Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Antagonistic effects of telomerase on cancer and aging in K5-mTert transgenic mice

Abstract

Many degenerative diseases that occur with aging, as well as premature aging syndromes, are characterized by presenting cells with critically short telomeres. Telomerase reintroduction is envisioned as a putative therapy for diseases characterized by telomere exhaustion. K5-mTert transgenic mice overexpress telomerase in a wide spectrum of tissues. These mice have a higher incidence of both induced and spontaneous tumors, resulting in increased mortality during the first year of life. Here, we show that in spite of this elevated tumor incidence and the initial lower survival, K5-mTert mice show an extension of the maximum lifespan from 1.5 to 3 months, depending on the transgenic line, which represents up to a 10% increase in the mean lifespan compared to wild-type littermates. This longer lifespan is coincidental with a lower incidence of certain age-related degenerative diseases, mainly those related to kidney function and germline integrity. Importantly, these effects of telomerase overexpression cannot be attributed to dramatic differences in telomere length in aged K5-Tert mice compared to wild-type mice, as shown by quantitative telomeric FISH. These findings indicate that telomerase overexpression extends the maximum lifespan of mice.

Introduction

The ends of vertebrate chromosomes are capped by telomeres, protective structures formed by tandem DNA TTAGGG repeats and associated proteins (Blackburn, 2001; Chan and Blackburn, 2002). Telomerase is a ribonucleoprotein that synthesizes telomere repeats in vivo (Greider and Blackburn, 1985). Adult somatic tissues generally lack or have insufficient telomerase activity to compensate telomere shortening coupled to cell duplication (Harley et al., 1990). Indeed, multiple evidences show that telomere shortening occurs associated to age-related diseases such as hypertension, atherosclerosis, diabetes mellitus insulin-independent, Alzheimer, and cancer (Aviv and Aviv, 1998; Cawthon et al., 2003), as well as associated to many chronic degenerative diseases such as ulcerative colitis (O'Sullivan et al., 2002; Cawthon et al., 2003). Telomerase re-introduction into cells derived from patients with premature aging syndromes rescues their critically short telomeres, as well as their premature senescence phenotype in vitro (Wyllie et al., 2000). Telomerase reintroduction into telomerase-deficient mice with short telomeres, Terc−/− mice (Blasco et al., 1997), is also sufficient to prevent critical telomere loss and premature aging phenotypes in these mice (Samper et al., 2000). These findings lead to the proposition that telomerase reintroduction in cells with limited telomere reserve could prevent some degenerative diseases associated to normal aging, as well as, may palliate the pathologies produced by critical telomere loss and decreased proliferative potential in patients with premature aging syndromes. However, considering that high telomerase activity levels are a common feature of tumors, it is important to study the relative impact of telomerase reactivation in both cancer and aging, as well as in the overall organismal survival.

In order to study the effects of telomerase overexpression in the context of the organism, we have generated a telomerase transgenic mouse, K5-mTert (Gonzalez-Suarez et al., 2001). These mice are viable and show a similar telomere length to that of wild-type littermate mice (Gonzalez-Suarez et al., 2001). However, K5-mTert mice show a higher rate of spontaneous and chemically induced tumors than wild-type mice as they age, as well as when in a p53 mutant background (Gonzalez-Suarez et al., 2001, 2002). These results suggested that Tert overexpression facilitates cell proliferation and cooperates with other mutations in tumor development in the absence of significant changes in telomere length.

Here, we show that in spite of the increased susceptibility of K5-mTert mice to develop tumors, these mice show a significant increase in their maximum lifespan. In addition, we show that the longer survival of K5-mTert mice is coincidental with a decreased incidence of degenerative diseases (in particular, kidney degenerative diseases) in these mice compared to wild-type littermates. Finally, using quantitative telomere FISH (Q-FISH), we show that both wild-type and K5-mTert mice have similarly long telomeres at 2 years of age, and that they do not show the presence of dramatically short telomeres. These results suggest that the longer lifespan of K5-mTert mice compared to wild-type mice is unlikely to be mediated by a role of telomerase in rescuing short telomeres, again suggesting a direct role of mTert in organismal survival, which is independent of telomere length.

Results

Increased maximum lifespan of K5-mTert telomerase transgenic mice

To study the impact of Tert overexpression on organismal lifespan, we generated in the past large colonies of wild-type and K5-mTert littermate mice from two independent transgenic lines (lines T1 and T8) with high levels of telomerase expression in most tissues, including stratified epithelia (skin, esophagus, forestomach, vagina, etc.), other epithelial tissues (stomach, colon, lung, etc.), lymphoid tissues (thymus, spleen), as well as male germline tissues (testis, seminal glands, etc.) (Gonzalez-Suarez et al., 2001, 2002). We had shown before that Tert overexpression in both transgenic lines favors tumor development in the absence of significant changes in telomere length (Gonzalez-Suarez et al., 2001, 2002). Here, we show that this increased tumorigenesis results in a higher mortality rate of T1 and T8 K5-mTert mice compared to wild-type littermates during the first months of life (Figure 1a). In particular, the number of mice included in the study was 65 wild-type (wt), 88 T1 K5-mTert mice (T1), and 52 T8 K5-mTert (T8) mice. Figure 1b shows a Kaplan–Meier representation of the survival curves including the error bars for each data point (s.e.m.), again illustrating the higher mortality of T1 and T8 K5-mTert mice during the first months of life compared to the wild-type controls.

