Genetic interaction of PGE2 and Wnt signaling regulates developmental specification of stem cells and regeneration

Abstract

Interactions between developmental signaling pathways govern the formation and function of stem cells. Prostaglandin (PG) E2 regulates vertebrate hematopoietic stem cells (HSC). Similarly, the Wnt signaling pathway controls HSC self-renewal and bone marrow repopulation. Here, we show that wnt reporter activity in zebrafish HSCs is responsive to PGE2 modulation, demonstrating a direct interaction in vivo. Inhibition of PGE2 synthesis blocked wnt-induced alterations in HSC formation. PGE2 modified the wnt signaling cascade at the level of beta-catenin degradation through cAMP/PKA-mediated stabilizing phosphorylation events. The PGE2/Wnt interaction regulated murine stem and progenitor populations in vitro in hematopoietic ES cell assays and in vivo following transplantation. The relationship between PGE2 and Wnt was also conserved during regeneration of other organ systems. Our work provides in vivo evidence that Wnt activation in stem cells requires PGE2, and suggests the PGE2/Wnt interaction is a master regulator of vertebrate regeneration and recovery.

Figure 1. Prostaglandin levels directly affect wnt activity in zebrafish embryos

(A–C )In situ hybridization for GFP in TOP:dGFP wnt reporter embryos at 36hpf shows widespread wnt activity; inset, close-up of GFP+ (black arrowheads) cells in the AGM. 10μM dmPGE2 enhanced GFP expression throughout the embryo, while 10μM indo decreased global wnt activity, and in the AGM. (D) Quantification of total GFP+ cells in the major trunk vessels and (E) qPCR analysis for GFP in whole embryo lysates following exposure to dmPGE2 or indo versus vehicle con reveals significant alterations in wnt activity (*significant (sig) across treatment groups, ANOVA, p<0.05, n=10/treatment). (F–H) Representative confocal microscopy images of the AGM region of TOP:dGFP; lmo2:DsRed embryos following exposure to con, dmPGE2, or indo are shown; differences can be seen in the wnt-active (green, left column), HSC/endothelial (red, middle), and colocalized (merged, right) populations. (I–K) Representative FL1 (green)/FL2 (red) FACS plots of individual TOP:dGFP; lmo2:DsRed embryos after exposure to con, dmPGE2 or indo confirm the confocal analyses (summarized in Sup. Fig. 2).

Figure 2. PGE2 regulates Wnt-mediated effects on HSC formation at the level of β-catenin

(A–D) wnt8 induction in hs:wnt8-GFP transgenic embryos increased runx1/cmyb+ HSCs compared to WT; indo decreased HSCs in WT and wnt8 embryos. (E–L) Induction of dkk diminished runx1/cmyb expression compared to WT. dmPGE2 enhanced HSCs in WT embryos, and rescued runx1/cmyb expression to approximately WT levels in dkk embryos. Axin (hs:axin-GFP) and dnTCF (hs:dnTCF-GFP) reduced runx1/cmyb expression, and dmPGE2 could not rescue those effects. (M) Schematic of confocal microscopy. Imaging was performed in the trunk/tail region of the embryo, centered around the tip of the yolk sac extension (YSE, blue arrowhead), encompassing the dorsal aorta (red dots) and vein (blue dots), as indicated by the pink bracket. (N–Y) In vivo confocal microscopy of wnt pathway inducible embryos crossed into the cmyb:GFP HSC reporter line confirmed the in situ hybridization analysis and demonstrated quantifiable effects on cell number. (Z) Cell counts (5 embryos/treatment, data represented as mean ± SD) were conducted in the major vessels (pink bracket) in a 40x field centered at the YSE; *sig vs con; **sig vs wnt8; *** sig vs dkk; ANOVA, p<0.001.

Figure 3. PGE2-mediated modulation of wnt signaling affects cell death and proliferation

(A,B) Indo treatment enhanced TUNEL+ cells in the AGM in both WT and wnt8 embryos. dkk, axin, or dnTCF enhanced apoptosis; dmPGE2 improved this effect only in dkk transgenics. Cell counts (5 embryos/treatment) in the AGM showed significant effects across treatment groups: *sig vs con; **sig vs wnt8; ***sig vs dkk; ANOVA; p<0.05. (C,D) wnt8 enhanced BrdU incorporation, while indo diminished BrdU in both WT and wnt8 embryos. wnt inhibition by dkk, axin, or dnTCF diminished BrdU incorporation; the effect of dkk could be rescued by dmPGE2. *sig vs con; **sig vs wnt8; ***sig vs dkk; ANOVA; p<0.05.

