Cell-type specific analysis of physiological action of estrogen in mouse oviducts

Abstract

One of the endogenous estrogens, 17β-estradiol (E₂ ) is a female steroid hormone secreted from the ovary. It is well established that E₂ causes biochemical and histological changes in the uterus. However, it is not completely understood how E₂ regulates the oviductal environment in vivo. In this study, we assessed the effect of E₂ on each oviductal cell type, using an ovariectomized-hormone-replacement mouse model, single-cell RNA-sequencing (scRNA-seq), in situ hybridization, and cell-type-specific deletion in mice. We found that each cell type in the oviduct responded to E₂ distinctively, especially ciliated and secretory epithelial cells. The treatment of exogenous E₂ did not drastically alter the transcriptomic profile from that of endogenous E₂ produced during estrus. Moreover, we have identified and validated genes of interest in our datasets that may be used as cell- and region-specific markers in the oviduct. Insulin-like growth factor 1 (Igf1) was characterized as an E₂ -target gene in the mouse oviduct and was also expressed in human fallopian tubes. Deletion of Igf1 in progesterone receptor (Pgr)-expressing cells resulted in female subfertility, partially due to an embryo developmental defect and embryo retention within the oviduct. In summary, we have shown that oviductal cell types, including epithelial, stromal, and muscle cells, are differentially regulated by E₂ and support gene expression changes, such as growth factors that are required for normal embryo development and transport in mouse models. Furthermore, we have identified cell-specific and region-specific gene markers for targeted studies and functional analysis in vivo.

Keywords: embryo development; embryo transport; estrogen; insulin-like growth factor 1; oviduct; scRNA-seq.

FIGURE 1

Ki67-positive cells and PGR expression in the mouse oviduct after E₂ treatment. A-B, Immunohistochemical (IHC) analyses of ESR1 and Ki67 expression of the infundibulum, ampulla, and the isthmus of the oviduct, respectively. Oviducts were collected from ovariectomized mice after 24 h of treatment with Veh or E₂ (0.25 μg/mouse), n = 3–4 mice/group. C, Quantification of Ki67-positive ciliated, secretory and stromal cells. D, Transcript levels of Pgr in the whole oviduct after 24 h of Veh or E₂ treatment (n = 4–9 mice/group). E. IHC analysis of PGR expression at the infundibulum, the ampulla and the isthmus after 24 h of treatment with Veh or E₂. Insets are enlarged images of area with PGR⁺ ciliated cells (arrowheads) in the ampulla. F. Quantification of PGR-positive ciliated, secretory and stromal cells. G. Transcript levels of Igf1 in the whole oviduct after 24 h of Veh or E₂ treatment (n = 4–9 mice/group). Scale bars = 25 μm. Each dot represents data from each mouse. *P < .05, significantly different when compared to the corresponding Veh, unpaired Student’s t-test with Welch’s correction

FIGURE 2

E₂ treatment alters the transcriptional profile of the oviduct. A, Cell singlets were isolated from either the “infundibulum + ampulla” (InfAmp) or “isthmus + uterotubal junction (IsthUTJ) and pooled from mice that were ovariectomized and treated with Veh or E₂ for 24 h (n = 5–6 mice/treatment). Single cell-RNA sequencing (scRNA-seq) data presented as Uniform Manifold Approximation and Projection (UMAP) plot. Each dot represents one cell. There are seven distinct oviductal cell clusters (#0–6) as indicated by different colors. B, Expression of the top marker gene from each cell cluster. C-D, All cell clusters differentiated by the region (InfAmp vs. IsthUTJ) or the treatment (Veh vs. E₂). E, Top biological processes using gene ontology analysis from genes enriched in all clusters after E₂ treatment. Numbers in brackets indicate P-values. F, Dot plots of genes enriched in cellular response to steroid hormone stimulus in all clusters. Percentage of cells expressing particular gene(s) in the dataset is visualized by the size of the dot. Mean expression within each category is visualized by color. G, Top biological processes from genes enriched in 0/ciliated cell cluster after E₂ treatment. H, Dot plots of genes enriched in blood vessel development in cluster 0/ciliated cells. I, Top biological processes from genes enriched in 1 and 2/secretory cell clusters after E₂ treatment. J, Dot plots of genes enriched in cytoplasmic translation in clusters 1 and 2/secretory cells