Figure 1
figure1

(a) Survival curves for wild-type (65 mice), T1 K5-mTert (88 mice), T8 K5-mTert (52 mice). (b) Kaplan–Meier survival curves (the error bars indicate the s.e.m. values). (c) Kaplan–Meier survival curves excluding data prior to week 100. (d) Kaplan–Meier survival curves excluding data prior to week 100 and with error bars. (e) Representation of the 10% longest-lived mice from the indicated genotypes. NOTE: mice belonging to the ‘sacrificed’ group were never included in the survival curves

The survival curve slope during the first year of life, which reflects on the survival trend, was m=−0.14 for wild-type, −0.24 for T1, and −0.42 for T8 mice, illustrating the lower survival of T1 and T8 K5-mTert mice compared to wild-type controls. Interestingly, this trend changed after week 100 and both T1 and T8 transgenic lines showed increased survival compared to wild-type mice, as indicated by curve slopes of −1.24 and −0.84 for T1 and T8 K5-Tert, respectively, compared to −1.56 for wild-type mice. In fact, at week 125 after birth, 13.6% of T1 mice and 7.7% of T8 mice were still alive compared to only 3.1% of wild-type mice (Figure 1a). These findings were also illustrated using a ‘Cox proportional hazards model’, which allows to test if death rates vary over time for different covariants (Materials and methods). This model indicated that the survival for both transgenic mouse groups (T1 and T8) is significantly different before and after week 75. In particular, T1 and T8 K5-mTert mouse colonies tend to show a higher risk of death in the first 75 weeks of life compared to the wild-type colony; however, after week 75, this trend is inverted and T1 and T8 transgenics have a lower or similar risk of death than wild-type mice (Supplementary Tables 1 and 2). Therefore, two different lines of K5-mTert mice showed a higher survival than wild-type mice at old ages, in spite of the fact that they showed a higher mortality during the first months of age.

Since there was a change in the survival trend of T1 and T8 K5-mTert mice at older ages, we represented Kaplan–Meier survival curves excluding the survival data before week 100 (Figure 1c and d). In particular, the number of mice that survived after 100 weeks was 34 wild-type (52.3%), 50 T1 (56.8%), and 17 T8 K5-mTert (32.7%) mice. Long-rank test analysis of this set of data indicated highly significant differences in survival between wild-type and T1 K5-mTert mice (χ2=7.486 P=0.0062). In the case of T8/wild-type comparison, the observed differences in survival did not reach statistical significance (χ2=2.468; P=0.1162); however, this fact could be due to the small sample size for the T8 line (only 17 T8 K5-mTert mice survived after 100 weeks). In support of this, a direct comparison between T1 and T8 K5-mTert transgenic lines indicated that they have a similar behavior and that they do not differ in survival (χ2=0.4665; P=0.4946).

The increased survival of both K5-mTert transgenic lines at old ages is also reflected by the fact that the survival curves of both T1 and T8 K5-mTert lines eventually crossed with that of wild-type mice (Figure 1a). Moreover, it is also apparent from Figure 1a–d that the ‘maximum lifespan’ is increased in both T1 and T8 K5-mTert lines compared to wild-type controls. This extension of the maximum lifespan was greater for T1 than for T8 K5-mTert mice. In particular, the extension of the maximum lifespan was of 14 weeks (3 months) or 7 weeks (1.5 months) for the longest-lived T1 and T8 K5-mTert mice, respectively, compared to the longest-lived wild-type mouse. However, direct ‘maximum lifespan’ comparisons are not reliable when the sample size is different for each genotype (in our case, a total of 65 wild-type, 88 T1, and 52 T8 mice). We circumvented this problem by calculating the lifespan of the 10% of each group that is longest-lived. This represented a total of seven wild-type (136, 128, 125, 124, 123, 123, 122 weeks), nine T1 K5-mTert (150, 150, 147, 140, 138, 136, 136, 133, and 130 weeks), and five T8 K5-mTert (143, 132, 129, 126, and 122 weeks) mice (Figure 1e). For longevity comparisons, we used the Tukey's multiple comparison test (Materials and methods). The results obtained indicate significant differences in the maximum lifespan between wild-type and T1 mice (P<0.01; medium difference −14.14, q=5.84). In the case of wild-type and T8 mice comparison, no significant differences were reached (P≥0.05; medium difference=−4.54 q=1.616). Again, most likely this is due to the fact that only 17 T8-mTert mice survived longer than 100 weeks, thus reducing the probability of reaching a significant lifespan extension in this group. In support of a similar behavior of T1 and T8 K5-mTert transgenic lines, no significant differences in the maximum lifespan were observed between these lines (P≥0.05; medium difference=+9.6, q=3.58). Indeed, when we performed the Tukey's test considering the 20% longest-lived mice out of the mice that reach 100 weeks of age (a total of seven wild-type, 10 T1, and three T8 mice), the differences between T1 and T8 mouse cohorts with respect to the wild-type group were further increased (wt/T1: P<0.01, medium difference=−13.19, q=5.506; wt/T8: P≥0.05, medium difference=−8.952, q=2.67). All together, these results suggest that the constitutive expression of high levels of Tert has an impact in extending the lifespan of mice, and that this lifespan extension was highly significant in the case of the T1 K5-Tert mouse line, in spite of the fact that these mice showed a lower survival than wild-type mice in the first year of life.

It is important to note that in spite of the fact that the T8 mTert transgenic line did not reach statistical significance in maximal lifespan extension when comparing with wild-type mice, both T1 and T8 transgenic lines showed a similar behavior. In particular, the Cox analysis data reveals that both transgenic lines have a similar behavior in survival (which is different from that of wild-type mice) after week 75 (of age). Furthermore, the survival curves of both transgenic lines crossed that of the wild-type mice. This similar trend in survival for T1 and T8 transgenic lines is in agreement with previous data showing that both transgenic lines have a similar behavior in proliferation and tumorigenesis assays (Gonzalez-Suarez et al., 2001, 2002).

