Fig 1.
Spatial and temporal contribution of BMDCs to decidua throughout mouse pregnancy.
(A) Biodistribution of BM-derived (GFP+) cells showing preferential recruitment to the pregnant uterus as compared with other body organs. Top panel, PBS nonpregnant control; middle panel, nonpregnant BMT; bottom panel, BMT at E9.5. (B) Engraftment of BMDCs (green) in uteri in nonpregnant state and throughout pregnancy in WT mice receiving BMT from GFP donors. PBS-injected pregnant mouse (E9.5) is shown as negative control, and GFP transgenic mouse is shown as positive control. White arrows indicate the preferential localization of BMDCs to the implantation site. (C) Uterine tissue sections of pregnant mice stained with anti-GFP antibody (brown) showing the localization of BMDCs in decidua during the course of pregnancy. Yellow arrows point to the maternal–fetal interface demarcating the maternal decidua basalis (DB) from fetal placenta (P). Scale bars, 100 μm (top panel) and 50 μm (bottom panel). (D) Representative graphs of flow cytometry of uterine cells demonstrating the temporal changes in hematopoietic (CD45+) and nonhematopoietic (CD45−) GFP+ BMDCs populations during the course of pregnancy. Numbers in each quadrant indicate percentage of cells. (E) Immunofluorescence of uterine tissue sections showing colocalization of PCNA-positive proliferating cells (red) and GFP-positive BMDCs (green) during the course of pregnancy; sections were counterstained with DAPI (blue). Scale bar, 50 μm. (F) Summary of flow cytometry analysis of percentage of GFP+ cells in the uterus during pregnancy (n = 5–7). **p ≤ 0.01 versus E9.5, E13.5, and E17.5. *p ≤ 0.05 versus E9.5. (G) Summary of flow cytometry analysis of percentage of GFP+ and GFP− cells in the uterus that are either CD45+ or CD45− during the course of pregnancy (n = 5–7). *p < 0.01, **p < 0.001. (H) Quantification of proliferating PCNA+GFP+ BMDCs in the uterus during the course of pregnancy (n = 4–7). All bar graphs are mean ± SEM. *p ≤ 0.05. See also S1 Fig. Underlying data are available in S1 Data. BM, bone marrow; BMDC, BM-derived cell; BMT, BM transplant; DB, decidua basalis; E, embryonic day; GFP, green fluorescent protein; NP, nonpregnant; P, fetal placenta; PCNA, proliferating cell nuclear antigen; WT, wild-type.
Fig 2.
Characterization of uterine BMDCs throughout pregnancy.
(A) Flow cytometry analysis of BMDCs in uterine implantation site on E9.5. Cells gated in R2 are BM derived (GFP+) uterine cells, while cells gated in R1 are non–BM-derived resident (GFP−) uterine cells. Histograms represent GFP+ cells (R2) from E9.5 implantation sites that are stained with the indicated antibodies (green line) and respective isotype controls (filled) (n = 4). (B) Quantification of percentage of BM-derived (GFP+) and non–BM-derived (GFP−) uterine cells expressing the various cell surface markers shown in (A) (n = 4), *p < 0.05. (C) Flow cytometry analysis of E9.5 BM-derived (GFP+) uterine cells (R2) using CD45 in combination with various surface marker antibodies. (D) Quantification of surface markers shown in (C), on BM-derived nonhematopoietic decidual cells (GFP+CD45−) and BM-derived hematopoietic cells (GFP+CD45+) (n = 4–5), *p < 0.01. (E, F, G, H, K) Immunofluorescence photomicrographs of E9.5 mesometrial decidua sections or nonpregnant mice uteri sections demonstrating co-staining of GFP+ BMDCs (green) with (E) CD45 (red), (F) CD31 (red), (G) Sca-1 (red), (H) DPRP (red), (K) progesterone receptor (PR) (pink), and CD45 (yellow). Sections were counterstained with DAPI showing nuclei (blue). Insets show higher magnification photomicrographs. White arrows point to BMDCs colocalizing with their respective markers. White dashes are encircling endometrial glands. Scale bars, 50 μm. (L and M) A z-stack series (L) and a 3D image (M) of the inset from (K) demonstrating a single BM-derived GFP+ cell co-expressing PR but negative for CD45 (white arrow). (I) Quantification of BMDCs (GFP+) or non-BMDCs (GFP−) positive for DPRP throughout gestation (n = 3). *p ≤ 0.01. (J) Quantification of BMDCs (GFP+) or non-BMDCs (GFP−) positive for PR throughout gestation (n = 3–5). *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. In all panels, bar graphs represent mean ± SEM. See also S3–S5 Figs. Underlying data are available in S1 Data. BM, bone marrow; BMDC, BM-derived cell; DPRP, decidual prolactin-related protein; GFP, green fluorescent protein; PR, progesterone receptor; Sca-1, stem cell antigen-1; SSC, side scatter; VEGFR2, vascular endothelial growth factor receptor 2.
