Thrombopoietin mimetic stimulates bone marrow vascular and stromal niches to mitigate acute radiation syndrome

Background Acute radiation syndrome (ARS) manifests after exposure to high doses of radiation in the instances of radiologic accidents or incidents. Facilitating the regeneration of the bone marrow (BM), namely the hematopoietic stem and progenitor cells (HSPCs), is a key in mitigating ARS and multi-organ failure. JNJ-26366821, a PEGylated thrombopoietin mimetic (TPOm) peptide, has been shown as an effective medical countermeasure (MCM) to treat hematopoietic-ARS (H-ARS) in mice. However, the activity of TPOm on regulating BM vascular and stromal niches to support HSPC regeneration has not yet been elucidated. Methods C57BL/6J mice (9–14 weeks old) received sublethal or lethal total body irradiation (TBI), a model for H-ARS, by 137Cs or X-rays. At 24 hours post-irradiation, mice were subcutaneously injected with a single dose of TPOm (0.3 mg/kg or 1.0 mg/kg) or PBS (vehicle). At homeostasis and on days 4, 7, 10, 14, 18, and 21 post-TBI with and without TPOm treatment, BM was harvested for histology, BM flow cytometry of HSPCs, endothelial (EC) and mesenchymal stromal cells (MSC), and whole-mount confocal microscopy. For survival, irradiated mice were monitored and weighed for 30 days. Lastly, BM triple negative cells (TNC; CD45−, TER-119−, CD31−) were sorted for single-cell RNA-sequencing to examine transcriptomics after TBI with or without TPOm treatment. Results At homeostasis, TPOm expanded the number of circulating platelets and HSPCs, ECs, and MSCs in the BM. Following sublethal TBI, TPOm improved BM architecture and promoted recovery of HSPCs, ECs, and MSCs. Furthermore, TPOm elevated VEGF-C levels in normal and irradiated mice. Following lethal irradiation, mice improved body weight recovery and 30-day survival when treated with TPOm after 137Cs and X-ray exposure. Additionally, TPOm reduced vascular dilation and permeability. Finally, single-cell RNA-seq analysis indicated that TPOm increased the expression of collagens in MSCs to enhance their interaction with other progenitors in BM and upregulated the regeneration pathway in MSCs. Conclusions TPOm interacts with BM vascular and stromal niches to locally support hematopoietic reconstitution and systemically improve survival in mice after TBI. Therefore, this work warrants the development of TPOm as a potent radiation MCM for the treatment of ARS.


BACKGROUND
Exposure to ionizing radiation, whether from accidental incidents or as a preparative regimen for allogenic stem cell transplantation to treat leukemias, results in profound bone marrow (BM) injury.Total body irradiation (TBI) can affect various organ systems, with the hematopoietic system being the most radiosensitive (1,2).Preserving and reconstituting hematopoietic stem and progenitor cells (HSPCs) in the BM is crucial for mitigating mortality from hematopoietic acute radiation syndrome (H-ARS), typically occurring after high-dose TBI over a short period of time (3,4).HSPCs comprise all blood and immune cells which support the vital function of eliminating infection among many others (5)(6)(7)(8).A comprehensive understanding of the mechanisms and biochemical pathways governing HSPC regeneration is essential for developing life-saving medical countermeasures (MCMs) against H-ARS and mitigating radiation injuries in clinical applications where radiation is used.
The BM microenvironment consists hematopoietic, vascular, and stromal niches which support and nurture each other (9)(10)(11).HSPCs can be broadly classi ed into myeloid progenitor cells (MPC) and lineage − , Sca-1 + , c-kit + (LSK) cells.The LSK population consists of short-term 'cycling' HSCs (ST-HSCs) and long-term 'quiescent' HSCs (LT-HSCs) (12)(13)(14).For nonhematopoietic cells, the vascular niche is comprised of endothelial progenitor cells (EPCs) and endothelial cells (ECs) that release canonical niche factors such as stem cell factor (SCF), CXCL12, and angiopoietin-1 to support HSPCs (14)(15)(16)(17)(18)(19)(20)(21).Additionally, the BM vascular niche can be divided into the sinusoidal niche, harboring both quiescent and proliferative HSCs and serving as the main site of BM entry and egress; and the arteriolar niche, supporting quiescent HSCs around small arterioles near the endosteal region of the BM (5,12,22).Furthermore, perivascular mesenchymal stromal cells (MSCs) belonging to the stromal niche represent a small, but essential part of the CD45 − TER-119 − CD31 − (triple negative cells; TNC) fraction in the BM (23).MSCs secrete growth factors and chemokines such as vascular endothelial growth factors (VEGFs), CXCL12, and SCF that support both BM ECs and HSPCs (24,25).High doses of ionizing radiation are known for inducing vascular injury; however, in-depth mechanistic studies focused on BM vascular niche injury have been limited, even more so for the stromal niche.Thus, therapeutic strategies aimed at mitigating BM vascular and stromal damage are currently lacking.Thrombopoietin (TPO), a glycoprotein hormone and pleiotropic growth factor, binds to its receptor, c-MPL, expressed on megakaryocytes, platelets, and HSPCs(26-28).TPO's primary biological function is to stimulate the generation of platelets (29,30).Clinical use of recombinant human TPO (rhTPO) for treating immune thrombocytopenia was discontinued due to subjects developing endogenous neutralizing antibodies, leading to immune-mediated thrombocytopenia (31).Consequently, a class of drugs with low immunogenicity, referred to as TPO mimetics, were developed to stimulate c-MPL signaling (32).JNJ-26366821, a TPO mimetic peptide (hereafter referred to as TPOm), is comprised of 29-amino acids conjugated to polyethylene glycol moieties with no sequence homology to endogenous TPO (33,34).Both rhTPO and other TPO mimetics, including JNJ-26366821, have demonstrated e cacy in mitigating H-ARS in murine and non-human primate models (35)(36)(37)(38)(39). Currently, the impact of TPO on the BM vascular and stromal niches for HSPC regeneration after irradiation has remained largely unexplored.