Increased hyperplasic lesions coincidental with decreased degenerative lesions in aged K5-mTert mice compared to wild-type mice

To understand the cause of the increased survival of K5-mTert mice at old ages, wild-type and K5-mTert mice from T1 and T8 lines were killed when they showed signs of suffering or disease (group designated as ‘moribund’). In addition, groups of 13 wild-type, 15 T1, and nine T8 mice (91–132-week old) were killed in parallel (group designated as ‘sacrificed’). An exhaustive histopathological pathology analysis was performed in all the killed mice and in each of the 42 wild-type, 64 T1, and 33 T8 moribund mice included in Tables 1 and 2. Neoplasias present in the ‘moribund’ and ‘sacrificed’ groups have been previously described (Gonzalez-Suarez et al., 2002). In this study, we focused on senile and degenerative lesions, which normally appear during the aging process (Mohr et al., 1996). The senile lesions found in both moribund and sacrificed mice were classified according to their cell-type origin and are shown in Table 1 (age-related proliferative lesions: hyperplasias and hypertrophies, as well as cysts – we excluded preneoplastic and neoplastic lesions) and Table 2 (age-related nonproliferative lesions: atrophies and degenerations) (Materials and methods).

Table 1 Proliferative senile lesions (excluding preneoplastic and neoplastic lesions) observed in moribund and sacrificed wild-type, T1 K5-mTert, and T8 K5-mTert mice
Table 2 Senile nonproliferative lesions (excluding preneoplastic and neoplastic lesions) observed in moribund and sacrificed wild-type, T1 K5-mTert, and T8 K5-mTert mice

As reported in Tables 1 and 2, some of these senile lesions affected a similar proportion of aged wild-type and K5-mTert mice. In particular, we found no differences between genotypes in the incidence of liver microgranulomas, ovary cysts, cystic endometrial hyperplasia, inflammation of the seminal vesicles, or atrophy of the spleen. However, some hyperproliferative lesions, such as mucosal hyperplasia of the stomach or the intestine, as well as acinar hyperplasia in the mammary glands, appeared to be more frequent in T1 and T8 K5-mTert mice than in wild-type controls, in agreement with the pattern of expression of the transgene (Table 1) (Gonzalez-Suarez et al., 2002). In particular, acinar hyperplasia of the mammary glands was detected in 5.5% of wild-type mice compared to 18.4 and 25% of the moribund T1 and T8 K5-mTert mice, respectively (Table 1). We also observed significant differences in the incidence of stomach mucosal hyperplasia between moribund wild-type and T8 K5-mTert mice (χ2=8.364; P=0.0038). Similarly, thyroid follicular cell hyperplasia and hyperplastic islet cells in the pancreas were also more frequent in moribund K5-mTert mice than in wild-type controls, and these differences reached statistical significance in the case of the T1 K5-mTert line (χ2=3.496; P=0.0615 and χ2=4.174; P=0.0411, respectively) (Table 1; Figure 2a, b). The incidence of thyroid follicular hyperplasia was also significantly increased in the ‘sacrificed’ T1 K5-mTert mice compared to the age-matched wild-type controls (χ2=3.125; P=0.0771) (Figure 2a, b).

Figure 2
figure2

(a) Thyroid gland hyperplasia in moribund and sacrificed wild-type, T1 K5-mTert, and T8 K5-mTert females. Each symbol represents a single mouse. Statistical significance is also indicated. (b) Percentage of mice that showed thyroid gland hyperplasia at the time of death. Asterisk indicates statistically significant differences

In addition to the higher incidence of hyperproliferative lesions, transgenic mice also showed a significant reduction in some degenerative lesions compared to wild-type controls (Table 2). In particular, K5-mTert mice showed a lower frequency of uterus atrophy (detected in 5.5% of wild-type females but never in the K5-mTert females), in agreement with the expression pattern of the transgene. K5-mTert mice also showed a lower frequency of ovary atrophy (present in 16.7% of wild-type females compared to 10.5% of T1 K5-mTert females and 8.3% of T8 K5-mTert females) (Table 2). The biggest differences between genotypes, however, were found in the lesions that affected kidney function, such as glomerulonephritis, as well as in testicular atrophies, which were more frequent in aged wild-type mice than in T1 and T8 K5-mTert mice (Table 2; Figures 3, 4) (see below for detailed analysis).

Figure 3
figure3

(a) Kidney membranoproliferative glomerulonephritis at the time of death in either moribund or sacrificed wild-type, T1 K5-mTert, and T8 K5-mTert mice. Each symbol represents a single mouse. Statistical significance is also indicated. (b) Percentage of mice that showed membranoproliferative glomerulonephritis at death. Asterisk indicates statistically significant differences

Figure 4
figure4

(a) Testicular atrophy at death in moribund and sacrificed wild-type, T1 K5-mTert, and T8 K5-mTert males. Each symbol represents a single mouse. Statistical significance is also indicated. (b) Percentage of males that showed testicular atrophy at death. Testicular atrophy has been classified as mild or severe as described in the text. Asterisk indicates statistically significant differences