Fig 3.
Co-staining of BMDCs with NK cell markers.
(A and B) Low-magnification immunofluorescence photomicrographs of E9.5 decidua, showing (A) GFP+ BMDCs (green) found predominantly in the mesometrial side but also on the antimesometrial side, while (B) DBA-lectin-positive NK cells (red) are found exclusively on the mesometrial side. (C) Quantification of DBA surface marker on BM-derived (GFP+) cells and non–BM-derived (GFP−) cells on E9.5 (n = 4). (D-F) Higher-magnification immunofluorescence photomicrographs of (D) E9.5 mesometrial decidua, (E) antimesometrial decidua, or (F) nonpregnant mice uteri sections demonstrating co-staining of a subset of GFP+ BMDCs (green) with NK cell–specific stain DBA-lectin (red). Sections were counterstained with DAPI showing nuclei (blue). The images on the right of each panel are higher magnification of the corresponding dashed areas. (G) Uterine sections from nonpregnant, E 9.5, or ED 13.5 pregnant mice co-stained with PAS (purple) and GFP antibody (brown) showing PAS+ NK cells (black arrow), BM-derived NK cells (red pointer) and other non-NK BMDCs (black pointer). Inset shows a higher magnification of the same photomicrograph. Scale bar, 20 μm. Bar graphs represent mean ± SEM. *p < 0.0001. Underlying data are available in S1 Data. BM, bone marrow; BMDC, BM-derived cell; DBA, Dolichos biflorus agglutinin; GFP, green fluorescent protein; NK, natural killer; PAS, Periodic Acid Schiff.
Fig 4.
BM transplantation from WT donors prevents pregnancy loss in Hoxa11+/− mice.
BMTs were performed from WT into Hoxa11+/− (Hoxa11+/−WT BMT), WT into HoxA11−/− (Hoxa11−/−WT BMT), Hoxa11−/− into Hoxa11+/− (Hoxa11+/−KO BMT), and HoxA11−/− into Hoxa11−/− (Hoxa11−/−KO BMT). WT mice receiving BMT from WT donors (WTWT BMT) served as controls. (A-C) Effect of BMT from different genotypes on (A) litter size, (B) time to delivery, and (C) mean pup weight per litter following breeding (n = 11–15/group). *p < 0.01. +p < 0.05 versus Hoxa11+/−KO BMT. (D and E) Representative uterine images (D) and quantitation of mean number of implantations per mouse on E5.5 (E) in WT, Hoxa11+/−, and HoxA11−/− mice that received BMT from either WT or KO donors (n = 10–16/group). (F-I) Pregnancy resorptions on ED 12.5 in WT and Hoxa11+/− mice that received BMT from either WT or KO donors (n = 10–12/group). (F) Representative images of uteri showing resorption sites (black arrows). (G) Mean number of viable implantations per mouse. (H) Resorption rate per mouse pregnancy. (I) Percentage of dams with at least one resorption. In all panels, bar graphs represent mean ± SEM. *p ≤ 0.05, **p ≤ 0.01. Underlying data are available in S1 Data. BM, bone marrow; BMT, BM transplant; Hoxa11, Homeobox a11; KO, knockout; WT, wild-type.
Fig 5.
BM transplantation from WT donors normalizes the uterine transcriptome of Hoxa11+/− mice.