In this study, we investigated the impact of TPOm on megakaryocytes, hematopoietic, endothelial, and stromal cell populations in the BM of healthy and TBI using histopathology and ow cytometry.We evaluated the e cacy of TPOm given subcutaneously 24 hours post-irradiation in enhancing the survival of TBI mice exposed to varying doses of 137 Cs and X-rays.In addition, we applied in situ whole-mount confocal microscopy to examine the effect of TPOm on the architecture of the arteriolar and sinusoidal vessels with high spatio-temporal resolution in the BM.We also measured the levels of VEGF-A and C in the BM and serum and employed IVIS (In Vivo Imaging System) to evaluate TPOm's effect on vascular permeability post-TBI.Lastly, we analyzed the effects and interactions of MSCs in the CD45 − TER-119 − CD31 − (triple negative cells; TNC) fraction with other hematopoietic TNCs using single-cell RNA-sequencing (scRNA-seq).Our study, for the rst time, elucidates a novel activity of TPOm in promoting the interaction of MSCs with other progenitors in BM, thereby sustaining BM vasculature and niche homeostasis, leading to HSPC regeneration and mitigation of H-ARS.

TPOm expands hematopoietic, endothelial, and stromal cells in murine bone marrow at homeostasis
To con rm the biological activity of TPOm through the TPO/c-MPL pathway, we initially measured platelets in the peripheral blood of normal C57BL/6J mice on days 1, 3, 6, and 13 following a single subcutaneous (sc) injection of TPOm dosed at 0.3 mg/kg.As expected, platelets (PLT) gradually increased over time, reaching a 3.2-fold peak compared to naïve 6 days after TPOm treatment (Fig. 1A).Furthermore, the numbers of white blood cells (WBCs), neutrophils (NE), and lymphocytes (LY) increased ~ 2.7-fold 3 days after TPOm treatment compared to naïve (Fig. 1B-D).Considering platelets arise from megakaryocytes, we then examined them in the BM after TPOm treatment.Histological analysis of sternal marrow with H&E revealed a signi cant 2.6-and 3.2-fold increase in megakaryocytes on days 3 and 6 post-TPOm treatment, respectively, compared to naïve (Fig. 1E, F).Morphologically in the TPOm-treated mice, the megakaryocytes are mostly mature, polylobated with abundant cytoplasm.A few young, smaller mononuclear megakaryocytes can also be visualized with rare mitoses.Lastly, the number of megakaryocytes in the TPOm-treated mice returned to baseline by day 13 (Fig. 1F).
As the number of BM megakaryocytes increased following TPOm treatment, we subsequently investigated the expansion of HSPCs in BM.The HSPC populations were assessed by ow cytometry on days 1, 3, 6, and 13 post-TPOm treatment with the gating strategy depicted in Figure S1A.MPCs (lineage − c-kit + ), exhibited a signi cant increase on day 1, followed by a signi cant depletion on day 6 after TPOm treatment relative to naïve mice (Fig. 1G and Figure S1G).Despite this decrease, there was a signi cant, transient expansion of MPCs on day 10, returning to baseline by day 13 after TPOm treatment (Fig. 1G and Figure S1B).The LSK, megakaryocyte progenitors (MkPs), and ST-HSCs (LSK, CD34 + ) exhibited a similar trend to MPCs after TPOm treatment (Fig. 1G and Figure S1B).Lastly, LT-HSCs (LSK, CD34 − CD48 − CD150 + ) only showed a signi cant increase by days 10 and 13 post-TPOm treatment (Fig. 1G and Figure S1B).
TPOm preserves murine bone marrow architecture and facilitates recovery of hematopoietic, endothelial, and stromal cells after sublethal irradiation Having established TPOm's capacity to expand HSPC, EC, and MSC populations in the BM of healthy mice, we subsequently investigated its potential to restore the BM of mice subjected to sublethal TBI.Following TBI with 137 Cs dosed at 7 Gy, mice received a single sc dose of TPOm at 0.3 mg/kg 24 h post-irradiation and assessed on days 2, 4, 7, and 14 post-TBI.First, the integrity of BM in irradiated mice was examined over time by H&E histology.On day 2 post-TBI, no gross differences were observed between vehicle-and TPOm-treated mice (Fig. 2A).On day 4 post-TBI, the size and extent of hemorrhage in the vehicle-treated mice were greater than the TPOm-treated mice (Fig. 2A).On day 7 post-TBI, early hematopoietic regeneration was evident near the endosteum in TPOm-treated mice, while not yet found in the vehicle-treated mice (Fig. 2A).Lastly, on day 14 after TBI, expansion of megakaryocytes was observed in the sternebrae of TPOm-treated mice, but not as prominently in the vehicle (Fig. 2A, B).Morphologically, the megakaryocytes resemble those present in the non-irradiated mice after TPOm treatment (Fig. 1E).Moreover, an increase in adipocytes was noted in the irradiated marrow on day 14 in both groups (Fig. 2A).Quantifying adipocytes using MarrowQuant (40) identi ed more adipocytes in the vehicle-treated BM relative to the TPOm-treated BM on day 14 (Fig. 2C).Assessment of cellularity, determined by counting live cells per femur, indicated a signi cant recovery on day 21 after TBI in the TPOm-treated mice, showing a 3.8-fold increase relative to vehicle-treated (Fig. 2D).
