mTOR mediates Wnt-induced epidermal stem cell exhaustion and aging

Abstract

Epidermal integrity is a complex process established during embryogenesis and maintained throughout the organism lifespan by epithelial stem cells. Although Wnt regulates normal epithelial stem cell renewal, aberrant Wnt signaling can contribute to cancerous growth. Here, we explored the consequences of persistent expressing Wnt1 in an epidermal compartment that includes the epithelial stem cells. Surprisingly, Wnt caused the rapid growth of the hair follicles, but this was followed by epithelial cell senescence, disappearance of the epidermal stem cell compartment, and progressive hair loss. Although Wnt1 induced the activation of beta-catenin and the mTOR pathway, both hair follicle hyperproliferation and stem cell exhaustion were strictly dependent on mTOR function. These findings suggest that whereas activation of beta-catenin contributes to tumor growth, epithelial stem cells may be endowed with a protective mechanism that results in cell senescence upon the persistent stimulation of proliferative pathways that activate mTOR, ultimately suppressing tumor formation.

Figure 1

Expression of Wnt1 leads to a persistent proliferative state of the hair follicles (HF). A) The HFs grow in a cyclic fashion, involving a growth phase or anagen, an involution phase (catagen), and a resting phase (telogen). B) H&E sections of the first synchronous hair cycle from K5rtTA control animals showing the histological characteristic of each phase at the indicated days of age (upper panel). Typical examples of K5rtTA/tet-Wnt1 mice treated with doxycycline showing early entry into a growth phase followed by failure to undergo catagen and telogen (lower panel).

Figure 2

Alterations in HF structures in K5rtTA/tet-Wnt1 mice. A) Upper panel: Frozen sections from skin of 21 days old (P21) K5rtTA control mice show typical localization of the terminal differentiation marker fillagrin in the upper most layers of epidermis. The normal localization of cytokeratin 1 (K1) and 10 (K10) in the spinous layer of the interfollicular compartment and the normal expression of cytokeratin 6 (K6) in the HF are shown. Expression of α6 integrin (α6) delimitates the epidermal from the dermal compartment. DAPI was used for nuclear DNA staining (blue). Lower panel: K5rtTA/tet-Wnt1 mice fed with doxycycline show normal localization of fillagrin, cytokeratin 1, 10 and absence of cytokeratin 6 in the interfollicular epidermis (IF). HFs of mutant mice do not express fillagrin and cytokeratin 1, however an abnormal expression of cytokeratin 10 is present in the IRS compartment, co-localized with normal expression of cytokeratin 6. B) During late stage morphogenesis (P14), K5rtTA control mice express Type I hair keratins AE13 and the IRS specific marker (mIRSa). Expression of these markers cease during telogen (P21). K5rtTA/tet-Wnt1 mice fed with doxycycline instead maintain the expression of both markers through P21, consistent with the failure to entering in a resting state. Cytokeratin 5 (K5) staining reveals the basal epithelial layer. High magnification (63x) shows details of AE13 and mIRSa expressing structures in K5rtTA/tet-Wnt1 mice.

Figure 3

Activation of Wnt downstream targets in the epidermis and HF of doxycycline-treated K5rtTA/tet-Wnt1 mice. A) Presence of total β-catenin in control and mutant mice. Representative immunofluorescence analysis for total β-catenin (green), cytokeratin 14 (K14) (red), and DAPI (blue) in skins from control and K5rtTA/tet-Wnt1 fed with doxycycline are shown. High magnifications show membrane and cytoplasmatic localization of β-catenin. B) Hyperproliferation of epithelial HF cells in K5rtTA/tet-Wnt1 mice. Graphic bar represents number of BrDU-positive cells as the average ± S.E.M. of 10 fields at 20× of magnification in representative animals of the indicated genotype fed with doxycycline (***P<0.001). C) IHC for luciferase (Luc) shows negative staining in control littermates on P14 and P21, and positive staining in the ORS of K5rtTA/tet-Wnt1 mice aligned with the expression of the transgenic construct. Accumulation of the unphosphorylated, active form of β-catenin (Act β-Cat) in the HF IRS of mutant mice was noticeable. However, the active form of β-catenin was not detected by IHC in P14 and P21 control littermates. Using a reporter mouse line for β-catenin (BAT-Gal mouse) bred into our mouse lines, we were able to detect β-catenin activity at P14 (growth phase) but not at P21 (telogen) in control mice. K5rtTA/tet-Wnt1 mice crossed with BAT-Gal mice however, showed a strong activation of the β-catenin pathway visualized by the expression of β-Gal staining in the IRS. IHC reaction for WISP-1 (Wnt1 induced secreted protein 1) was observed in the IRS of the HF of K5rtTA/tet-Wnt1 mice but not in the K5rtTA control mice. Accumulation of high levels of the phosphorylated form of S6 (p-S6) was readily detectable in the IRS compartment of the HF upon Wnt-1 expression, and during late stage morphogenesis in control littermates (P14), but absent during telogen (P21).