Tert overexpression preserves renal function in aged K5-mTert mice

Kidney is one of the tissues most frequently damaged in old mice. Kidney dysfunction can be the result of a general organismal failure including defects in the digestive tract, the immune system, as well as the cardiovascular system. We saw a significant protection in both transgenic lines from kidney disease. In particular, we saw a clear protection from membranoproliferative glomerulonephritis, a renal inflammatory process associated to senility in mice, which when chronic, could give rise to interstitial nephritis, tubular dilatations, cysts, and tubular calcifications in the kidney. Membranoproliferative glomerulonephritis is characterized by a thickening of the glomerular capillary basement membrane and by an increase of mesangial cells and mesangial matrix. In severe cases, it may be accompanied by the formation of semicircular hypercellular lesions known as glomerular crescent that can lead to glomerular fibrosis and chronic renal failure (see Figure 5e for an illustrative example) (Table 2). Wild-type mice were more frequently affected with this pathology than K5-mTert mice (Table 2). In particular, only 20.3% of T1 K5-mTert and 18.2% of T8 K5-mTert mice presented membranoproliferative glomerulonephritis at time of death, compared to 35.7% of wild-type mice (Figure 3a, b; Table 2) (‘moribund’ mice), in spite of the fact that the K5-mTert mice generally reached older ages (Figure 1). The differences between genotypes were statistically significant as indicated by χ2 test values (wt/T1: χ2=3.09, P=0.0785; wt/T8: χ2=2.82, P=0.0932). Other pathologies affecting the kidney, such as chronic interstitial nephritis and amyloidosis, were only detected in wild-type mice but never in K5-mTert mice (Table 2). Tubular calcifications, degenerations, and cysts, however, were present in both genotypes at a similar frequency (Table 2). These results were confirmed with the age-matched ‘sacrificed mice’ group, where membranoproliferative glomerulonephritis affected 23.1% of wild-type mice compared to no T1-K5-mTert mice affected and 11.1% of the T8 K5-mTert mice affected (wt/T1: χ2=3.877, P=0.049; wt/T8: χ2=0.512, P=0.4763) (Figure 3a, b; Table 2).

Figure 5
figure5

Hematoxylin–eosin staining of representative degenerative lesions. (a, b) Severe testicular degeneration showing a reduced number of germinal cells and vacuolated Sertoli cells from a 2-year-old wild-type mouse. (c, d) Normal testis in a 2-year-old T1 K5-mTert mouse. (e) Section of a kidney showing chronic membranoproliferative glomerulonephritis with diffuse hyalin thickening of glomerular capillary basement membranes and hypercelullarity from a 2-year-old wild-type mouse. (f) Section of a normal kidney from a 2-year-old K5-mTert mouse. a, c: × 40 magnification. b, d: × 200 magnification. e, f: × 200 magnification

To address whether the lower number of degenerative kidney lesions in aged K5-mTert mice compared to similarly aged wild-type controls was correlated to mTert mRNA expression levels, we used real-time PCR to determine mTert mRNA abundance in the kidneys from 1-year-old T1 and T8 K5-mTert mice compared to those of wild-type mice. As shown in Table 3, there was an approximately fourfold increase in mTert levels in T1 and T8 kidneys compared to wild-type controls.

Table 3 Tert mRNA overexpression in 1-year-old K5-mTert mice compared to wild-type mice

All together, these results indicate that Tert overexpression decreases degenerative pathologies of the kidney associated with aging in two independent lines of transgenic K5-mTert mice, possibly as a combined effect of the modest increase of mTert expression in the kidney in old mice, as well as of the wide K5-Tert transgene expression pattern in these mice (Gonzalez-Suarez et al., 2002), which could impact, in general, organismal fitness. The significant protection to kidney disease shown by K5-mTert transgenic mice could be in part responsible for their higher survival at older ages compared with wild-type mice.

Preservation of the male germ line in aged K5-mTert mice

The fact that male germ tissues (testis, seminal glands, prepucial glands, and penis) showed high levels of Tert overexpression in K5-mTert transgenic lines (Gonzalez-Suarez et al., 2002; see also Table 3) prompted us to study testicular function with age in these mice. Testicular atrophy occurs spontaneously in older mice and is characterized by a decrease in the size of the testes accompanied by a decrease in the diameter of the seminiferous tubules. At the cellular level, there is atrophy of the germinal epithelium and massive loss of spermatogenic cells, so that only a few spermatogonias remain (see Figure 5a, b for examples). To asses the impact of telomerase overexpression in the male germ line of K5-mTert mice, we classified the testicular atrophies found in these mice as mild or severe depending on the extension of the depletion of germinal cells and the number of seminiferous tubules affected. Mild testicular atrophy is a focal process where the seminiferous tubules affected show a decreased germinal epithelium (spermatids and spermatocytes). In severe atrophies, however, most tubules lose the germinal epithelium and are only lined by vacuolated Sertoli cells showing thickening of the basement membrane, and giving rise in some cases to calcifications.

Interestingly, K5-mTert mice preserved the male germ line to a higher extent than the wild-type controls. In particular, 20.8% of moribund wild-type mice show severe testis atrophy at the time of death compared to only 11.5% of T1 K5-mTert mice and none of the T8 K5-mTert mice. These differences reached statistical significance in the case of the T8 transgenic line (wt/T1: χ2=0.802, P=0.3704; wt/T8: χ2=4.922, P=0.0267) (Figure 4a, b; Table 2). We also detected calcifications in the seminiferus tubules in 4.7% of wild-type males, but never in the K5-mTert mice (Table 2). Some of transgenic mice, 3.8% of T1 and 14.3% of T8, only showed mild testicular atrophies at the time of death. In addition, severe testis atrophy appeared earlier in wild-type mice than in the K5-mTert transgenics. In particular, we detected severe testis atrophy in an 86-week-old wild-type mouse, whereas T1 males did not show severe atrophy until week 110, and some T1 mice reached 136 weeks of age with a normal testis structure (Figure 4a). Again, these results were confirmed with the ‘sacrificed’ group of mice (Figure 4a). In this case, we analysed the testis of mice killed in parallel and found that 22.2% of wild-type males (two out of nine) showed testis atrophy (mild or severe), while none of the T1 and T8 K5-mTert mice showed any signs of testicular atrophy (Figure 4a, b, Table 2).