To precisely evaluate the impact of irradiation with and without TPOm on HSPC, EC, and MSC populations in the BM, we conducted ow cytometry.In TPOm-treated mice, the frequency of MPCs signi cantly increased by day 10 after irradiation (Figure S2A), while the absolute count surpassed that of the vehicle-treated mice from day 7 through day 21 after TBI (Fig. 2E).Similarly, the absolute count of MkPs in TPOm-treated mice signi cantly exceeded that in vehicle-treated mice from day 7 to day 21 after TBI, excluding day 14 (Fig. 2E).For the LSK population, a signi cant increase in frequency was observed on days 7 and 10 in the TPOm-treated mice (Figure S2A), while their absolute counts remained elevated through day 14 relative to vehicle-treated mice (Fig. 2E).ST-HSCs exhibited a similar trend to LSK cells (Fig. 2E).Conversely, for the rare LT-HSCs, a signi cant increase in both frequency and absolute count was only observed on day 21 after TBI in TPOmtreated mice compared to vehicle-treated mice (Fig. 2E and Figure S2A).
In the BM vascular and stromal niches, EPCs in TPOm-treated mice exhibited a signi cant increase in both frequency and absolute count on day 7, compared to vehicle-treated mice (Fig. 2F and Figure S2B).Total BM ECs showed a signi cant increase in frequency only on day 7 in TPOm-treated mice compared to the vehicle-treated mice (Figure S2B).Moreover, the absolute count of total ECs in TPOm-treated mice was signi cantly higher than the vehicle-treated mice from days 7 through 21, with a trend for higher counts on day 14 that was not signi cant (Fig. 2F).For the subsets of ECs, the absolute count of AECs showed no statistical difference between vehicle-and TPOm-treated mice, while the absolute count of SECs signi cantly increased in TPOm-treated mice from day 7 through day 21 (Fig. 2F).Finally, the absolute count of MSCs also signi cantly increased on days 7 and 14 in TPOm-treated mice compared to the vehicle-treated mice after TBI (Fig. 2F).These results emphasize TPOm's role in stimulating repopulation and regeneration of HSPCs, ECs, and MSCs in the BM, thereby contributing to the preservation of the BM architecture in mice after sublethal irradiation.

TPOm increases survival of mice exposed to lethal irradiation
As TPOm demonstrates the capability to mitigate BM damage in mice following sublethal irradiation, we explored its potential as a radiation MCM for H-ARS.The primary method for evaluating the e cacy of a radiation MCM candidate in animal models is by assessing its impact on the 30-day survival post-lethal TBI.To evaluate the e cacy of TPOm, C57BL/6J irradiated with lethal dose of TBI at 8.8 Gy from a 137 Cs source.This radiation dose was selected because 8.8 Gy TBI is the lethal dose for 70% (LD 70 ) of mice within 30 days post-TBI.24 hours post-TBI, mice received a single sc dose of TPOm at 0.3 mg/kg or vehicle.As illustrated in Fig. 3A, TPOm treatment signi cantly increased the survival from 28.6% in vehicle-treated mice to 93.3% in the TPOm-treated mice.Moreover, TPOm treatment signi cantly prevented body weight loss compared to the vehicle-treated mice (Fig. 3B).
We further assessed the e cacy of TPOm in enhancing survival in mice exposed to X-ray TBI.The LD 100 using orthovoltage X-rays is lower than 137 Cs, as established in our previous publication (41).Notably, a signi cant increase in survival was observed in mice exposed to 6.7 and 7.2 Gy TBI (LD 50 and LD 100 , respectively) with TPOm treatment compared to the vehicle-treated mice (Fig. 3C, E).Additionally, the male surviving mice maintained their body weight after irradiation (Fig. 3D, F).In female mice, no signi cant differences were observed in survival or weights (Figure S3A, B).
Upon closer examination of BM damage after 8.8 Gy ( 137 Cs) TBI by H&E histology, we noted a substantial reduction in BM cellularity in both vehicle-and TPOm-treated mice on day 4 (Figure S3C).On day 14 after TBI, sinusoidal dilatation was evident in both groups.However, TPOm-treated mice exhibited more de ned, intact vessels, known as BM angiectasis.In contrast, vehicle-treated mice displayed extensive hemorrhaging of the vessels with erythrocytes present in the parenchyma, indicating compromised BM sinusoids (Figure S3C).Overall, these data demonstrate that TPOm can serve as a potent radiation MCM by increasing the survival of mice exposed to a lethal dose of radiation from various sources.
TPOm accelerates and promotes restoration of the vascular niche in mice after lethal irradiation Extensive vascular dilatation and an increase in vascular area in the diaphysis are inherent responses to BM stress and injury, particularly for the BM sinusoids (9).Therefore, we investigated the sinusoidal niche in situ using whole-mount confocal microscopy of the femur.BM SECs were identi ed by vascular endothelial growth factor receptor 3 (VEGFR3) labeling, which is exclusively expressed on the BM sinusoids (Fig. 4A).Combined labeling of CD31 and CD144 was employed to examine the total BM vasculature, with DAPI used for nuclei identi cation in mice irradiated at 8.8 Gy ( 137 Cs) TBI.The VEGFR3 + -stained area in naïve mice was 19.8 ± 1.45 × 10 3 µm 2 (Fig. 4A, B).By day 4 after irradiation, this area increased to 53.7 ± 2.30 × 10 3 µm 2 in vehicle-treated mice, while TPOm signi cantly inhibited the dilation to 43.3 ± 2.39 × 10 3 µm 2 (Fig. 4A, B).Furthermore, by day 10, the VEGFR3 + -stained area of vehicle-treated mice reached 45.7 ± 3.36 × 10 3 µm 2 , while with TPOm, it was further reduced to 35.6 ± 2.08 × 10 3 µm 2 (Fig. 4A, B).These results demonstrate that TPOm enhances the restoration of BM sinusoids in irradiated animals.