Figure 4

Progressive premature hair loss in K5rtTA/tet-Wnt1 mice. A) K5rtTA/tet-Wnt1 transgenic mice fed with doxycycline show complete loss of their fur by 1 year of age when compared with K5rtTA control littermates. B) H&E histological sections of K5rtTA and K5rtTA/tet-Wnt1 by 21, 90, and 360 days of age showing progressive HF degeneration when compared to normal K5rtTA control mice. K5rtTA/tet-Wnt1 mice present a growth phase, anagen-like phenotype by 21 days of age, which evolves into altered HF structures and subsequent HF loss. C) Quantification of hair follicles was performed using histological sections cut in parallel to the surface of the epidermis. Graphic bar represent the number of hair follicles at the indicated age as the average ± S.E.M. of 10 fields at 10X magnification from representative animals of the indicated genotype (***P<0.001; **P<0.01; ns: P > 0.05).

Figure 5

Induction of senescence and loss of CD34⁺ epithelial stem cells in K5rtTA/tet-Wnt1 mice. A) Frozen skin sections of K5rtTA and K5rtTA/tet-Wnt1 mice fed with doxycycline (Doxy) showing SA β-Gal reactivity as a marker of senescence. Mutant HF are positive for this senescence marker while control sample present no staining for SAβ-Gal. Foci of IF staining for the double strand DNA-damage marker γ-H2AX was observed in the nuclei of HF from K5rtTA/tet-Wnt1 mice (γ-H2AX-red, and nuclear staining with DAPI). Higher magnifications are shown below. B) Quantification of γ-H2AX foci in HF from mutant and K5rtTA control mice. Bar graphs represent the quantitative analysis of positive HF, as the average ± S.E.M. of the number of HF presenting γ-H2AX positive cells in 10 fields at 20x magnification from representative animals of the indicated genotype and treatment (**P<0.01; *P <0.05). C) Frozen skin samples were sectioned and stained for keratin 5 (K5) (red) and CD34 (green) markers. K5rtTA mice display strong and well demarked staining for CD34 at the bulge compartment of the HF while no CD34⁺ cells were observed in the HF of K5rtTA/tet-Wnt1 mice fed with doxycyline. Higher magnifications (lower left) show the presence of CD34⁺ cells (Arrow) in K5rtTA control mice but absence of CD34⁺ cells in the K5rtTA/tet-Wnt1 mice (Arrow). D) A cohort of 6 mice per group was shaved simultaneously with the administration of doxycycline diet. By day 28, K5rtTA mice have completely recovered dorsal fur while K5rtTA/tet-Wnt1 doxycycline treated mice failed to recover the shaved area. Representative mice are shown.

