It has been demonstrated that damage to marrow sinusoidal endothelium impairs recovery of peripheral blood counts (54). Considering our data in which platelet and MK recovery were accelerated by treatment with dmPGE₂ individually as well as in tandem with lisinopril, we next investigated whether the mitigation therapies affected the recovery of sinusoidal endothelial cells. Interestingly, only the combined dmPGE₂ and lisinopril regimen significantly increased sinusoidal endothelial cell numbers compared to vehicle controls at day 10 after 4 Gy TBI (Fig. 5B).

FIG. 5.

TBI at 4 Gy causes acute injury to sinusoidal endothelium in marrow. Panel A: Sinusoidal endothelial cell numbers were significantly decreased in marrow at days 7 and 10 after 4 Gy TBI relative to nonirradiated controls. Levels recovered by day 21 after TBI. Panel B: Combined treatment with dmPGE₂ and lisinopril significantly increased sinusoidal endothelial cell numbers in marrow at day 10 after 4 Gy TBI relative to vehicle controls, back to nonirradiated control levels. n = 5–22/group, six experiments. Statistical analyses were performed using one-way ANOVA with multiple comparisons between each group. Dashed lines with shading represent the average values for nonirradiated controls. *P < 0.05, **P < 0.005 and ***P < 0.001.

DISCUSSION

We have determined that combined therapy with dmPGE₂ and the ACE inhibitor lisinopril significantly enhances survival compared to each single agent in mice after lethal-dose TBI. Interestingly, this mitigation strategy appears to target both the MK lineage and its niche in the marrow microenvironment, the sinusoidal endothelium. Several studies have established PGE₂ and its longer-acting analogue, dmPGE₂ (55), as effective regulators of hematopoiesis both in mice and in humans. In mice, hematopoietic stem cells pulsed with PGE₂ ex vivo display increased engraftment, survival and proliferation after serial transplantations (56). A published in vivo mouse study demonstrated expansion of short-term hematopoietic stem cells and multi-potent progenitors in marrow after treatment with PGE₂ without impairing their ability to differentiate at steady state (57). Moreover, in a phase 1 clinical trial, ex vivo treatment of umbilical cord blood with dmPGE₂ expanded hematopoietic stem cells that engrafted safely in 10 out of 12 study participants (58). Daily delivery of dmPGE₂ for 72 h after sublethal TBI accelerated recovery of platelet, neutrophil and hemoglobin levels in mice relative to vehicle controls (38). Consistent with these findings, we observed statistically significant increases in circulating platelets when a single dose of dmPGE₂ was delivered 24 h after 4 Gy TBI (Fig. 2A).

In contrast to dmPGE₂, ACE inhibitors are known to target endothelium. Endothelial cells present an attractive target of mitigation because thrombopoiesis requires interactions between MKs and endothelial cells both in the marrow (22, 23) and in the lungs (28). Radiation reduces the ability of endothelium to attract migrating MKs, impairing adhesion and reducing platelet production (59). Various ACE inhibitors mitigate radiation-induced dysfunction of endothelial cells in rats (40–42). Furthermore, we have demonstrated that delayed delivery of ACE inhibitors attenuates endothelial injury after TBI (43), which may circumvent logistical difficulties for patient care that may arise after mass radiation exposure incidents. We found that combined dmPGE₂ and lisinopril treatment, but not lisinopril alone, increased the recovery of sinusoidal endothelial cell numbers in the marrow after sublethal irradiation (Fig. 5B). Some published studies suggest that endothelial cells contribute to pathology after irradiation by promoting systemic inflammation (60–62). However, we did not observe a robust mitigating effect on inflammation through any therapeutic regimen in either of our TBI models (Supplementary Fig. S3; https://doi.org/10.1667/RADE-20–00113.1.S1). Although the precise mechanisms by which ACE inhibitors preserve endothelial cell function are unknown (63), ACE inhibitors may augment therapies delivered after radiation exposure by mitigating injury to the stroma. Taken together, both PGE₂ and lisinopril are FDA-approved therapies that are generally well tolerated by patients, making them ideal candidates for future studies to better understand their effectiveness as mitigators of acute and delayed radiation injury, as well as their mechanisms of action (64, 65).

Since thrombocytopenia is such an important risk factor for hemorrhage and death in H-ARS, we investigated the effects of radiation and mitigation strategies on the MK lineage. Consistent with previous findings (45), we observed acute reductions in the numbers both of marrow MKs and of circulating platelets between 5 and 7 days after 4 Gy TBI, which persisted through day 10 (Fig. 2A and B). We considered not only MK numbers but also their maturational status by examining MKs of different ploidies. At day 5 after 4 Gy TBI, the distribution of MK ploidy in the bone marrow was altered (Fig. 2C), and higher ploidy MKs were significantly smaller than normal MKs (Fig. 2G), suggesting radiation-induced abnormalities in MK maturation (Fig. 2D, E and G; Supplementary Fig. S2A and B; https://doi.org/10.1667/RADE-20–00113.1.S1). While treatment with dmPGE₂ alone increased the number of low ploidy MKs in the marrow at day 10 after 4 Gy TBI, combined dmPGE₂ and lisinopril treatment increased MKs of almost all ploidies, suggesting that the dual therapy regimen promotes increased maturation of recovering MKs after radiation injury (Fig. 4D). It is not known whether the combined dmPGE₂ and lisinopril treatment acts directly on the MK lineage or indirectly through the more rapid recovery of sinusoidal endothelial cells, which constitute the microenvironmental niche for MKs.