To assess the effect of lethal TBI at 9.0 Gy ( 137 Cs) on the arteriolar niche, we used angiopoietin-1 receptor, known as TIE2, as a marker to distinguish AECs, given their high expression of TIE2 (42).BM arterioles also express Sca-1, typically a marker of for hematopoietic progenitors (Figure S1A).On days 2 and 4 after TBI, arterioles, marked as Sca-1 + and TIE2 + , were readily detected in both naive and irradiated mice and appeared unchanged (Fig. 4C).Moreover, on day 4, the Sca-1 + hematopoietic cells were found more clustered in the TPOm-treated mice compared to the vehicle, particularly around the arterioles (Fig. 4C).The number of Sca-1 + cells in the vehicle mice began to decrease on day 2 and further decreased on day 4, compared to the naïve (Fig. 4D).In contrast, the number of Sca-1 + cells in the TPOm-treated mice were well maintained, reaching a level similar to the naïve mice and signi cantly higher than the vehicle (637 ± 41 vs. 1150 ± 58 cells per eld) on day 2, although it dropped to similar counts as the vehicle on day 4 (Fig. 4D).Despite not detecting noticeable changes in the BM arterioles after irradiation, there were increased numbers of Sca-1 + cells near the endosteum and arterioles in TPOm-treated mice.
To further analyze the impact of TPOm on vascular integrity and function, we utilized IVIS imaging to evaluate vascular permeability.Mice were exposed to 7.2 Gy TBI with X-rays, followed by TPOm treatment 24 hours post-irradiation.On day 3 after irradiation, mice received an injection of a vascular dye (AngioSense 750EX) intravenously and were imaged 48 hours later.The IVIS imaging showed that there was a greater amount of dye present in the tissues of the vehicle-treated mice compared to the TPOm-treated mice (Fig. 4E).After quanti cation of the region of interests (ROIs), the total radiant e ciency in the TPOm-treated group was signi cantly lower than the vehicle-treated group by 53.2% (Fig. 4F).These ndings suggest that TPOm contributes to vascular integrity, as evidenced by the reduced leakage of vascular dye into surrounding tissues.
VEGFs play a crucial role in regulating ECs, in uencing growth and repair processes (43).Consequently, we investigated the effect of TPOm on the levels of VEGF-A and VEGF-C in the BM and serum.In healthy mice, following TPOm injection, VEGF-A levels signi cantly increased on day 3 in serum and on day 13 in the BM (Figure S4A, B).Concurrently, VEGF-C levels signi cantly increased on days 1 and 3 in the BM and serum, respectively, after TPOm treatment (Figure S4C, D).
However, when mice were subjected to 7 Gy ( 137 Cs) TBI, TPOm treatment did not elevate VEGF-A levels in either serum or BM, unlike healthy mice (Fig. 4G, H).Nevertheless, the levels of VEGF-C in the BM and serum of TPOm-treated mice were markedly increased on days 2 and 4, respectively, compared to the vehicle-treated and naïve mice (Fig. 4I, J).These results highlight that TPOm can selectively stimulate the release of VEGFs, distinctively VEGF-C, both systemically and locally, promoting the repair of vascular damage in the BM and potentially other organs after irradiation.
TPOm elicits distinct changes in cellular heterogeneity and cell cycle dynamics of murine bone marrow cells postirradiation evaluated by single-cell RNA-sequencing analysis BM MSCs play a signi cant role in the regeneration of HSPCs and ECs, particularly after irradiation (24,44).Given the observed increase in MSCs after TPOm treatment (Fig. 1H and 2F), we further investigated the effect of TPOm on MSCs.
We sorted BM cells from mice using the markers CD45 − , TER-119 − , and CD31 − [triple negative cells (TNC)](23, 45) as this fraction is enriched for MSCs and conducted single-cell RNA-sequencing.Mice were divided into four groups: naïve, TPOm alone, 6 Gy TBI (X-rays), and 6 Gy TBI followed by TPOm treatment 24 hours post-irradiation.BM was harvested on day 10 after irradiation.A heatmap of the top 10 enriched genes was generated for each cluster to identify the populations of TNCs (Fig. 5A).An overall UMAP was generated by combining clusters from all four groups (Fig. 5B) as well as individual UMAPs for each group (Fig. 5C).Notably, neutrophil progenitors (Neutro_prog), megakaryocyte progenitors (Mk_prog), and eosino-basophil progenitors (Eo-Baso_prog) were noticeably depleted after irradiation, while Pro-B cells and both clusters of erythroblasts were increased (Fig. 5C).Analyzing the percentage of each cluster per group revealed that TPOm alone increased the percentage of Neutro_prog and Mk_prog from 32.9-38.5% and 13-17.9%,respectively, compared to naïve (Fig. 5D).Following irradiation, most identi ed clusters, with the exception of Pro-B cells, were increased after TPOm treatment relative to the irradiation alone group (Fig. 5D).
