Effects of Telomerase and Telomere Length on Epidermal Stem Cell Behavior

See allHide authors and affiliations

Science  19 Aug 2005:
Vol. 309, Issue 5738, pp. 1253-1256
DOI: 10.1126/science.1115025


A key process in organ homeostasis is the mobilization of stem cells out of their niches. We show through analysis of mouse models that telomere length, as well as the catalytic component of telomerase, Tert, are critical determinants in the mobilization of epidermal stem cells. Telomere shortening inhibited mobilization of stem cells out of their niche, impaired hair growth, and resulted in suppression of stem cell proliferative capacity in vitro. In contrast, Tert overexpression in the absence of changes in telomere length promoted stem cell mobilization, hair growth, and stem cell proliferation in vitro. The effects of telomeres and telomerase on stem cell biology anticipate their role in cancer and aging.

Tumor formation and aging are associated with alterations in the number or functional competence of tissue stem cells (13). Both processes have also been linked to alterations at the telomere (47), the nucleoprotein structure that caps chromosome ends (8, 9), and to changes in the activity of telomerase, the reverse transcriptase that elongates telomeres (10, 11). The catalytic subunit of telomerase (Tert) is expressed in the stem cell compartment of several adult tissues (12), although telomerase levels in these tissues are not sufficient to prevent progressive telomere shortening with age (10). Reduced telomerase activity due to mutations in telomerase components in the human diseases dyskeratosis congenita and aplastic anemia (10) leads to accelerated telomere shortening and premature loss of tissue regeneration, which suggests that telomerase levels in the adult organism are rate limiting and influence organ homeostasis. Further evidence for a role of telomerase and telomere length in organ homeostasis comes from the study of telomerase-deficient mice (Terc–/– mice), which show premature aging and a decreased proliferative potential of adult stem cell populations (1315).

To investigate the role of telomerase and telomere length on stem cell biology, we used mouse models with altered telomerase activity (16). We compared epidermal stem cell number in different generations of telomerase-deficient mice, which have telomeres ranging from slightly reduced in length (first generation, G1 Terc–/– mice) to critically short (third generation, G3 Terc–/– mice) (17, 18). Because telomerase activity per se is not required for cell proliferation when telomeres are long, the study of G1 and G3 Terc–/– mice allowed us to assess independently the impact of telomerase deficiency and telomere length on epidermal stem cells. To visualize epidermal stem cells, we used a labeling technique previously shown to mark self-renewing and multipotent epidermal cells, the so-called “label-retaining cells” (LRCs) (16, 19) (Fig. 1A). Confocal microscopy revealed that LRCs are enriched at the bulge area of the hair follicle in Terc+/+ (wild-type) mice, which corresponds to the niche of epithelial stem cells (Fig. 1B, fig. S1A, and SOM Text) (19). Interestingly, G1 Terc–/– hair follicles contained significantly more LRCs than did Terc+/+ follicles (Fig. 1, B and C). Even greater numbers were present in the G3 Terc–/– follicles (Fig. 1, B and C). As with Terc+/+ mice, LRCs in Terc–/– mice accumulated in the niche/stem cell compartment (Fig. 1B and fig. S1A). The increased number of LRCs in mice with short telomeres was an unexpected finding, because Terc–/– mice are resistant to tumor-inducing protocols (20, 21) and enlarged numbers of cells with stem characteristics have been associated with increased tumor formation (2).

Fig. 1.

Telomere attrition and Tert overexpression are independent determinants of LRC mobilization upon TPA treatment. (A) Strategy to assay mobilization of LRCs (16). Mice were injected with bromodeoxyuridine (BrdU) and, 69 days later, whole mounts of tail epidermis were collected from untreated and TPA-treated mice and stained with an antibody to BrdU. (B) Representative confocal micrographs of tail follicles from wild-type (WT), G1 Terc–/–, and G3 Terc–/– mice in a C57BL6 background (left) and from WT and K5-mTert mice in a C57BL6 x DBA/2 background (right), stained for BrdU (green) after LRC labeling and treatment (16). The tail-hair follicles are grouped in sets of three (triplets), two outer follicles (OF) and one inner follicle (IF), which generally contains fewer LRCs (23). LRCs of all genotypes accumulate in the bulge (Bu) region of the hair follicle. Scale bars, 80 μm. (C) Quantification of LRCs in the conditions and genotypes shown in (B). Comparison P values are P < 0.0001 (significant) in all cases between genotypes, as well as between control treatment and TPA treatment, except P < 0.05 (significant) for WT (control) versus G1 Terc–/– (control) and for G3 Terc–/– (control) versus G3 Terc–/– (TPA). SD is represented on each bar for each genotype and condition.