This increased preservation of the male germ line in aged K5-mTert mice compared to similarly aged wild-type controls was correlated with increased mTert mRNA levels in the testis from 1-year-old T1 and T8 K5-mTert mice compared to those of wild-type mice (Table 3). As shown in Table 3, there was an approximately 5–7-fold increase in mTert mRNA levels in T1and T8 testis compared to wild-type controls.

In summary, these results suggest that Tert constitutive expression in the testis results in a preservation of the male germ line in the context of an organism.

Telomere length in wild-type and K5-mTert mice

We have previously described that basal layer skin keratinocytes from young K5-mTert mice have a similar telomere length to that of age-matched wild-type controls (Gonzalez-Suarez et al., 2001). To address whether the differences in survival between wild-type and the T1 and T8 K5-mTert transgenic mice could be attributed to differences in telomere length in the context of aging, we performed quantitative telomere FISH on age-matched 2-year-old mice from the different genotypes (Materials and methods). First, we measured telomere length in basal layer skin keratinocytes, which have the highest transgene expression (Gonzalez-Suarez et al., 2002), as well as on kidney sections. Figure 6a shows telomere length distributions in basal layer skin keratinocytes from individual wild-type and K5-mTert mice of both transgenic lines (2-year-old mice). It is interesting to note that there is a high mouse-to-mouse variation in telomere length distribution within each genotype (Figure 6a). When grouped by genotype, there was a slight increase in the average telomere fluorescence in T1 and T8 mice compared to wild-type controls, which reached statistical significance only in the case of the T8 K5-mTert mice (Figure 6a). To test whether these findings could be extended to the kidney, which showed an important difference in senile lesions between genotypes, we measured telomere length in kidney sections from 2-year-old wild-type and K5-mTert mice from both T1 and T8 transgenic lines. As shown in the case of skin keratinocytes, there was a slight increase in average telomere fluorescence in T1 and T8 mice compared to wild-type controls, and this increase reached statistical significance only in the case of the T8 K5-mTert mice (Figure 6b). These very minor differences in telomere length between genotypes do not correlate with the differences in survival, since mice from the T1 K5-mTert line are the ones with the greater lifespan extension (Figure 1). These results suggest that the effect of telomerase overexpression on mouse lifespan is independent of a role of telomerase in rescuing short telomeres.

Figure 6
figure6

Telomere fluorescence in skin and kidney sections from 2-year-old mice. Quantification of telomere fluorescence in skin (a) and kidney (b) sections from wild-type and K5-mTert mice from the T1 and T8 transgenic lines. Each dot represents the average telomere fluorescence expressed in arbitrary units of one single nucleus. n=total number of nuclei analysed per mouse or genotype. On the left, telomere length distributions for individual mice of the indicated genotype. On the right, telomere length distributions are grouped by genotype. The average telomere fluorescence for each genotype is also shown together with the P-value for Student's t test

Discussion

Critically short telomeres, due to lack of telomerase activity in the context of the telomerase-deficient mice (Terc−/−), result in a general loss of organismal viability. In particular, telomerase-deficient mice with short telomeres show defects in all highly proliferative tissues (Herrera et al., 1999a, 1999b; Rudolph et al., 1999), as well as in vital organs such as the heart and the kidney (Ruiz-Torres et al., 2002; Leri et al., 2003), resulting in decreased survival with age (Herrera et al., 1999a, 1999b; Rudolph et al., 1999). This general loss of tissue fitness is accompanied by increased genomic instability, increased apoptosis, and loss of proliferative capacity (Herrera et al., 1999a, 1999b; Rudolph et al., 1999; Leri et al., 2003). Telomerase reintroduction in Terc−/− mice with critically short telomeres is sufficient to elongate the shortest telomeres and to prevent chromosomal instability, as well as degenerative pathologies in these mice, such as testis atrophy, bone marrow aplasia, and atrophy of the intestinal epithelium (Samper et al., 2001). Similarly, telomerase also prevents premature senescence of tissue-cultured fibroblasts derived from patients suffering a variety of premature aging syndromes through its ability to rescue short telomeres (Ouellette et al., 2000; Wyllie et al., 2000; Choi et al., 2001).