To further pinpoint clusters demonstrating active proliferation, we analyzed the expression of the proliferation marker Mki67.The combined UMAP displayed elevated levels of Mki67 in the erythroblasts and Pro-B cells (Fig. 5E).As anticipated, TPOm increased Mki67 expression in Mk_prog cluster (Figure S5A).Irradiation increased Mki67 expression in Neutro_prog, Mk_prog, and erythroblast 2 clusters, an effect further ampli ed in the irradiation plus TPOm-treated group (Fig. 5F and Figure S5A).Additionally, we assessed the cell cycle status of each cluster for all the individual groups (Figure S5B).For comparison, the percentage of each cell cycle phase (G1, G2M, or S) in each cluster was plotted for all groups (Fig. 5G).In healthy mice, TPOm notably increased the percentage of Pro-B cells in the G2M phase, Neutro_prog in the G1 phase, and erythro-progenitors (Erythro_prog) in the G1 phase (Fig. 5G).Radiation increased the percentages of several clusters in S phase, while TPOm treatment slightly decreased the percentage of all the Erythro-clusters in the S phase (Fig. 5G).Particularly, MSCs exhibited a slight increase in the S phase after TPOm treatment in the irradiated groups; however, the percentage of MSCs in G2M was increased with TPOm treatment compared to irradiation alone (Fig. 5G).Collectively, these data highlight that TPOm regulates both the proliferation and cell cycle dynamics of erythroid, B lymphoid, and MSCs after irradiation.

TPOm enhances the interaction of mesenchymal stromal cells with other hematopoietic progenitors in the mouse bone marrow after irradiation
To explore cell-cell interaction among the different clusters, we used the CellChat(46) program to analyze the single-cell RNA sequencing data.The analysis revealed that MSCs acted as a central signaling hub, engaging in robust interactions with other clusters in the dataset (Fig. 6A).Identi cation of MSCs was based on the expression levels of several genes, including canonical MSC markers such as Pdgfra and Lepr (Figure S6A).The predicted ligand-receptor communications from MSCs to each cluster, along with the intensity of these interactions is illustrated in Fig. 6B.Collagens expressed in MSCs emerged as the primary contributors to the cell-cell interactions, with Col1a2-Cd44 exhibiting the highest contribution (Fig. 6C).Cd44 was highly expressed in Neutro_prog, Mk_prog, and Erythro_prog (Fig. 6B and Figure S6B).
Another signi cant molecule in mediating cell-cell interactions was Sdc4, which was highly expressed in Pro-B cells and MSCs (Fig. 6B and Figure S6B).We further analyzed the expression levels of several genes in the collagen family in each experimental group.In TPOm-treated mice, there was an increased expression of Col1a1 and Col1a2 in MSCs, and irradiation also heightened this expression (Fig. 6D).Particularly in the irradiated groups, TPOm treatment increased the expression of Col1a2, Col4a1, and Col4a2 in MSCs compared to radiation alone, while it decreased the expression of Col6a2 (Fig. 6D).
We further explored the expression of genes in MSCs that might be in uenced by TPOm treatment.MSCs inherently expressed canonical EC ligands, such as VEGFs, which would contribute to EC regeneration (Figure S6C).Differential expression analysis between TPOm treatment and naïve mice revealed that one gene, Col8a1, was signi cantly downregulated, and 12 genes were signi cantly upregulated, including various collagens and osteoblastic genes (Fig. 6E).
The single cell pathway analysis (SCPA) using Gene Ontology Biological Pathways database (GOBP) of differentially expressed genes demonstrated a notable increase in the regeneration pathway in TPOm-treated mice compared to naïve (Fig. 6F).When comparing the differential expression between TBI and TBI plus TPOm-treated mice, 11 genes were signi cantly downregulated, and 3 genes were signi cantly upregulated (Fig. 6G).The SCPA of these genes indicated that the humoral immune response, multicellular organismal response to stress, and response to oxidative stress were signi cantly downregulated in the TPOm-treated mice after TBI (Fig. 6H).Together, these results suggest that TPOm upregulates several genes in the collagen family in MSCs, promoting their interaction with other hematopoietic TNCs in the BM.Moreover, TPOm stimulates regeneration and suppresses the humoral immune response in mice with treatment alone or after TBI, respectively.

DISCUSSION
In today's geopolitical climate, individuals face the looming threat of exposure to high doses of ionizing radiation due to nuclear or radiological incidents, which carries the risk of developing ARS.In addition, patients undergoing myeloablative BM transplant conditioning suffer from radiation-induced toxicities and mortality.The hematopoietic system, consisting of highly proliferative stem cells, stands out as one of the most susceptible organ to radiation-induced injury (47).In established animal models for TBI, four FDA-approved drugs targeting the hematopoietic system-Neupogen, Neulasta, Leukine, and Nplate (Romiplostim)-have demonstrated e cacy in increasing HSPCs in irradiated animals (2,(48)(49)(50).
However, their activity regulating the BM microenvironment for HSPC regeneration remains unexplored.In this study, we have evaluated the potential of TPOm (JNJ-26366821) as a radiation MCM (37) and agent than can mitigate radiation induced toxicities, focusing on its role in regulating BM vascular and stromal niches for HSPC regeneration in mice exposed to TBI from 137 Cs and X-ray sources.
Our ndings show that TPOm effectively expanded HSPCs, ECs, and MSCs in the BM of both healthy and irradiated mice.