To investigate whether LRCs in Terc–/– mice mobilize (exit their quiescent state and migrate) out of the niche, we studied the response of wild-type, G1 Terc–/–, and G3 Terc–/– LRCs to treatment with 12-O-tetradecanoylphorbol 13-acetate (TPA), a potent tumor promoter (22) that activates LRCs to give numerous progeny (16). TPA treatment results in rapid disappearance of LRCs (23), skin hyperplasia (24), and entry of hair follicles (HF) in their anagen (growing) phase (25). After TPA treatment, wild-type epidermis showed ∼70% reduction in the number of LRCs localized within the stem cell niche (Fig. 1C). In contrast, TPA-treated G1 Terc–/– mice showed ∼43% reduction, which suggests a defect in their mobilization (Fig. 1C). This defect was even more pronounced in G3 Terc–/– mice, which showed only ∼14% reduction after TPA treatment (Fig. 1C). Coincidentally, the proliferation index in different compartments of the G3 Terc–/– follicle was lower than that of wild-type follicles (fig. S2, A and B, and SOM Text). Furthermore, in contrast to wild-type mice, interfollicular skin thickness (IFE) and HF length were not significantly increased in G3 Terc–/– mice (Fig. 2, A to C), reflecting defective hyperplasia and anagen responses in these mice after TPA treatment. An alternative explanation for the differences in the number of LRCs between genotypes, such as different apoptotic rates (16) (fig. S3 and SOM Text) or migration of stem cells out of the niche without division, were ruled out (fig. S1A and SOM Text). All Terc–/– mice used had a histopathologically normal skin at the time of analysis (16) (fig. S4), which suggests that stem cell mobilization defects anticipate the aging and cancer-resistant phenotypes of these mice (18, 20).

Fig. 2.

Telomere attrition and Tert overexpression influence both TPA-induced hyperplasia and anagen. (A) Representative tail-skin sections from mice of the indicated genotypes before and after TPA treatment. Continuous black double-pointed arrows mark interfollicular (IFE) thickness. Black dashed double-pointed arrows mark hair-follicle (HF) length from sebaceous glands (SG) to dermal papilla (DP). Quantification of IFE thickness (B) and of HF length from SG to DP (C) in tail skin from mice of the indicated genotype. Histomorphometry was performed in three mice of each genotype before and after TPA treatment, quantifying a total of 30 follicles in the control group and 90 follicles in the TPA-treated group. (D) Representative back skin sections from 8-week-old male mice before and 10 days after plucking. Double-pointed arrows mark dermis thickness. Quantification of HF length from IFE to DP (E) and dermis thickness (F) in 30 different sections per genotype before and after plucking. NS, not significant, P > 0.05; *, significant, P < 0.05; ***, significant, P < 0.001. SD is represented on each bar for each genotype and condition. Scale bars, μm.

Next, we performed clonogenic assays to compare the proliferation potential of Terc+/+ and Terc–/– epidermal stem cells (16). Individual colonies in this assay have been proposed to derive from single stem cells (26). In agreement with the in vivo results, keratinocytes from G1 and G3 Terc–/– mice formed fewer and smaller colonies than those from Terc+/+ controls (Fig. 3, A and B), reflecting the defective capacity of G3 Terc–/– cells to proliferate. Interestingly, colony formation was particularly impaired in those G3 Terc–/– mice showing a small-size phenotype (Fig. 3, A and C), which in turn is associated with shorter telomeres (18). Finally, G1 and G3 Terc–/– colonies were able to fully differentiate upon high calcium treatment (16) (fig. S5 and SOM Text), which suggests that decreased colony number in these mice is not due to defects in stem cell multipotency.