Recent evidence suggests that telomerase may favor survival and growth independently of its role in elongating critically short telomeres (reviewed in Blasco, 2002). Part of this evidence comes from the study of mouse models with altered telomerase expression in the absence of significant changes in telomere length. In particular, early generation telomerase-deficient mice are more susceptible to the genotoxic effects of ionizing radiation and MNU treatment than wild-type mice despite showing comparable telomere length (Goytisolo et al., 2000; Gonzalez-Suarez et al., 2003). In addition, early generation telomerase-deficient mice show abnormal heart function parameters such as aberrant left ventricular developed pressure and increased myocyte volume (Leri et al., 2003), as well as imbalanced oxidant capacity in the renal cortex (Ruiz-Torres et al., 2002) when compared to wild-type mice. These findings suggest that telomerase activity may have a protective role in organismal aging, which seems to be independent of telomere length. In addition, we have also shown that early generation telomerase-deficient mice are more resistant to (DMBA+TPA)-induced skin tumors (Gonzalez-Suarez et al., 2000; Nat Genet, 2000), as well as more resistant to MNU-induced tumors (Gonzalez-Suarez et al., 2003), suggesting that telomerase deficiency protects against tumorigenesis independently of telomere length. In support of this, mice that express constitutive telomerase, such as K5-mTert mice, show a faster rate of wound-healing and a higher proliferation of skin keratinotyces in response to phorbol esters than wild-type mice with a similar telomere length, suggesting a role of telomerase in proliferation (Gonzalez-Suarez et al., 2001). K5-mTert mice also show an increased susceptibility to both induced and spontaneous tumorigenesis when compared to wild-type mice (Gonzalez-Suarez et al., 2001, 2002). Similarly, two independent telomerase transgenic mouse models, Lck-mTert mice (Tert is overexpressed in thymocytes) and β-actin-mTert mice (Tert is ubiquitously expressed) also showed increased tumors with age (Artandi et al., 2002; Canela et al., 2004). All together, these findings suggest that telomerase has a role in promoting tumorigenesis at the same time that it may prevent or delay some pathologies associated with aging by increasing tissue fitness. Importantly, the data presented here provides evidence for these antagonistic roles of telomerase in cancer and aging. In particular, we show that K5-mTert mice show an increase in the maximum lifespan compared to wild-type mice, in spite of an initial decreased survival due to the higher incidence of cancer. In other words, telomerase expression impacts on the normal aging pattern of K5-mTert mice, by increasing the risk of neoplasias at the same time that it decreases the incidence of degenerative lesions. Similar scenarios have been proposed for the antagonistic roles of tumor suppression in cancer and aging (Roussel, 2003), as well as for a role of myc in promoting both cancer and tissue fitness through increasing cell competition (Moreno and Basler, 2004).

In particular, we show here that Tert overexpression preserves the integrity of the male germ line in K5-mTert mice compared to wild-type controls, in agreement with the fact that K5-mTert mice overexpress Tert in testis (Gonzalez-Suarez et al., 2002). This protection of the male germ line in old K5-mTert mice compared to similarly aged wild-type controls suggests that the high levels of telomerase activity allow the maintenance of a healthy and functional testicular histology independent of the role of telomerase in telomere length maintenance (Lee et al., 1998). This is in accordance with the recent finding that it is possible to generate spermatocytes and spermatids from Tert-immortalized mouse type A spermatogonial cells (Feng et al., 2002), again indicating that Tert can immortalize certain mouse cell types in a manner that seems to be independent of telomere length.

In addition, we describe here a decrease in the frequency of pathologies of the kidney in aged K5-mTert mice compared to wild-type controls. The fact that Tert overexpression protects old K5-mTert mice against kidney lesions is in agreement with the observation that K5-mTert transgene is slightly overexpressed in the kidney in 1-year-old mice (fourfold) compared to age-matched wild-type controls. In turn, this could act in synergy with a general better health of the K5-mTert mice compared to the wild-type littermates, in agreement with the fact that K5-mTert is expressed in a wide variety of tissues including the digestive tract and lymphoid tissues. In this regard, we have recently reported kidney dysfunction in late generation telomerase-deficient mice with short telomeres, which show premature loss of organismal viability (Ruiz-Torres et al., 2002). Furthermore, early generation telomerase-deficient mice also showed significant alterations in kidney function when compared to wild-type littermates, suggesting a protective role of telomerase in kidney function which appeared to be independent of telomere length (Ruiz-Torres et al., 2002). It is possible that the decreased incidence of renal degenerative pathologies in K5-mTert mice could contribute to the increased survival of these mice compared to the wild-type controls at old ages.

Therefore, Tert constitutive expression results in a decreased incidence of degenerative pathologies in the mouse, which is accompanied by an increase in hyperproliferative lesions. The mechanisms underlying these telomere-length independent effects of Tert overexpression could be related to the fact that K5-mTert mice show a higher proliferative response upon mitogenic stimuli, such as treatment with phorbol esters or wound-healing assays (Gonzalez-Suarez et al., EMBO, 2001). It is possible that the increased survival of K5-mTert mice could be due to a selection of the fitness during early life, also it could be related to a role of telomerase on tissue renewal. Other groups have also shown that Tert overexpression increases survival and proliferation (Lu et al., 2001; Mattson and Klapper, 2001; Smith et al., 2003). A role of Tert in promoting proliferation may be beneficial for maintaining tissue fitness and at the same time it may promote hyperplasic lesions and tumorigenesis with aging.

In summary, we describe here for the first time in a mammalian organism that telomerase overexpression results in significant lifespan extension, which is coincidental with increased tissue fitness. Furthermore, we show that this effect of telomerase overexpression cannot be explained by a role of telomerase in rescuing critically short telomeres. This is in agreement with the fact that mice have very long telomeres and mouse aging is unlikely to be the result of critical telomere shortening. These findings suggest that telomerase regulation has a direct impact on mouse aging, and predict a similar scenario in the case of human aging.

Materials and methods

Generation of K5-mTert mice

K5-mTert mice were generated as previously described (Gonzalez-Suarez et al., 2001). In particular, all the K5-mTert and wild-type mice used for the analysis were obtained after crossing of the founders (C57Bl/6xDBA/2) with C57Bl/6xDBA/2 females. Genetic background was stabilized after 4–5 generations. In the different experiments described here, K5-mTert mice were always compared to their wild-type littermates, which share identical genetic background. All wild-type and K5-mTert mice from lines T1 and T8 were aged at the National Center of Biotechnology animal facility, where pathogen-free procedures are employed in all mouse rooms. Quarterly health monitoring reports were negative for all pathogens in accordance with Federation of European Laboratory Animal Science Association recommendations.