Figure 4. PGE2 regulates the effects of wnt activity on HSCs via cAMP/PKA signaling

(A,B,E,F; I,J,M,N) The/PGE2/wnt interaction affected runx1/cmyb+ HSC formation, as seen in Fig. 2. (C,G) cAMP enhancement by forskolin increased runx1/cmyb expression in WT embryos and further expanded HSCs in wnt8 embryos. (D,H) Forskolin counteracted the inhibitory effect of indo on HSC formation in WT and wnt8 embryos. (K,O) PKA inhibition by H89 decreased runx1/cmyb expression in WT embryos and eradicated HSC formation in dkk embryos. (L,P) The enhancing effect of dmPGE2 on HSCs was reduced back to baseline levels by PKA inhibition; similarly the dmPGE2-induced rescue of HSCs in dkk embryos was blocked by H89.

Figure 5. The PGE2/wnt interaction is conserved in HSC regeneration after injury

(A) wnt activity in the kidney marrow was enhanced in FACS profiles on day 3 post irradiation (dpi) in TOP:dGFP fish; representative FACS analyses are shown on the left, and summarized (mean ± SD) on the right. dmPGE2 further increased and indo inhibited wnt activity; *sig vs unirradiated con, **sig vs irradiated con, ANOVA, p<0.001, n=5 fish/treatment. (B) wnt8 (bottom left panel) enhanced the precursor population (red), while indo suppressed that effect significantly; *sig vs con, **sig vs wnt8, t-test, p<0.02, n=10–15. (C) dkk reduced progenitor recovery, which was rescued by dmPGE2; *sig vs con, ***sig vs dkk, ANOVA, p<0.05, n=6.

Figure 6. PGE2 influenced wnt-mediated repopulation of murine HSC and progenitors

(A) Ex vivo treatment of purified C57Bl/6 KSL cells with indo, BIO, and/or dmPGE2 prior to transplantation into lethally irradiated mice revealed that PGE2 modulation significantly impacts hematopoietic progenitors. BIO significantly enhanced colony formation, which could be suppressed by indo. dmPGE2 further increased the effect of BIO; *sig vs con, ANOVA, p<0.05, n≥7. (B) Effect of ex vivo treatment with dmPGE2 or indo on CFU-S₁₂ colony formation in KSL cells from APC^Min mice. Genetic activation of wnt signaling in APC^Min hematopoietic progenitor cells has comparable effects on CFU-S₁₂ colony formation to chemical stimulation by BIO, and indo exposure blocks this enhancement; *sig vs con, ANOVA, p<0.05, n≥7. (C,D) Wnt activation through the GSK3β inhibitor BIO or in APC^Min marrow enhanced chimerism rates at 24 weeks; each effect could be inhibited by indo. Test cell chimerism of individual mice is shown (C), with the mean/group indicated by a solid black line. The dashed black line demonstrates the 5% cut-off value used to determine engraftment frequencies (D); *sig vs con, Fisher’s exact, p=0.045; n≥8.

Figure 7. The PGE2/wnt interaction is a master regulator of liver regeneration

(A–D) Representative photomicrographs of en bloc dissections following liver resections at day 3 are shown; the liver is highlighted by a yellow dotted line, the resection site by a blue line, and the black arrow indicates the amount of liver regrowth. Wnt activation in APC mutant zebrafish enhanced liver regeneration compared to WT. Indo stymied liver regeneration. (E) Quantification of zebrafish liver regeneration showed significant differences across treatment groups; *sig vs con, **sig vs APC^+/−, ANOVA, p<0.001, n≥6. (F) qPCR for cyclin D1 gene expression; effects are coordinately regulated by the PGE2/wnt interaction; *sig vs con, **sig vs APC, ANOVA, p<0.001, n=7. (G–J) wnt and PGE2 modulation has significant effects on murine liver regeneration following 2/3 partial hepatectomy. APC^Min mice exhibit enhanced β-catenin staining (left panel), particularly in the periportal areas (top middle panel), with noticeable nuclear staining (bottom middle panel). BrdU incorporation (top right panel) and cyclin D1 staining (bottom right panel) indicated enhanced regenerative activity. Indo diminished global β-catenin staining (left and top middle panels), excluded β-catenin from the nuclei (bottom middle panels), and resulted in a corresponding decrease of both BrdU incorporation and cyclin D1. (K,L) Quantification of BrdU incorporation and cyclinD1 staining in corresponding serial sections of regenerating livers; *sig vs con, **sig vs APC, ANOVA, p<0.05, n=10 sections/treatment.