TPOm also signi cantly improved the 30-day survival of TBI-exposed mice, a necessary endpoint for evaluating drug e cacy in treating H-ARS according to the FDA animal rule (2).Furthermore, our study reveals a novel activity of TPOm in alleviating BM vascular dilation in the sinusoidal niche and maintaining the arterioles of the arteriolar niche post-TBI.TPOm reduced vascular permeability, a typical consequence of exposure to high doses of radiation and increased the levels of VEGF-C in BM and serum.scRNA-seq analysis unveiled another novel function of TPOm in upregulating the expression of speci c collagens in MSCs, thereby promoting their interaction with other rare hematopoietic progenitors in the BM.Additionally, TPOm upregulated regeneration and dampens the humoral response in MSCs.
We have veri ed that TPOm exhibited functionality akin to endogenous TPO by exerting its role as a regulator of platelet production from megakaryocytes through the differentiation of HSCs (29).TPOm was developed by screening peptides capable of binding to c-MPL with a phage display library (51).After a single sc injection, TPOm signi cantly increased the number of megakaryocytes in the BM of the healthy mice on days 3 and 6 which is in line with its role of promoting Mk differentiation.Moreover, TPOm induced a signi cant expansion of the HSPCs, consistent with prior studies demonstrating direct binding of TPO to HSCs and the expansion of LSK cells (52).Recognizing the critical role of vascular and stromal niches in the BM for HSPC regeneration (9,24,44), our study examined the broader effects of TPOm treatment on other constituents of the BM microenvironment.We observed a signi cant increase in EPCs, ECs, AECs, SECs, and MSCs in the BM of healthy mice post-TPOm treatment.Notably, some organ-speci c ECs expressing c-MPL, such as liver sinusoidal ECs and human umbilical vein ECs, have been reported (53,54) suggesting the potential of direct interactions of ECs with TPOm.Furthermore, investigators have found that BM osteoblasts and -clasts express c-MPL (55).Osteoblasts are derived from BM MSCs which may explain how TPOm is interacting with MSCs at homeostasis and after irradiation.While our ndings point to a potential role of TPOm in the expansion of EC and MSC populations in the BM, the nature of this effect, whether direct or indirect, warrants further investigation.
For mice exposed to TBI at sublethal 7.0 Gy ( 137 Cs) doses, the architecture and cellularity of BM were damaged, re ecting severe depletion of HSPCs, ECs, and EPCs that persisted for at least 14 days after TBI.With TPOm treatment, the architecture and cellularity of BM were more preserved, exhibiting less hemorrhage and adipocytes.A signi cant recovery of MPCs, LSK cells, and ST-HSCs by day 7 was observed in TBI mice treated with TPOm.Similarly, a recent study demonstrated that endogenous TPO mainly produced from liver promotes the regeneration of HSCs after chemo-and radio-induced myeloablation, an example of cross-organ signaling(56, 57).Moreover, TPOm increased EPCs at day 7 after TBI, which would differentiate into mature ECs, resulting in a marked elevation of EC counts starting from day 10 after TBI.
Next, we evaluated the effectiveness of TPOm as a potential radiation MCM by subjecting mice to lethal doses of radiation: 8.8 Gy ( 137 Cs), 6.7 Gy (X-ray), and 7.2 Gy (X-ray).Administering TPOm 24 hours post-TBI resulted in a signi cant increase in 30-day survival rates at all three radiation doses, exceeding the vehicle-treated mice by at least 45%.Notably, TPOm demonstrated its e cacy not only in male mice but also in improving the survival of female mice exposed to X-ray irradiation.In addition to the enhanced survival rates, TPOm treatment effectively mitigated body weight loss following TBI.This mitigative effect aligns with our previous study, which demonstrated that TPOm signi cantly increased the survival of CD2F1 and C57BL/6J mice exposed to TBI from a 60 Co γ-radiation source in a dose-dependent manner (37).An important consideration is that various radiation sources, as described in our prior publication, can have signi cantly different effects on the composition of the BM depending on the dose; as such, γ-radiation and X-rays at isodoses are not equivalent (41).
Our ndings are consistent with the activity of other c-MPL agonists in mitigating H-ARS.For instance, administration of rhTPO enhanced the HSPC recovery in irradiated mice and signi cantly improved the survival of both mice and nonhuman primates exposed to lethal TBI(58).Romiplostim, another TPO mimetic recently approved by USA FDA to treat patients acutely exposed to myelosuppressive doses of radiation, has also demonstrated its e cacy in conferring a survival bene t in murine and non-human primate models of H-ARS(38, 59, 60).To our knowledge, this is the rst study to examine the role of TPOm in mitigating radiation-induced vascular and stromal injuries to support hematopoietic regeneration, marking a signi cant advancement in our understanding of TPO's multifaceted mitigative mechanisms.
The vasculature within the BM can be divided into sinusoids and arterioles (17).In mice exposed to TBI, we observed increased vessel dilation of the sinusoids, quanti ed by the area of VEGFR3, which was reduced by TPOm treatment.In the arteriolar niche, the structure of arterioles remained intact after lethal irradiation, albeit a marked decrease in the number of Sca-1 + cells was observed.Previous studies have reported differences in radiosensitivity between sinusoids and arterioles in the BM (9,13).The impact of TPOm on vasculature was further evident in the reduction of vascular dye leakage throughout the body of TBI mice, as detected by IVIS imaging.Likewise, we demonstrated that TPOm increased the levels of VEGFs in the BM and serum.Our ndings are supported with previous reports indicating that TPO released from BM stromal cells can bind to HSCs to stimulate VEGF, implying a potential role in vascular regeneration (25,61).