Telomere Q-FISH analysis on skin and kidney paraffin sections

For Q-FISH determinations, paraffin-embedded skin and kidney sections from 2-year-old mice were hybridized with a PNA-tel probe, and fluorescence intensity of telomeres was determined as described (Gonzalez-Suarez et al., 2001). Slides were deparaffinated in three xylene washes and then treated with 100, 95, and 70% ethanol. Keratinocyte nuclei for skin and nuclei from tubules of kidney were captured for each mouse at a × 100 magnification and the telomere fluorescence was integrated using spot IOD analysis in the TFL-TELO program. For statistical analysis, a two-tailed Student's t test was used to assess significance.

Real-time quantitative PCR for Tert mRNA detection

Two 1-year old mice of each K5-mTert transgenic line T1 and T8 as well as four wt mice were killed and total RNA was prepared from kidney using Trizol (Invitrogen). For testis, two 1-year-old wild-type mice and one animal of each T1 and T8 K5-mTert transgenic lines were killed and total RNA was prepared. Reverse transcription was performed with random hexamers with Superscript II (Invitrogen) following the manufacturer's instructions. Real-time PCR was performed for two dilutions in duplicates on an ABI Prism 7700 Sequence Detector (Applied Biosystems), using SYBR Green (Applied Biosystems), where each reaction contained 1 × SYBR-Green mix, 3 mM MgCl2, 0.4 μ M each primer, 0.5 mM dNTPs, and 5 μl of the sample. For each mRNA sample, Tert expression was corrected by the actin mRNA content in each sample. The program was an initial incubation of 2 min at 50°C, followed by 10 min at 95°C and 40 cycles of 15 s 95°C, 2 min 15 s 68°C, 15 s 95°C, 20 s 63°C, 15 s 95°C. In each cycle, the melting point of the product was determined. The PCRs employed a set of primers specific for the Tert gene: TERT-F, IndexTerm5′ GGATTGCCACTGGCTCCG 3′; TERT-R, IndexTerm5′TGCCTGACCTCCTCTTGTGAC 3′. For Actin: ACTIN-F, IndexTerm5′GGCACCACACCTTCTACAATG 3′, ACTIN-R IndexTerm5′GTGGTGGTGAAGCTGTAG 3′.

Histopathological analyses

Mice designated as ‘moribund’ were allowed to live out their normal lifespan or were killed when they showed evident signs of suffering or illness. Mice designated as ‘sacrificed’ were killed between weeks 91 and 132 after birth even if they did not show any signs of disease. Full-body histopathological analysis was performed on all animals that died (‘moribund’) or were killed as described to score for age-related lesions or ‘senile’ lesions. Tissues were fixed in 10% PBS-buffered formalin and embedded in paraffin. Sections of 4 μm were stained with hematoxilin–eosin. Images were captured with an Olympus-Vanox microscope.

The types of lesions that were considered ‘senile’ lesions were:

  1. i)

    Hyperplasias and hypertrophies. Lesions common in aged mice including gastric mucosal hyperplasia (gastric plaque), intestinal mucosal hyperplasia, mammary acinar hyperplasia, cystic endometrial hyperplasia (HEQ) in uterus, prostate hyperplasia, hyperplastic pancreatic islets, and thyroid follicular hyperplasia (Table 1). Hyperplasias present in lymphoid tissues were not included since they could also appear due to tumors or infections.

  2. ii)

    Cysts and dilatations in different tissues.

  3. iii)

    Inflammation in several organs: normally occurred in liver and seminal vesicles, although their severity was minimal in the majority of the mice. In the case of the kidney, however, the occurrence of glomerulonephritis, a renal inflammatory/degenerative pathology associated with aging, was common.

  4. iv)

    Atrophies and degenerations in various tissues: adipose subcutaneous tissue, uterus, ovary, testis, pancreas, spleen, and kidney. In the case of the kidney, amyloidosis (deposition of amyloid fibrils), hemosiderosis, and lipidosis (deposition of pigments), as well as calcifications were also observed (Table 2).

Statistical analyses

We used Kaplan–Meier analysis to calculate the significance of differences in survival between the different genotypes, including a long-rank data for Kaplan–Meier comparisons after week 100. The Tukey's test was used to compare the longevity of the 10% longest-lived mice from each genotype. The statistical significance of the different pathologies was calculated using the χ2 test in a 2 × 2 contingency table. The Kaplan–Meier, the Tukey's test, and the χ2 test were performed using the GraphPad Prism software. To investigate whether the death rate varied over time in the different genotypes, we used a Cox proportional hazards model as fit by the SAS software system using the PHREG procedure.

References

  1. Artandi SE, Alson S, Tietze MK, Sharpless NE, Ye S, Greenberg RA, Castrillon DH, Horner JW, Weiler SR, Carrasco RD and DePinho RA . (2002). Proc. Natl. Acad. Sci. USA, 99, 8191–8196.

