Decidual PTEN expression is required for trophoblast invasion in the mouse

Abstract

Trophoblast invasion likely depends on complex cross talk between the fetal and maternal tissues and may involve the modulation of phosphatidylinositol 3-kinase (PI3K)/AKT signaling activity in maternal decidual cells. In this report, we studied implantation in Pten(tm1Hwu/tm1Hwu);Amhr2(tm3(cre)Bhr/+) mice, which lack the PI3K signaling antagonist gene Pten in myometrial and stromal/decidual cells. Primiparous Pten(tm1Hwu/tm1Hwu);Amhr2(tm3(cre)Bhr/+) mice were found to be subfertile because of increased fetal mortality at e11.5. Histopathological analyses revealed a failure of decidual regression in these mice, accompanied by reduced or absent invasion of fetal trophoblast glycogen cells and giant cells, abnormal development of the placental labyrinth, and frequent apparent intrauterine fetal growth restriction. Unexpectedly, the loss of phosphate and tensin homolog deleted on chromosome 10 (PTEN) expression in Pten(tm1Hwu/tm1Hwu);Amhr2(tm3(cre)Bhr/+) decidual cells was not accompanied by a detectable increase in AKT phosphorylation or altered expression or activation of PI3K/AKT downstream effectors such as mammalian target of rapamycin or glycogen synthase kinase-3β. Terminal deoxynucleotidyl transferase-mediated nick end labeling and bromodeoxyuridine incorporation analyses attributed to the lack of decidual regression mainly to decreased apoptosis in Pten(tm1Hwu/tm1Hwu);Amhr2(tm3(cre)Bhr/+) decidual cells, rather than to increased proliferation. Remodeling of the maternal vasculature was delayed in Pten(tm1Hwu/tm1Hwu);Amhr2(tm3(cre)Bhr/+) uteri at e11.5, as evidenced by persistence of vascular smooth muscle and decreased infiltration of uterine natural killer cells. In addition, thickening of the myometrium and disorganization of the muscle fibers were observed before and throughout gestation. Almost all Pten(tm1Hwu/tm1Hwu);Amhr2(tm3(cre)Bhr/+) mice failed to carry a second litter to term, apparently attributable to endometrial hyperplasia and uterine infections. Together, these data demonstrate novel roles of PTEN in the mammalian uterus and its requirement for proper trophoblast invasion and decidual regression.

Fig. 1.

Amhr2^tm3(cre)Bhr mice exhibit Cre activity in myometrial and decidual cells and in the blood vessels. Histological sections of uteri from d11.5 pc pregnant Gt(ROSA)26Sor^tm1Sho/+ (A, control) and Gt(ROSA)26Sor^tm1Sho/+; Amhr2^{tm3(cre)Bhr/+} mice (B–D) were stained for β-galactosidase activity (blue). Results showed Cre activity in the myometrium (M, B), decidual cells (D, C; also faintly visible in B), and blood vessels (arrow, D) in Gt(ROSA)26Sor^tm1Sho/+;Amhr2^{tm3(cre)Bhr/+} uteri. Original magnification, ×400 (A and B) and ×630 (C and D).

Fig. 2.

Loss of phosphate and tensin homolog deleted on chromosome 10 (PTEN) expression in stromal, decidual, and myometrial cells in Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} mice. Immunohistochemical analysis of PTEN expression in uteri of nonpregnant (A and B) and d11.5 pc (C and D) Pten^{tm1Hwu/tm1Hwu} and Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} mice showing a loss of PTEN expression in the myometrium, uterine stroma, maternal decidual cells (MD), and some uterine glandular epithelial cells (B, inset). The same pattern of expression was also observed in Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} uteri at d9.5 and d14.5 pc (data not shown). Loss of PTEN expression did not result in appreciable differences in phospho-AKT or phospho-mammalian target of rapamycin (mTOR) expression in Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} mice (F and H) at d11.5 pc, compared with controls (E and G). FTGC, fetal trophoblast giant cells; L, labyrinth. Original magnification, ×100 (A–H) and ×400 (B, inset).

Fig. 3.