Vascular swelling after irradiation can be alleviated by HSC transplant supplemented with VEGF-A(9), and MSC-secreted VEGF-C has been shown to be crucial in regeneration of the vascular niche after irradiation (24).Remarkably, TPOm distinctively increased VEGF-C levels within the BM and serum in healthy and irradiated mice.Consequently, TPOm may exert bene cial effects on regeneration and recovery of ECs post-irradiation, which subsequently affects BM HSPCs, potentially through the release of VEGFs.
To investigate the effect of irradiation with and without TPOm treatment on BM MSCs, we isolated the TNCs from the BM using cell sorting for single-cell RNA sequencing.Previous studies have established that the CD45 − , TER-119 − , CD31 − fraction of the BM is enriched with a heterogeneous population of MSCs and devoid of hematopoietic, erythroid, and endothelial cells (45).However, recent ndings have challenged this notion, revealing that TNCs contain cells of hematopoietic origins, particularly B lymphoid and erythroid lineages, which are dependent on signals from MSCs (23).
Our scRNA-seq data revealed a signi cant loss of Mk_prog, Neutro_prog, and Eo-Baso_prog populations after irradiation, with TPOm treatment mitigating this loss to some extent.Conversely, B lymphoid and erythroid lineages expanded after irradiation, with the erythroid clusters showing further enhancement with TPOm treatment.One study reveals that B lymphoid lineage and plasma cells derived from BM are resistant to radiation(62).The elevation of Ter-119 low/− erythroid cells, a hallmark of stress-erythropoiesis, is typical after irradiation(63).Further, TPO has been demonstrated to synergize with erythropoietin (EPO) and support erythroid recovery following myeloablative injury(64).These ndings suggest a potential role for TPOm in in uencing the dynamics of various hematopoietic lineages post-irradiation.
The scRNA-seq data also uncovered the pivotal role of MSCs as central regulators of various hematopoietic TNCs.
Interactions between MSCs and the other clusters were predominantly mediated by Col1a2/Col1a1 and syndecan-4 (Sdc4) or Cd44.Speci cally, Sdc4 exhibited high expression on the Pro-B cluster and has been shown to modulate cell migration and adhesion(65).Alternatively, the Neutro_prog, Erythro_prog, and Mk_prog clusters predominantly interacted with MSCs through Cd44, which is known for its critical functions in cell migration, adhesion, and homing(66).Moreover, Cd44 has been used to differentiate stages of erythroid lineage development(67).Sdc4 and Cd44 are known as cellsurface heparan sulfate proteoglycans that are indispensable for humoral immune system development and maintenance of hematopoiesis, in general(68).These ndings collectively suggest that MSCs play a crucial role in maintaining early B lymphoid and erythroid cells, priming them for HSPC recovery after irradiation.It is noteworthy that TPOm appears to augment these interactions, indicating a potential enhancement of MSC-mediated recovery of hematopoietic cells after irradiation.
TPOm has undergone thorough nonclinical toxicology evaluations, including chronic toxicity studies, and no issues have been identi ed that would preclude its clinical development (33).A Phase 1 clinical study involving healthy volunteers further supported the safety and tolerability of TPOm.Particularly, TPOm dose-dependently elevated platelet counts and increased total colony-forming unit (CFU) counts compared to the placebo, with no evidence of antibody formation against endogenous TPO in humans(69).
In conclusion, our study has unveiled novel functions of TPOm (JNJ-26366821) in regulating the vascular and stromal niches in the BM, fostering the regeneration of HSPCs in irradiated mice.TPOm's stimulation of VEGF secretion contributed to the maintenance vascular integrity in irradiated mice.Additionally, TPOm promoted MSCs to interact with other progenitors in the BM.These TPOm-induced effects collectively resulted in a signi cant improvement in the survival of the TBI mice, a model of H-ARS (Fig. 7).Taken together, TPOm is positioned as a clinical ready drug, meriting further development as radiation MCM for potential FDA approval.Animals C57BL/6J (wild-type; stock no.000664), B6.Cg-Tg(Tek-cre)1Ywa/J (TIE2-cre; stock no.008863), B6.Cg-Gt(ROSA)26Sor tm14(CAG-tdTomato)Hze /J (tdTomato; stock no.007914) mice were purchased from Jackson Laboratories (Bar Harbor, ME).B6.Cg-Tg(Tek-cre)1Ywa/J and B6.Cg-Gt(ROSA)26Sor tm14(CAG-tdTomato)Hze /J were crossed together to generate constitutive Tek-cre; tdTomato (TIE2-tdTomato) mice in our facilities.All mice were acclimated for 1 week prior to experiments and group housed (no more than 5 per cage) in pathogen-free conditions under a 14:10 hour light:dark cycle.

Materials and Methods
Moreover, mice were housed at 20°C to 22°C with 30-70% humidity and fed ad libitum (Lab Diet 5001).To limit pathogen transmission, water was acidi ed to a pH of 2.5 to 3.0 with HCl for survival studies.All experiments were carried out using gender-matched littermate controls where appropriate.All mice in this study were used at 9-14 weeks of age.Both males and females were used for experiments.
Preparation and injection of (JNJ-26366821) TPOm was supplied as a powder for reconstitution at 1mg/mL in sterile PBS.TPOm dosing formulations were stored protect from light, refrigerated (set to 2-8°C) pending use for dosing within one day of preparation.Drug substance and stock solutions were stored protected from light in a -80 °C freezer.Stock solutions in the concentration at 1 mg/mL can be stored in the above referenced freezer conditions for up to 8 weeks.Either TPOm or its vehicle were injected once subcutaneously at the nape, 24 hours post-TBI.Mice were dosed at 0.3 mg/kg or 1.0 mg/kg.

