Conditional deletion of Sox17 reveals complex effects on uterine adenogenesis and function

Abstract

The importance of canonical Wnt signaling to murine uterine development is well established. Mouse models in which uterine-specific Wnt ligands, β-catenin, or Lef1 are disrupted result in failure of postnatal endometrial gland development. Sox17 is a transcription factor characterized in numerous tissues as an antagonist of Wnt signaling. Thus, we hypothesized that conditional ablation of Sox17 would lead to hyperproliferation of endometrial glands in mice. Contrary to our prediction, disruption of Sox17 in epithelial and stromal compartments led to inhibition of endometrial adenogenesis and a loss of reproductive capacity. Epithelium-specific Sox17 disruption resulted in normal adenogenesis although reproductive capacity remained impaired. These findings suggest that non-epithelial, Sox17-positive cells are necessary for adenogenesis and that glands require Sox17 to properly function. To our knowledge, these findings are the first to implicate Sox17 in endometrial gland formation and reproductive success. The data presented herein underscore the importance of studying Sox17 in uterine homeostasis and function.

Keywords: Adenogenesis; Development; Endometrial glands; Mouse model; Sox17; Uterus.

Figure 1. Expression of Sox17 in the mouse uterus is dramatically decreased in Pgr-Sox17 and Sprr2f-Sox17 cKO mice

(A–B, D) Immunohistochemical labeling of Sox17 in wildtype (A), Pgr-Sox17 cKO (B), and Sprr2f-Sox17 cKO (D) adult mice. Arrows indicate isolated cells labeling positive for Sox17. Control, n=5; Pgr-Sox17 cKO, n=5; Sprr2f-Sox17 cKO, n=4. (C) qRT-PCR demonstrates a decrease in Sox17 mRNA expression in adult Pgr- and Sprr2f-Sox17 cKO mice. Data are presented as fold-change in 2^−ΔΔCt relative to the lowest expressing Pgr-Sox17 or Sprr2f-Sox17 cKO sample. Control, n=3; Pgr-Sox17 cKO, n=4; Sprr2f-Sox17 cKO, n=5. Lines indicate median. *P = 0.011. LE, luminal epithelium; GE, glandular epithelium. H&E scale bar: 100 μm, unless otherwise indicated.

Figure 2. Pgr-Sox17 cKO mice have severely reduced endometrial gland development while Sprr2f-Sox17 cKO mice do not

(A–F′) H&E photomicrographs of postnatal day (PND) Pgr-Sox17 cKO mice. PND 9 (A, B, A′, B′), PND 12 (C, D, C′, D′), PND 28 (E, F, E′, F′). For each age/genotype, n ≥ 3. Notice luminal invagination (arrow, B); glandular budding (arrow, D); stromal glands (“GE” arrow, F) in control mice. Luminal invagination (arrow, B′) and glandular budding (arrow, D′) are present but stromal integration is reduced in Pgr-Sox17 cKO mice. (G–I) H&E photomicrographs of 8-week-old mice. The endometrial compartment of control and Sprr2f-Sox17 cKO mice contain glands (G, I) while Pgr-Sox17 cKO mice do not (H). (J, K) Number of glands per micrographic field in PND 28 and 8-week-old adult mice, respectively. For all groups, n=4. Lines indicate median. **P ≤ 0.01; ****P ≤ 0.0001. LE, luminal epithelium; GE, glandular epithelium; Myo, myometrium. H&E scale bar: 100 μm.

Figure 3. mRNA expression of Sox17, Wnt7a, Lef1, and Foxa2 are altered in the neonatal Pgr-Sox17 cKO and adult Sprr2f-Sox17 cKO mouse uteri

qRT-PCR data of genes necessary for endometrial adenogenesis. For each age/genotype group, n ≥ 3. Data are presented as fold-change in 2^−ΔΔCt relative to a composite mixture of 4 cDNA samples from PND 28 control mice (A) or adult control mice (B). Lines indicate median. *P ≤ 0.05; **P ≤ 0.01; ***P≤ 0.0003; ****P ≤ 0.0001.

Figure 4. Immunofluorescence reveals cytokeratin-negative, Sox17;Pgr-dual positive cells within the stromal compartment of wildtype and Sprr2f-Sox17 cKO mouse uteri

Wildtype adult uterus (A–D). Sprr2f-Sox17 cKO uterus(E–G). Prg-Sox17 cKO uterus lacking glands both glands and stromal cytokeratin-negative, Sox17;Pgr-dual positive cells (H–K). Clustering phenotype of cytokeratin-negative, Sox17;Pgr-dual positive cells in wildtype, Sprr2f-Sox17 cKO, and incomplete deletion sections of Pgr-Sox17 cKO uteri (L–O). LE, luminal epithelium; GE, glandular epithelium. Cytokeratin = red. Sox17 = green. Pgr = magenta. DAPI = blue. Arrows indicate cytokeratin-negative, Sox17;Pgr-dual positive cells within uterine stroma (A–G) and region of incomplete luminal deletion of Sox17 with in the Pgr-Sox17 cKO mouse (I–J). For each genotype, n ≥ 4.

Figure 5. Proposed model for Sox17 regulation of adenogenesis in endometrial tissue

In the canonical Wnt signaling pathway, Wnt ligands stimulate the activation of β-catenin to induce transcription of Wnt target genes via interaction with the Lef1 transcription factor promoter. The absence of adenogenesis in the Pgr-Sox17 cKO mouse and presence of it in the Sprr2f-Sox17 cKO mouse suggests that Sox17 in isolated cells within the stromal compartment of the endometrium is necessary for adenogenesis while Sox17 within the epithelial compartment is dispensable. Depending on the cell type expressing Sox17 within the stroma, the mechanism by which Sox17 effects adenogenesis could be non-cell autonomous via target gene signaling (A) or cell autonomous via differentiation of a stem cell population into epithelial cells (B).

Figure 6. Label retaining assay in adult wildtype uterine tissue suggests Sox17-positive, cytokeratin-negative cells within the stromal compartment serve a supportive role in adenogenesis

Wildtype adult uterus (A–D). Sox17-positive, cytokeratin negative cells with a clustering phenotype are closely associated with label retaining cells. Cytokeratin = red. Sox17 = green. CldU = blue. Circles indicate a cluster of two or more Sox17-positive cells that are adjacent to CldU positive cells. n = 2.