Prostaglandin E₂ Prevents Apoptosis and Promotes Homeostasis of Hematopoietic Stem Cells after Lethal Irradiation through Alteration of p53 and TNF Signaling

Background: The hematopoietic system is the most sensitive organ to therapeutic or unintended radiation. In a mouse model of the hematopoietic acute radiation syndrome (H-ARS) a single dose of 16,16 dimethyl prostaglandin E2 (dmPGE2) given 15 min to 3 h before LD50/30-90/30 total body irradiation (TBI) provides 100% survival and accelerates hematopoietic recovery. The mechanisms responsible, however, are not known.

Methods: C57BL/6 or FGD5-ZsGreen (JAX) mice received 35 µg dmPGE2 or vehicle 30 min before LD50/30 (853 cGy) γ-irradiation, along with non-irradiated dmPGE2 and vehicle controls. Bone marrow (BM) cells were harvested at multiple time points post-TBI and analyzed by flow cytometry for hematopoietic stem cell (HSC) markers as well as cleaved-caspase 3, cleaved-PARP, and p21. HSC were sorted for RNA-seq 1 h post-TBI based on SLAM-LS (Sca-1, CD150, CD48, Lineage-neg) and ZsGreen fluorescence. FGD5-ZsGreen mice express ZsGreen specifically in HSC, and were used to maximize HSC purity as the standard cKit marker is affected by irradiation. RNA-seq data were analyzed using Ingenuity Pathway Analysis.

Results: The number of marker-defined HSC in dmPGE2-treated mice was preserved at 24 h post-TBI, but then declined to levels of vehicle-treated mice by day 3, implying that a qualitative rather than quantitative difference in HSC may be facilitating animal survival. To investigate gene expression changes in HSC after dmPGE2 treatment and irradiation, BM cells were harvested 1 h post-TBI and HSC sorted for RNA-seq. We first reasoned that changes observed with dmPGE2 alone may be involved in priming HSC to respond differently to irradiation, perhaps preserving HSC properties and function. The most activated upstream regulator by dmPGE2 alone was TNF, based on the expression patterns of known TNF-regulated genes, with an activation z-score of +5.26 (p = 1.39 x 10^-23). Interestingly, TNF was also the upstream regulator most activated by irradiation alone (z-score +8.26, p = 5.88 x 10^-37), to a greater degree than dmPGE2 alone, and different sets of TNF regulated genes were increased or decreased in each case. Significantly, with dmPGE2 plus irradiation, TNF-regulated genes in HSC adopted an expression pattern similar to dmPGE2-only, and generally opposite to irradiation alone. The direct comparison of dmPGE2-irradiation with vehicle-irradiation resulted in a significant net inhibition of TNF activation. Approximately one third of these altered genes are also NF-kB-regulated, suggesting that NF-kB-mediated TNF signaling induced by irradiation is prevented by dmPGE2 treatment. TNF mRNA itself was highly upregulated in both irradiated groups, but while TNF receptor 1 expression was not affected, dmPGE2 (with and without irradiation) significantly increased expression of TNF receptor 2, likely influencing TNF downstream effects.

Strong changes were also seen in the p53-mediated apoptosis pathway. This pathway was not increased by dmPGE2 alone, but was greatly activated by irradiation (p = 1.47 x 10^-27) and was the top inhibited regulator by dmPGE2-irradiation compared to irradiation alone (p = 1.86 x 10^-18). At the protein level, a corresponding decrease in HSC production of cleaved caspase-3 and cleaved-PARP was detected from 1 h to 9 d post-TBI in PGE2-treated animals after irradiation, supporting its protection of HSC from irradiation-induced apoptosis. One dramatically affected gene involved in both the TNF and p53 pathways was the cdk inhibitor p21 (Cdkn1a). Confirming the observed gene expression pattern, p21 protein was substantially increased in HSC by 24 h post-TBI, and this was prevented by dmPGE₂.

Finally, irradiation stimulated a widespread increase in HSC transcriptional activity that was generally prevented by dmPGE2, suggesting preservation of a more quiescent state. In addition, transcription factors implicated in HSC maintenance/function were significantly decreased with irradiation and protected by dmPGE2, including Gata2, Meis1, Hoxa9, Tal1, and Gfi1.

Conclusions: DmPGE2 induces rapid gene expression changes in HSC that dampen transcriptional activation induced by high-dose irradiation. This includes inhibiting activation of p53 and NFKB-mediated TNF activity, reducing apoptosis and maintaining essential HSC properties. This work provides insight into the mechanism of how dmPGE2 facilitates survival and hematopoietic recovery from lethal irradiation.