Irradiation
For 137 Cs TBI, mice were anesthetized with 60:9 mg/kg ketamine:xylazine, which was equivalent to about 100 μL/mouse, and placed in single chambers of a round brass animal holder for the Shepherd Mark I irradiator.Brass container was placed on a rotating plate to expose them to uniform total body γ-irradiation according to the manufacturer's speci cations with a dose rate of about 1.90 Gy/min.TBI doses of 7.00 and 8.80/9.00Gy were used for sublethal and lethal doses, respectively.
For X-ray total body irradiation, using a CIX-3 orthovoltage source (Xstrahl), unanesthetized mice were placed into a Plexiglas jig.The X-ray irradiator was operated at 300 kVp, 10 mA with either 1 mm Cu at a dose rate of 1.89 Gy/min or 4 mm Cu ltration (for scRNA-seq data) at a dose rate of 1.12 Gy/min at a 40 cm source surface distance.All irradiation was performed in the morning.Doses and dosimetry were determined as described in our previous publication comparing 137 Cs γ-radiation to orthovoltage X-rays (41).

Complete blood count
At the time of euthanasia, mice were subjected to iso urane overdose and blood was collected via cardiac puncture into K2EDTA coated microtainer tubes (BD Pharmingen, cat# 365967).Automated complete blood count with differential was performed using a Hemavet 950FS instrument (Drew Scienti c).
Flow cytometry and cell sorting

ELISA
Plasma was collected after complete blood count analysis using 8,000xg for 5 minutes to spin down whole blood and stored at -80°C.Bone marrow supernatant was collected by ushing a 200 μL of 1X PBS through two femurs.The solution was centrifuged at 300xg for 5 minutes to separate cells and supernatant was separately stored at -80°C.Protein concentrations for the BM supernatant was determined using BCA Protein Assay kit (Thermo Scienti c, cat# 23225) according to manufacturer's instructions with BSA as a standard.ELISA for VEGF-A (R&D, cat# MMV00) and VEGF-C (Novus Biologicals, cat# NBP2-78893) were performed according to manufacturer's instructions.Standard dilutions for plasma and 10 μg of total protein for BM supernatant were loaded for either ELISA for standardization.
In vivo imaging IVIS was performed on the Caliper Life Sciences IVIS Spectrum system.Mice were intravenously perfused with AngioSense750 EX (PerkinElmer, cat# NEV10011EX) on day 3 after irradiation.On day 5 after irradiation, mice were anesthetized with iso urane (2% v/v with oxygen as the carrier gas) in an inhalation chamber (VetEquipt, cat# 911103) and maintained as mice were in the IVIS.The radiant e ciency, a relative measure of photon emission from the animal (photons/s/cm 2 ), was measured in a standardized region of interest (ROI) with the variables of exposure time, binning, and focal length/stop also standardized.Fluorescence measurements were acquired with Living Image (Perkin Elmer, v4.3.1) and are expressed as a pseudocolor on a gray background, with red representing the lowest intensity and blue the highest.

Library preparation and sequencing
Single-cell RNA sequencing libraries involved sorted bone marrow cells stained with CD45, TER-119, and CD31 markers.
These libraries were generated from a total of ~20,000 individual cells, combining cell-multiplexing oligos (CMOs) from one male and one female mouse, contributing about ~10,000 cells each.The process involved generating cDNA within individual cell-gel bead emulsion micro-reactors, during which barcodes were added at both cellular and molecular levels.
This barcoding allowed for the combination of the cDNA from individual cells for further library processing.Unique molecular barcodes (UMIs) were utilized to ensure that ampli cation artifacts did not distort the analysis.The prepared libraries underwent sequencing for 4000 M reads (PE150), with approximately 400 million read pairs for gene expression libraries and about 100 million read pairs for CMO libraries, all sequenced on an Illumina HiSeq 2500 system.

1
Abbreviations TBI Total body irradiation BM Bone marrow H-ARS Hematopoietic acute radiation syndrome TPOm Thrombopoietin mimetic, JNJ-26366821 HSPC Hematopoietic stem and progenitor cell MSC Mesenchymal stromal cell

Figure 6
Figure 6 For analysis of hematopoietic cells, femurs were ushed with 2% FBS-PBS 2mM EDTA (FPE) buffer with a 21G needle.For analysis of endothelial and stromal cells, tibias were ushed and digested with 1 mg/mL of collagenase IV (Gibco, cat# with 2% FBS-PBS solution and incubated with secondary antibodies for 30 minutes, if necessary.Cells were resuspended in FPE buffer and acquired on Cytek Aurora with SpectroFlo software or BD LSRII with FACS Diva software on ow cytometer.Cell sorting was performed on FACSAria Cell Sorter (BD Biosciences).Dead cells and debris were excluded by FSC, SSC, and Live/Dead staining.Data analysis was done through FlowJo (Tree Star, v10.1) software.Images were acquired using a water immersion lens on the ZEISS AXIO examiner D1 microscope (Zeiss) with a confocal scanner unit, CSUX1CU (Yokogawa), and reconstructed in three dimensions with Slide Book software (Intelligent Imaging Innovations, v6.0) or analyzed using Volocity software (Quorum Technologies, v6.5.1).Brie y, original images were loaded into Volocity as .TIFF le formats.Brightness-contrast and noise reduction modi cations were applied to each channel for the whole image.Quanti cation of vessel area and quanti cation of Sca-1 + cells were performed in Volocity.