Representative immunoblot analyses of Pten^{tm1Hwu/tm1Hwu} and Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} uterine wall samples from nonpregnant mice and at d11.5 pc. In pregnant mice, uterine tissues surrounding healthy and nonviable fetuses were analyzed separately. Densitometric analyses were performed using n = 4 samples per pregnancy/viability status and per genotype and normalized to corresponding β-actin (ACTB) loading control values before statistical analyses. Comparisons between genotypes for each pregnancy/viability status failed to show statistically significant differences for any of the proteins examined, except for PTEN as mentioned in the text. Arrows indicate bands of interest. GSK, glycogen synthase kinase.

Fig. 4.

Defects in trophoblast invasion and maternal decidual regression in Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} mice. A: photomicrograph of a Pten^{tm1Hwu/tm1Hwu} conceptus at d11.5 pc showing a well-developed labyrinth and spongiotrophoblast (S), normal migration of the FTGC, and a regressing MD. B: photomicrograph of a Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} conceptus at d11.5 pc showing an underdeveloped labyrinth, reduced trophoblast migration, and a persistent MD. HPS, hematoxylin-phloxin-saffron. C: cytokeratin immunohistochemical analysis of a Pten^{tm1Hwu/tm1Hwu} conceptus at d14.5, showing normal invasion of trophoblast glycogen cells (border indicated with a green dashed line) beyond the trophoblast giant cell border (red dotted line). Glycogen cells are seen as negative images (gaps) in the immunohistochemistry specimen. D: same as C, in a Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} conceptus. Note the failure of the glycogen cells to invade the decidua. E and F: alkaline phosphatase staining (blue) of Pten^{tm1Hwu/tm1Hwu} and Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} decidual cells. Bromodeoxyuridine (BrdU) incorporation (G and H) and terminal deoxynucleotidyl transferase-mediated nick end labeling (TUNEL) analysis (I and J) of Pten^{tm1Hwu/tm1Hwu} and Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} uteri at d11.5 pc showing increased proliferation and reduced apoptosis in the maternal decidual cells. Arrows indicate examples of BrdU- and TUNEL-positive cells, and these are also shown at higher magnification in the subpanels. Original magnification, ×50 (A and B), ×100 (C–J), ×1,000 (G–J, inset).

Fig. 5.

Decreased uterine natural killer cell (uNK) infiltration and poorly transformed blood vessels in the maternal decidua of Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} mice. Transgelin (TAGLN) immunohistochemistry (A and B) and Dolichos biflorus agglutinin lectin analyses (C, D) on uteri from d11.5 pc mice, showing decreased numbers of uNK cells and poorly transformed blood vessels (arrowheads, also shown at higher magnification in inset) in Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} maternal decidua compared with well-transformed blood vessels (arrows, also shown at higher magnification in the inset) in the Pten^{tm1Hwu/tm1Hwu} controls. Original magnification, ×100 (A and B), ×200 (C and D), ×1,000 (A and B, inset).

Fig. 6.

Increased myometrial thickness in Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} mice. A and B: photomicrographs of d14.5 Pten^{tm1Hwu/tm1Hwu} and Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} uteri illustrating a thicker myometrium in the latter. C: histogram of myometrial thickness measurements in nonpregnant, d9.5, d11.5 and d14.5 pc Pten^{tm1Hwu/tm1Hwu} and Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} mice. Columns and error bars denote means ± SE. Asterisk denotes significant difference from control (P < 0.05). D and E: TAGLN immunohistochemical analysis of Pten^{tm1Hwu/tm1Hwu} and Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} uteri at d11.5 pc. Note the disorganized staining pattern in Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} myometrial cells. Original magnification, ×200 (A and B) or ×1,000 (D and E).

Fig. 7.

Histopathological analysis of Pten^{tm1Hwu/tm1Hwu};Amhr2^{tm3(cre)Bhr/+} uteri during the second pregnancy. Photomicrographs depict epithelial hyperplasia at d9.5 pc (A, arrow), residual maternal decidua at d9.5 pc (B, circumscribed by dashed line), necrotic maternal decidua at d22.5 pc (C, circumscribed by dashed line, arrows indicate labyrinth and fetal membranes), and mucometra at d41.5 pc (D, arrows). Original magnification, ×400 (A), ×200 (B), ×20 (C), and ×12.5 (D).