The relationship between thrombocytopenia and the increased risk of hemorrhage has been well studied (9, 10). Other studies have informed guidelines for the management of platelet counts in thrombocytopenic patients to prevent adverse bleeding events. While there are no clear clinical indicators to predict spontaneous bleeds (66), hemorrhage is most likely to occur in patients with less than 5,000 platelets per μl of blood and a history of active bleeding over the prior 5 days (67). These patients often receive prophylactic platelet transfusions to keep levels above 50,000 platelets per μl of blood (68). While the focus has been on platelet numbers, we also observed previously unrecognized alterations in platelet function after sublethal TBI. These include decreased surface P-selectin expression and PF4 release following platelet activation, as well as decreased capacity to contribute to thrombus formation in vivo. These alterations in platelet function may contribute to the hemorrhaging seen in the setting of H-ARS. Interestingly, PF4 is known to inhibit MK maturation (69), raising the possibility that the decrease in platelet PF4 release after sublethal TBI may represent an adaptive mechanism facilitating MK recovery. While dmPGE₂-based therapies accelerated recovery of MKs and circulating platelet numbers (Fig. 4B and C), we did not observe a mitigating effect on platelet abnormalities (Fig. 4E). Further work is required to determine whether these platelet alterations resolve and to delineate their recovery kinetics. Taken together, our studies highlight the need to better understand the effects of radiation and of mitigation strategies not only on blood cell numbers but also on cellular function.

Given the ongoing risk of large-scale radiological accidents, there is a heightened demand for effective medical countermeasures to treat thrombocytopenia in the setting of H-ARS (70–72). Significantly, our studies demonstrate the ability of a combined dmPGE₂ and lisinopril therapy to improve murine survival and platelet recovery after gamma radiation exposure. Both agents are well tolerated in humans and are already in clinical use for other purposes (64, 65). Future research is needed to investigate the ability of a combined dmPGE₂ and lisinopril treatment regimen to mitigate H-ARS in large animal models, such as in non-human primates, to satisfy the FDA Animal Rule for approval (13). Such efforts will aid progress towards the goal of developing more effective and more readily available therapies for radiation-induced thrombocytopenia.

Supplementary Material

Supplementary file

Fig. S1. Necropsy examination reveals multifocal intracranial hemorrhage after TBI. Panel A: 8–10-week-old, female C57BL/6J mice received 7.75 Gy TBI and subsequent supportive care. Three moribund mice (226, 227, 228) were euthanized at day 18 postirradiation. All had evidence of intracranial bleeding. Panel B: Images of periventricular hemorrhages (mice 226, 227, upper panels), and perivascular hemorrhages (mouse 228, lower panels). Size bars = 50 μM.

Fig. S2. TBI at 4 Gy transiently alters BM MK maturation. Panel A: The nuclear area of 8N and greater ploidy MKs was significantly decreased in vehicle-treated mice compared to nonirradiated controls at day 5 after 4 Gy TBI. Panel B: Cytoplasmic area was also decreased at day 5 after TBI. Panels C and D: Differences in nuclear or cytoplasmic area were not observed for MKs of any ploidy at day 10 after TBI. n = 3–8/group, three experiments. Statistical analyses were performed using two-way ANOVA with multiple comparisons between each group. Panel E: Activation assays of adult platelets stained with calcein red-orange or with calcein green, compared with unstained control platelets. Platelets were activated with 0.5 mg of thrombin (+) and the surface expression of P-selectin (CD62P) was compared to no thrombin (−) controls. Panels F and G: dmPGE₂ and lisinopril treatment, alone or in combination, did not affect recovery of hemoglobin levels or white blood cell counts (WBC), at day 10 after 4 Gy TBI. n = 12/group, three experiments. Statistical analyses were performed using one-way ANOVA with multiple comparisons between each group. Dashed lines represent nonirradiated control levels. *P < 0.05, ***P < 0.001 and ****P <0.0001.

Fig. S3. dmPGE₂ and lisinopril therapies do not appear to mitigate systemic inflammation after TBI. Panel A: Circulating PLA levels were increased above nonirradiated control levels at day 5 after 4 Gy TBI. Panel B: PLA levels were decreased in all irradiated groups at day 10 after 4 Gy. Panel C: Plasma CRP levels were increased in all irradiated groups at day 10 after 4 Gy. Panel D and E: Plasma CRP levels were not altered from nonirradiated control levels at days 10 and 12 after 7.75 Gy. n = 3–10/group, six experiments. Statistical analyses were performed using one-way ANOVA with multiple comparisons between each group. *P < 0.05.