Theophylline derivatives promote primordial follicle activation via cAMP-PI3K/Akt pathway and ameliorate fertility deficits in naturally aged mice

Abstract

In elderly women and patients with premature ovarian insufficiency (POI), activating their remaining dormant primordial follicles in vivo is challenging. In this study, we found that phosphodiesterase (PDE) subtypes were expressed mainly in primordial follicle oocytes. The specific PDE inhibitors and theophylline derivatives (aminophylline, dyphylline, and enprofylline) activated primordial follicles in neonatal mice by ovary culture and intraperitoneal injection. These inhibitors also increased the levels of ovarian cyclic adenosine monophosphate (cAMP) and oocyte phosphorylated protein kinase B (p-Akt). The blockade of gap junctions using carbenoxolone (CBX) increased the levels of ovarian cAMP and pre-granulosa cell phosphorylated mammalian target of rapamycin (p-mTOR), suggesting that oocyte PDEs hydrolyze cAMP from pre-granulosa cells through gap junctions to maintain primordial follicle dormancy. Importantly, oral aminophylline improved ovulated oocyte quantity and quality, and increased offspring numbers in naturally aged mice. In addition, theophylline derivatives also activated human primordial follicles and increased p-Akt levels. Thus, theophylline derivatives activate primordial follicles by accumulating cAMP levels and activating phosphatidylinositol 3-kinase (PI3K)/Akt pathway in oocytes, and oral aminophylline increased fertility in naturally aged female mice by improving ovulated oocyte quantity and quality. As oral medications, theophylline derivatives may be used to improve fertility in elderly women and patients with POI.

Keywords: fertility; phosphodiesterase; premature ovarian insufficiency; primordial follicle activation; theophylline derivative.

Figure 1

Effects of specific PDE inhibitors on mouse primordial follicle activation in vitro. 4 dpp mouse ovaries were collected for qRT‒PCR analysis (A-B). 3 dpp mouse ovaries were cultured in medium alone or with 10 μM nimodipine, 2.5 μM EHNA, 5 μM zaprinast, 50 μM BRL-50481, 100 nM PF-04957325, PDEI (the combination of 10 μM nimodipine, 2.5 μM EHNA, and 5 μM zaprinast), and/or 10 μM LY294002 (LY) for 1 (G, I), 2 (E-F, H, and J-K), or 4 days (C-D and L-M). A, The phosphodiesterase subtype mRNA levels in the ovaries of mice at 4 dpp. The mRNA value of Pde4d was set as 1 (n = 3 biological replicates). B, The comparison of highly expressed phosphodiesterase subtype mRNA levels in ovarian somatic cells and oocytes of 4 dpp mouse primordial follicles. *P < 0.05 and **P < 0.01 vs. ovarian somatic cells group. C-D, The comparison of ovarian morphology (C) and primordial and growing follicle number (D) across various groups. The ovarian sections were hematoxylin-stained. E, The comparison of Gdf9 and Zp3 mRNA levels between control and PDEI groups. F, The comparison of GDF9, ZP3, and DDX4 protein (F) levels between control and PDEI groups. G-H, The comparison of PCNA, BAX, BCL-2, Cleaved Caspase-3, and Ki-67 mRNA (G) and/or protein (H) levels between control and PDEI groups. I, The comparison of p-mTOR and p-Akt protein levels between control and PDEI groups. J-K, FOXO3a localization in primordial follicle oocyte nuclear (arrowheads) or cytoplasm (arrows. J) and the comparison of FOXO3a nuclear export percentage in primordial follicle oocytes (K) between control and PDEI groups. DDX4, red; FOXO3a, green. L-M, The comparison of ovarian morphology (L) and primordial and growing follicle number (M) across various groups. The representative images were displayed. PF, primordial follicle; GF, growing follicle. Red arrows, growing follicles. PF, primordial follicle; GF, growing follicle. Scale bars, 50 µm. In each experiment, n ≥ 3 biological replicates. Bars indicate the mean ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001. Two-tailed unpaired t-test was used to assess statistical significance.