Figure 6

Expression of Wnt1 in isolated primary epidermal keratinocytes causes cell senescence through mTOR. A) Primary cultures of keratinocytes derived from K5rtTA, and K5rtTA/tet-Wnt1 mice were treated with doxycycline for 5 days. Detection of senescence was observed in K5rtTA/tet-Wnt1 cells by SA-βgal staining (arrows). B) Monolayers of keratinocytes from K5rtTA, tet-Wnt1, and K5rtTA/tet-Wnt1 mice were starved overnight and treated with doxycycline (24 h). Cells were lysed and analyzed by Western Blotting for phosphorylated S6 (p-S6), total S6 (S6), and tubulin content as a loading control. Total cell lysates from each mouse genotype without doxycycline treatment were used as controls. Cells from K5rtTA mice starved overnight and pretreated for 30 min with EGF (30ng/ml) served as additional controls. C) Primary culture of keratinocytes from transgenic mice were plated onto cover slips and treated with doxycycline (1μg/ml) with or without rapamycin (50nM) for 24 h, as indicated. DAPI nuclear staining and IF for γ-H2AX, were performed as indicated, and images merged in the right panel. Quantification of γ-H2AX staining in K5rtTA, and K5rtTA/tet-Wnt1 derived cells treated with doxycycline plus vehicle or rapamycin. Data are represented as the percentage of γ-H2AX⁺ cells, expressed as the average ± S.E.M. of 10 fields at 20x magnification from representative cells of the indicated genotype and treatment (**P<0.001, ns: P >0.05). D) Keratinocytes from 7 weeks old mice were sorted into two distinct cellular populations (CD34⁺ and CD34⁻). Sorted cells were plated in collagen IV coated dishes and treated as indicated. Cells sorted for CD34 show membrane CD34 staining in vehicle treated cells, which was lost concomitant with the nuclear accumulation of γ-H2AX foci when treated with Wnt3 for 6 days. CD34⁻ cells also accumulated nuclear γ-H2AX foci in response to Wnt3. Rapamycin inhibited γ-H2AX accumulation in both CD34⁺ and CD34⁻ cells, and prevented the decreased expression of CD34 in CD34⁺ sorted keratinocytes caused by Wnt3. Quantitative analysis of CD34 and γ-H2AX staining of treated CD34⁺ sorted keratinocytes is shown in the right. Data are represented as the percentage of positive cells expressed as the average ± S.E.M. in triplicate cultures for each indicated treatment (***P<0.001; **P<0.01; ns: P >0.05).

Figure 7

Blockade of mTOR with rapamycin protects from hair follicle alterations, cell senescence, and the depletion of CD34⁺ epithelial stem cells caused by Wnt1. A) Twenty one days old mice fed with doxycycline pellets received daily doses of rapamycin from 3 days of age, while control littermates received only the vehicle. Rapamycin treatment prevented the development of changes in the fur, a phenotype that can be readily observed in of K5rtTA/tet-Wnt1 treated with doxycycline and vehicle control. Frozen sections were stained for p-S6 immunoreactivity. Rapamycin treatment prevented the activation of p-S6 in the IRS compartment and the histological changes caused by doxycycline in K5rtTA/tet-Wnt1 mice, as observed in H&E histological sections. Note the failure of the lower half of the HF to regress in K5rtTA/tet-Wnt1 mice fed with doxycycline and the re-establishment of a normal hair cycle (telogen) in these mice upon rapamycin treatment. B) Rapamycin did not prevent the expression from the β-catenin reporter system, as judged by the detection of β-Gal in 14 days old K5rtTA/tet-Wnt1/BAT-Gal mice treated with doxycycline and rapamycin. C) K5rtTA/tet-Wnt1 mice were fed with doxycycline by 3 days of age and received daily IP injections of rapamycin or vehicle and sacrificed at 21 days of age. Frozen sections of K5rtTA control mice show the presence of CD34⁺ epithelial stem cells in the bulge region of the HF, and absence of nuclear γ-H2AX foci. Instead, K5rtTA/tet-Wnt1 mice fed with doxycycline present high number of cells displaying γ-H2AX staining, and lack of CD34⁺, but both were prevented by the rapamycin treatment. D) Quantification of CD34⁺ and γ-H2AX staining in K5rtTA/tet-Wnt1 mice fed with doxycycline treated with vehicle control or rapamycin. Data are represented as the percentage of HF displaying CD34⁺ cells, expressed as the average ± S.E.M. of 10 fields at 20× magnification from representative animals of the indicated genotype and treatment (***P<0.001, *P<0.05), and as the percentage of nuclei exhibiting γ-H2AX positive foci, expressed as the average ± S.E.M. after counting 300 HF cells at 20× magnification from representative animals of the indicated genotype and treatment (**P<0.01; ns: P >0.05).