  2. Aviv A and Aviv H . (1998). Hum. Genet., 103, 2–4.

  3. Blackburn EH . (2001). Cell, 106, 661–673.

  4. Blasco MA . (2002). Nat. Rev. Cancer, 2, 627–633.

  5. Blasco MA, Lee HW, Hande MP, Samper E, Lansdorp PM, DePinho RA and Greider CW . (1997). Cell, 91, 25–34.

  6. Canela A, Martín-Caballero J, Flores JM and Blasco M . (2004). Mol. Cell. Biol., 24, 4275–4293.

  7. Cawthon RM, Smith KR, O’Brien E, Sivatchenko A and Kerber RA . (2003). Lancet, 361, 393–395.

  8. Chan SW and Blackburn EH . (2002). Oncogene, 21, 553–563.

  9. Choi D, Whittier PS, Oshima J and Funk WD . (2001). FASEB J., 15, 1014–1020.

  10. Feng LX, Chen Y, Dettin L, Pera RA, Herr JC, Goldberg E and Dym M . (2002). Science, 297, 392–395.

  11. Gonzalez-Suarez E, Flores JM and Blasco MA . (2002). Mol. Cell. Biol., 22, 7291–7301.

  12. Gonzalez-Suarez E, Goytisolo FA, Flores JM and Blasco MA . (2003). Cancer Res., 63, 7047–7050.

  13. Gonzalez-Suarez E, Samper E, Flores JM and Blasco MA . (2000). Nat. Genet., 26, 114–117.

  14. Gonzalez-Suarez E, Samper E, Ramirez A, Flores JM, Martin-Caballero J, Jorcano JL and Blasco MA . (2001). EMBO J., 20, 2619–2630.

  15. Goytisolo FA, Samper E, Martin-Caballero J, Finnon P, Herrera E, Flores JM, Bouffler SD and Blasco MA . (2000). J. Exp. Med., 192, 1625–1636.

  16. Greider CW and Blackburn EH . (1985). Cell, 43, 405–413.

  17. Harley CB, Futcher AB and Greider CW . (1990). Nature, 345, 458–460.

  18. Herrera E, Samper E and Blasco MA . (1999a). EMBO J., 18, 1172–1181.

  19. Herrera E, Samper E, Martin-Caballero J, Flores JM, Lee HW and Blasco MA . (1999b). EMBO J., 18, 2950–2960.

  20. Lee HW, Blasco MA, Gottlieb GJ, Horner II JW, Greider CW and DePinho RA . (1998). Nature, 392, 569–574.

  21. Leri A, Franco S, Zacheo A, Barlucchi L, Chimenti S, Limana F., Nadal-Ginard B, Kajstura J, Anversa P and Blasco MA . (2003). EMBO J., 22, 131–139.

  22. Lu C, Fu W and Mattson MP . (2001). Brain Res. Dev. Brain Res., 131, 167–171.

  23. Mattson MP and Klapper W . (2001). J. Neurosci. Res., 63, 1–9.

  24. Mohr U, Dungworth DL, Capen CC, Carlton WW, Sundberg JP and Ward JM . (1996). Pathobiology of the Aging Mouse Vol 2, ILSI Press, pp. 505.

  25. Moreno E and Basler K . (2004). Cell, 117, 117–129.

  26. O’Sullivan JN, Bronner MP, Brentnall TA, Finley JC, Shen WT, Emerson S, Emond MJ, Gollahon KA, Moskovitz AH, Crispin DA, Potter JD and Rabinovitch PS . (2002). Nat. Genet., 32, 280–284.

  27. Ouellette MM, McDaniel LD, Wright WE, Shay JW and Schultz RA . (2000). Hum. Mol. Genet., 9, 403–411.

  28. Roussel MF . (2003). Cancer Cell, 2, 434–435.

  29. Rudolph KL, Chang S, Lee HW, Blasco M, Gottlieb GJ, Greider C and DePinho RA . (1999). Cell, 96, 701–712.

  30. Ruiz-Torres MP, Perez-Rivero G, Franco S, Blasco MA, Diez-Marques ML and Rodriguez-Puyol D . (2002). J. Am. Soc. Nephrol., 13, 163A.

  31. Samper E, Flores JM and Blasco MA . (2001). EMBO Rep., 2, 800–807.

  32. Samper E, Goytisolo FA, Slijepcevic P, van Buul PP and Blasco MA . (2000). EMBO Rep., 1, 244–252.

  33. Smith LL, Coller HA and Roberts JM . (2003). Nat. Cell Biol., 5, 474–479.

  34. Wyllie FS, Jones CJ, Skinner JW, Haughton MF, Wallis C, Wynford-Thomas D, Faragher RG and Kipling D . (2000). Nat. Genet., 24, 16–17.

Download references

Acknowledgements

We are indebted to Rosa Serrano, Elisa Santos, and Jessica Freire for mouse care and genotyping. We thank Todd Devries for statistical calculations and discussion, and Susa Alcamí for proofreading of the manuscript. EG-S is a predoctoral fellow of the Fondo de Investigaciones Sanitarias (FIS). CG is supported by a Marie Curie Fellowship within the 6th EU Framework Programme. MAB laboratory is funded by The Swiss Bridge Cancer Research Award 2000, by The Josef Steiner Cancer Research Award 2003, by grants GEN20014856C1308 and SAF20011869 from Ministry of Science and Technology of Spain, by grant 08.1/0054/01 from the CAM, by EU grants FIGHCT199900009, FIGHCT199900002, QLG1199901341, FIS5200200078, and by the Department of Immunology and Oncology.

Author information

Affiliations

Authors

Corresponding author

Correspondence to María A Blasco.

Additional information

Supplementary Information accompanies the paper on Oncogene website (http://www.nature.com/onc)

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

González-Suárez, E., Geserick, C., Flores, J. et al. Antagonistic effects of telomerase on cancer and aging in K5-mTert transgenic mice. Oncogene 24, 2256–2270 (2005). https://doi.org/10.1038/sj.onc.1208413

Download citation

Keywords

  • telomerase
  • cancer
  • aging
  • mouse models
  • maximum lifespan

Further reading

Search

Quick links