Figure 2

Effects of CBX on mouse primordial follicle activation in vitro. 3 dpp mouse ovaries were cultured in medium alone or with 10 μM CBX and/or 5 mM 2DG for 1 (D, G, and L), 2 (C, E-F, H-K, and M-N), or 4 days (A-B). A-B, The comparison of ovarian morphology (A) and primordial and growing follicle number (B) between control and CBX groups. C-D, The comparison of DDX4 (C), p-mTOR and p-Akt (D) protein levels between control and CBX groups. E-F, FOXO3a localization in primordial follicle oocyte nuclear (arrowheads) or cytoplasm (arrows. E) and the comparison of FOXO3a nuclear export percentage in primordial follicle oocytes (F) across various groups. G, The comparison of Glut4, Hk1, Pfkl, Aldoa, Tpi, Eno1, Pkm2, and Ldhb mRNA levels between control and CBX groups. H-I, The comparison of HK1, PFKL, PKM2, and GLUT4 protein levels between control and CBX groups. J-K, The comparison of ovarian morphology (J) and primordial and growing follicle number (K) across various groups. L, The comparison of p-mTOR and p-Akt protein levels between control and CBX groups. M-N, FOXO3a localization in primordial follicle oocyte nuclear (arrowheads) or cytoplasm (arrows. M) and the comparison of FOXO3a nuclear export percentage in primordial follicle oocytes (N) across various groups. Red arrows, growing follicles. The ovarian sections were hematoxylin-stained. DDX4, red; FOXO3a green. The representative images were displayed. PF, primordial follicle; GF, growing follicle. Scale bars, 50 µm. In each experiment, n ≥ 3 biological replicates. Bars indicate the mean ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001. Two-tailed unpaired t-test was used to assess statistical significance.

Figure 3

Effects of theophylline derivatives on mouse primordial follicle activation in vitro. 3 dpp mouse ovaries were cultured in medium alone or with 50 μM aminophylline (Ami), 160 μM dyphylline (Dyp), or 10 μM enprofylline (Enp) for 1 (E), 2 (C-D and F-I), or 4 days (A-B). A-B, The comparison of ovarian morphology (A) and primordial and growing follicle number (B) across various groups. The ovarian sections were hematoxylin-stained. C-D, The comparison of Gdf9 and Zp3 mRNA (C) and DDX4 protein (D) levels across various groups. E-G, The comparison of PCNA, BAX, BCL-2, Cleaved Caspase-3, and Ki-67 mRNA (E) and/or protein (F, G) levels across various groups. H, PCNA, Ki-67, BrdU, and Cleaved Caspase-3 immunofluorescence stain (green) across various groups. FOXL2, red. I, The comparison of PCNA- and Ki-67-positive granulosa cell percentage and BrdU- and Cleaved Caspase-3-positive cell number across various groups. The representative images were displayed. Red arrows, growing follicles. PF, primordial follicle; GF, growing follicle. TD, theophylline derivatives. Scale bars, 50 µm. In each experiment, n ≥ 3 biological replicates. Bars indicate the mean ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001. Two-tailed unpaired t-test was used to assess statistical significance.

Figure 4

Effects of theophylline derivatives on the PI3K/Akt pathway activation in vitro. 3 dpp mouse ovaries were cultured in medium alone or with 50 μM aminophylline (Ami), 160 μM dyphylline (Dyp), 10 μM enprofylline (Enp), and/or 10 μM LY294002 (LY) for 1 (A and F-G), 2 (B-C and H-I), or 4 days (D-E). A, The comparison of p-mTOR and p-Akt protein levels across various groups. B-C, FOXO3a localization in primordial follicle oocyte nuclear (arrowheads) or cytoplasm (arrows. B) and the comparison of FOXO3a nuclear export percentage in primordial follicle oocytes (C) across various groups. D-E, The comparison of ovarian morphology (D) and primordial and growing follicle number (E) across various groups. The ovarian sections were hematoxylin-stained. F-G, The comparison of p-mTOR and p-Akt protein levels across various groups. H-I, FOXO3a localization in primordial follicle oocyte nuclear (arrowheads) or cytoplasm (arrows. H) and the comparison of FOXO3a nuclear export percentage in primordial follicle oocytes (I) across various groups. DDX4, red; FOXO3a, green. The representative images were displayed. PF, primordial follicle; GF, growing follicle. Red arrows, growing follicles. Scale bars, 50 µm. In each experiment, n ≥ 3 biological replicates. Bars indicate the mean ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001. Two-tailed unpaired t-test was used to assess statistical significance.

Figure 5

Effects of PDEI and theophylline derivatives on mouse primordial follicle activation in vivo. 3 dpp female mice were intraperitoneally injected twice a day with physiological saline, PDEI (the combination of 4.18 mg/kg nimodipine, 0.79 mg/kg EHNA, and 1.36 mg/kg zaprinast), aminophylline (Ami, 21.02 mg/kg), dyphylline (Dyp, 40.68 mg/kg), or enprofylline (Enp, 1.94 mg/kg) for two consecutive days. The ovaries were collected after 12 h (C-E) or 2 days (A-B). A-B, The comparison of ovarian morphology (A) and primordial and growing follicle (GF) number (B) across various groups. The ovarian sections were hematoxylin-stained. C, The comparison of p-mTOR and p-Akt protein levels across various groups. D-E, FOXO3a localization in primordial follicle oocyte nuclear (arrowheads) or cytoplasm (arrows. D) and the comparison of FOXO3a nuclear export percentage in primordial follicle oocytes (E) across various groups. DDX4, red; FOXO3a, green. The representative images were displayed. PF, primordial follicle; GF, growing follicle. Red arrows, growing follicles. Scale bars, 50 µm. In each experiment, n ≥ 3 biological replicates. Bars indicate the mean ± SD. **P < 0.01 and ***P < 0.001. Two-tailed unpaired t-test was used to assess statistical significance.