Fig. 3.

Telomere attrition and Tert overexpression influence the proliferative potential of stem cells ex vivo. (A) Quantification of size and number of macroscopic colonies obtained from isolated keratinocytes of the indicated genotype purified from 2-day-old mice and cultured for 1 week on J2-3T3 mitomycin-C-treated feeder fibroblasts. Comparison P values are P < 0.0001 (significant) in all cases except P = 0.041 (significant) for WT versus G1 Terc–/– and P = 0.087 (not significant) for WT versus K5-mTert/Terc–/–. (B) Note that a high proportion of G3 Terc–/– colonies comprise less than 50 cells. (C) G3 Terc–/– newborn mice with a smaller size than littermates present an exacerbated deficiency in colony formation. SD is represented on each bar for each genotype and condition.

To study the impact of telomerase up-regulation on the stem cell compartment and cancer, we used K5-mTert transgenic mice, which have increased telomerase activity and increased Tert expression in skin, including the stem cell niche (24, 27). K5-mTert mice have an increased susceptibility to tumorigenesis in the absence of changes in telomere length (16, 24, 28, 29) (fig. S6 and SOM Text). To determine whether increased mTert expression affected the number and mobilization of epidermal stem cells, we compared LRC number and functionality in wild-type mice versus K5-mTert mice (16). In K5-mTert mice, the basal number of LRCs in the absence of TPA treatment was ∼65% lower than in wild-type mice (Fig. 1, B and C). After TPA treatment, ∼94% of the LRCs mobilized in K5-mTert mice, compared with ∼67% in wild-type mice (Fig. 1, B and C). In fact, at the end of the experiment, only 1 to 2 cells were labeled in K5-mTert outer follicles, compared with an average of 12.7 cells in wild-type mice (Fig. 1, B and C). Thus, mTert overexpression in epidermal stem cells appears to promote LRC mobilization. Again, this occurred in the absence of changes in apoptotic rates (fig. S3) or LCR accumulation out of the stem cell niche (fig. S1B). This Tert-dependent effect is consistent with an increased number of proliferative keratinocytes (fig. S2, A and B), an increased skin hyperplasia (IFE thickness), and increased anagen response (increased HF length) after TPA stimulation (Fig. 2, A to C) or after plucking (16) (Fig. 2, D to F, and SOM Text), and it is further supported by a high colony-forming efficiency of isolated K5-mTert keratinocytes compared with the wild-type controls (Fig. 3A) in the absence of differences in multipotency (fig. S5).

To address whether Terc is required for the high clonogenicity effect of Tert overexpression, we generated K5-mTert mice in a Terc–/– genetic background, K5-mTert/Terc–/– mice. The absence of Terc abolished the effect of Tert overexpression on colony formation (Fig. 3A), which suggests that the formation of Tert/Terc complexes is required for the enhanced clonogenicity of K5-mTert cells. Again, no differences in the ability of K5-mTert or K5-mTert/Terc–/– colonies to differentiate upon high calcium conditions were observed (fig. S5).

Telomere length and telomerase activity are independent determinants of the mobilization efficiency and proliferative capacity of epidermal stem cells (fig. S7). On the one hand, critically short telomeres inhibit the mobilization of epidermal stem cells, resulting in the persistence of LRCs in the hair follicle niche. This in turn results in a reduced proliferative potential and reduced anagen response after TPA treatment, as well as in a low in vitro clonogenicity. On the other hand, Tert overexpression promotes mobilization of epidermal stem cells and decreases the number of LRCs in the stem cell niche after TPA treatment, coincidental with increased numbers of proliferating cells, increased skin thickness and anagen response, and a high efficiency for colony formation. These effects of Tert anticipate the increased susceptibility of K5-mTert mice to develop skin tumors (24).

Supporting Online Material

SOM Text

Materials and Methods

Figs. S1 to S7


References and Notes

View Abstract

Stay Connected to Science

Navigate This Article