FIGURE 3

Exogenous and endogenous E₂ induce differential transcriptional signatures in the oviduct. Endogenous E₂ single cell samples were collected from the entire oviduct at the estrus stage (n = 5 mice). scRNA-seq data from estrus (endogenous E₂) were combined with data from ovariectomized E₂-treated samples (exogenous E₂) shown in Figure 2 for comparison analyses. A, Estrus and E₂-treated (InfAmp and IsthUTJ) datasets were overlapped. B, After the dataset from E₂-treated and estrus samples were combined, scRNA-seq data were analyzed and separated into 7 cell clusters containing the similar cell populations to that in Figure 2. C, Overall similarity and difference between endogenous vs. exogenous E₂ in the oviductal cells. D, Dot plot of top 10 differentially expressed genes between exogenous and endogenous E₂ in the oviduct. E, Top biological processes using gene ontology analysis of gene enrichment in exogenous E₂ treatment. F, Dot plots of genes enriched in sensory perception of chemical stimulus in exogenous E₂ treatment. G, Top biological processes of genes enriched in estrus samples. H, Dot plots of genes enriched in positive regulation of type III hypersensitivity in estrus samples

FIGURE 4

Cells isolated from estrus alone show similarity of cell clusters identified in the E₂-treated dataset. A, UMAP plot of scRNA-seq analysis from estrus samples. Cells were separated into 7 cell clusters. B, Heatmap plot of top ten differentially expressed genes in each cell cluster. Color-coded bar on the left is coordinated with the cell clusters in A. In situ hybridization analyses of (C) Pdxk, (D) Serpina1e, (E) Dcn, and (F) Crabp2 transcripts in the oviduct from samples collected randomly at different stages of the estrous cycle from ovarian intact adult female mice

FIGURE 5

Unique cell populations in mouse oviducts and human Fallopian tube produce embryotrophic factor(s). A, Dot plot of embryotrophic factors and associated proteins in each cell cluster in the mouse oviduct collected at estrus. Insulin-like growth factor 1 (Igf1), epidermal growth factor (Egf), IGF receptor (Igfr), IGF-binding protein (Igfbp), fibroblast growth factor (Fgf), compliment C3 (C3), colony-stimulating factor 1 (Csf1), and demilune cell parotid proteins 1–3 (Dcpp). Different sizes of the circle represent percentages of the cells with expression of each gene. Color scale represents normalized expression value. B, scRNA-seq analysis of cell clusters from human Fallopian tube isolated from one individual as a proof-of-concept. Marker genes for each cell cluster are indicated in Figure S6. Stromal cells (decorin, DCN+) also express IGF1, FGF1, and C3. C, Expression of Igf1 in the mouse oviduct using ISH analysis. D and E, ForceAtlas2 (FA) plot, phase contrast, expression, and RNA project velocities of Igf1 from mouse oviductal cells at (D) estrus or (E) after E₂ treatment. Gray lines in the FA plot represent all velocity-inferred cell-to-cell connections/transitions but do not differentiate the directionality of velocity. Phase portraits depict a spliced vs. unspliced Igf1 mRNA. Black dotted line represents constant transcriptional state. The expression of Igf1 is mainly in the fibroblast cell populations, but the inferred direction of RNA velocity (projected velocity) showed the trajectory towards ciliated (InfAmp), secretory (InfAmp), and secretory (IsthUTJ) cell populations

FIGURE 6

Oviductal ablation of Igf1 increases developmentally delayed and nonviable embryos in mouse oviducts and inhibits embryo transport. A, Igf1 transcripts in Igf1^f/f (control) compared to Pgr^Cre/+; Igf1^f/f oviducts. B, Embryos collected from the uterus and/or oviduct from Igf1^f/f compared to Pgr^Cre/+; Igf1^f/f females at 3.5 days post coitus (dpc). C, Number of embryos at each developmental stage at 3.5 dpc. D, Number of developmentally delayed or nonviable embryos. P = .0595, unpaired Student’s t-test with Welch’s correction. E, Total number of all eggs or embryos present in the oviduct at 3.5 dpc. F, Number of embryos collected from Igf1^f/f (left bar) compared to Pgr^Cre/+; Igf1^f/f (right bar) at each region of the oviduct at 3.5 dpc. *P < .05, significantly different when compared to the corresponding Igf1^f/f mice, unpaired Student’s t-test with Welch’s correction. G, Percentage of mice with developmentally delayed or nonviable embryos in the oviduct. *P < .05, significantly different when compared to the corresponding Igf1^f/f group, Fisher’s exact test for contingency plot. n = 9–12 mice/genotype