Figure 6

Effects of oral administration of aminophylline on the fertility of naturally aged female mice. The naturally aged mice were administered water either with or without 4.0 mM aminophylline, and then the fertility was examined. A-B, The comparison of ovarian morphology (A) and follicle number at different stages (B) between control and aminophylline groups. The ovarian sections were hematoxylin-stained. Yellow arrows, early antral follicles; red arrows, late antral follicles. C-D, The comparison of ovulated oocytes between control and aminophylline groups. E, Normal and aberrant spindles morphology. F, The comparison of aberrant spindles proportion between control and aminophylline groups. G-H, JC-1 staining showed the ΔΨm of oocytes (G) and the comparison of relative red/green fluorescence intensity ratio (H) between control and aminophylline groups. Red, higher ΔΨm; green, lower ΔΨm. I-J, The ROS (green) levels of oocytes (I) and the comparison of ROS relative fluorescence intensity (J) between control and aminophylline groups. Green, ROS; BF, bright field. K-M, The comparison of different fertility status (K), pups per female (L), and pups per litter (M) number in naturally aged mice between control and aminophylline groups. N, The comparison of pups per litter mean body weight between control and aminophylline groups. The representative images were displayed. PF, primordial follicle; PrF, primary follicle; SF, secondary follicle; EA, early antral follicle; LA, late antral follicle; CL, corpus luteum. Scale bars, 50 µm. In each experiment, n ≥ 3 biological replicates. Bars indicate the mean ± SD. *P < 0.05. Two-tailed unpaired t-test was used to assess statistical significance.

Figure 7

Effects of aminophylline on the metabolome and transcriptome of naturally aged mice. The naturally aged mice were administered water either with or without 4.0 mM aminophylline, and then the metabolome of granulosa cells (A-E) and transcriptome of oocytes (F-I) were examined. A-B, Principal-component analysis (PCA, A) and Orthogonal partial least squares-discriminant analysis (OPLS-DA, B) of granulosa cells from control and aminophylline (Ami) groups. C, Statistical validation of OPLS-DA by 7-fold cross validation and 200 × response permutation testing of metabolites. D-E, Volcano plot and metabolome data displayed differential metabolites (D) and the levels of phospholipid metabolites (E) in granulosa cells from control and aminophylline groups. F-G, Heatmap (F) and volcano plot (G) showed the differentially expressed genes in oocytes from control and aminophylline groups. H, KEGG analysis of the differentially expressed genes in oocytes from control and aminophylline groups. I, Heatmap showed the differences in the expression of a set of transcripts involving different processes between control and aminophylline groups. Bars indicate the mean ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001. Two-tailed unpaired t-test was used to assess statistical significance.

Figure 8

Effects of theophylline derivatives on human primordial follicle activation in vitro. The fragments of human ovarian tissue were directly collected (uncultured) or cultured in medium alone (control) or with 50 μM aminophylline (Ami), 160 μM dyphylline (Dyp), or 10 μM enprofylline (Enp) for 4 days, followed by cultured in medium alone for additional 2 days. The fragments were collected after 4 (D) or 6 days (B-C). A, The PDE subtype expression levels in pre-granulosa cells (n = 8 follicles) and oocytes (n = 17 follicles) of human primordial follicles. pre-GCs, pre-granulosa cells; OOs, oocytes. B-C, The comparison of human ovarian tissue fragmented morphology (B) and primordial and growing follicle (C) proportion across various groups. The human ovarian sections were hematoxylin-stained. D, The comparison of p-mTOR and p-Akt protein levels across various groups. The representative images were displayed. Arrowheads, primordial follicles; arrows, growing follicle. Scale bars, 50 µm. In each experiment, n ≥ 3 biological replicates. Bars indicate the mean ± SD. *P < 0.05 and **P < 0.01. Two-tailed unpaired t-test was used to assess statistical significance.

Figure 9

A model depicting theophylline derivatives-mediated primordial follicle activation. Cyclic AMP produced by pre-granulosa cells enters oocytes through gap junctions, which is degraded by PDEs to maintain the dormant state of primordial follicles. Theophylline derivatives, as PDE inhibitors, activate the PI3K/Akt signaling pathway in oocytes by the accumulation of cAMP, leading to the activation of primordial follicles.