(A-C) RNA-seq analysis of implantation sites on E5.5 in Hoxa11+/−WT BMT, Hoxa11+/−KO BMT, and WTWT BMT. (A) Heat map of the expression fold change of 498 commonly DEGs in the comparisons of Hoxa11+/−KO BMT versus Hoxa11+/−WT BMT, and Hoxa11+/−KO BMT versus WTWT BMT. (B) The top 10 KEGG pathways enriched in genes commonly differentially expressed in the comparisons of Hoxa11+/−KO BMT versus Hoxa11+/−WT BMT, and Hoxa11+/−KO BMT versus WTWT BMT. (C) Three-way Venn diagram of the comparisons of Hoxa11+/−KO BMT versus Hoxa11+/−WT BMT, Hoxa11+/−KO BMT versus WTWT BMT, and Hoxa11+/−WT BMT versus WTWT BMT. (D) RT-PCR validation of gene expression of select genes from the 498 commonly DEGs that have important roles in decidualization. *p ≤ 0.05 and **p ≤ 0.01 versus Hoxa11+/−KO BMT (n = 3–5/group). (E-H) Hoxa11 expression in implantation sites on ED5.5 in Hoxa11+/−WT BMT, Hoxa11+/−KO BMT, and WTWT BMT (n = 5–6/group). (E) Hoxa11 mRNA expression and (F) Hoxa11 protein expression in ED5.5 implantation sites. (G and H) Sections of implantation sites from ED5.5 stained with Hoxa11 antibody (brown) (H). Inset (below) shows high magnification photomicrograph of the dashed area and the embryo (e). The primary decidual zone (PDZ) and secondary decidual zone (SDZ) are noted. (G) Corresponding quantification of Hoxa11-positive cells in the PDZ. Bar graphs represent mean ± SEM. *p ≤ 0.05, **p ≤ 0.01. Scale bars, 200 μm (upper panel), 100 μm (lower panel). See also S12 and S16 Fig. Underlying data are available in S1 Data. BM, bone marrow; BMT, BM transplant; DEG, differentially expressed gene; Hoxa11, Homeobox a11; KEGG, Kyoto Encyclopedia of Genes and Genomes; KO, knockout; PDZ, primary decidual zone; RNA-seq, RNA sequencing; SDZ, secondary decidual zone; WT, wild-type.
Fig 6.
BM transplantation from WT donors leads to stromal expansion and induces glandular formation in HoxA11-null mice.
(A and B) Uterus photographs (A) and mean uterine weight (B) of nonpregnant KOWT BMT and KOKO BMT mice. (C) HE and cytokeratin immunostaining (brown) of uterine sections from KOWT BMT and KOKO BMT mice. The endometrial area is encircled in the left panel and shown at higher magnification in the middle (HE) and right (cytokeratin) panels. Glands are seen only in KOWT BMT mice and are surrounded by black dashed circles in the middle panel, corresponding to the cytokeratin-positive (brown) areas in the right panel. The luminal epithelium (Lu) is encircled by a red dash. Scale bars, 100 μm (left) and 50 μm (middle and right panels). (D-F) Quantification of total stromal area (μm2) (D), percent stromal area out of endometrial area (E), and percent mice with endometrial glands (F) in KOWT BMT and KOKO BMT groups. n = 6–7/group. (G) GFP immunostaining of uterine sections from KOWT BMT showing localization of GFP+ BMDCs (brown) in the stroma. The Lu) is encircled by a red dash and the gland is encircled by black dash. Bar graphs represent mean ± SEM. *p ≤ 0.05, **p ≤ 0.01. See also S15 Fig. Underlying data are available in S1 Data. BM, bone marrow; BMDC, BM-derived cell; BMT, BM transplant; GFP, green fluorescent protein; HE, hematoxylin–eosin; Hoxa11, Homeobox a11; KO, knockout; Lu, luminal epithelium; WT, wild-type.
Fig 7.
BM transplantation from WT donors leads to decidualization reaction in Hoxa11-null mice.
(A-H) Sections of E5.5 implantation sites from WTWT BMT, KOWT BMT, and KOKO BMT stained with HE (A and B), CD31 (C and D), PR (E and F), and PCNA (G and H). B, D, F, and H are higher magnification photomicrographs of A, C, E, and G, respectively. Black arrows point to embryos. (J, K and L) Quantification of pecent PR-positive cells (J), percent PCNA-positive cells (K), and mean blood vessel luminal area (μm2) in the endometrium (n = 3–4 mice/group). (I) Sections of E5.5 implantation sites from WTWT BMT, KOWT BMT, and KOKO BMT co-stained with Hoxa11 (green), CD45 (red), and counterstained with DAPI (blue). Arrows point to Hoxa11-expressing cells. (M) Relative uterine mRNA expression of HoxA11 and implantation-related genes Prl8a2, Lif, and Msx1 normalized to GAPDH in WTWT BMT versus KOWT BMT versus KOKO BMT on E5.5 (n = 4–5 mice/group). (N) Uterine Hoxa11 protein expression in KOWT BMT versus KOKO BMT on E5.5. Bar graphs represent mean ± SEM. p-values are noted on the graphs. Scale bars, 100 μm (A, C, E, G), 50 μm (B, D, F, H). Underlying data are available in S1 Data. BM, bone marrow; BMT, BM transplant; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HE, hematoxylin–eosin; Hoxa11, Homeobox a11; KO, knockout; PCNA, proliferating cell nuclear antigen; PR, progesterone receptor; WT, wild-type.