Constitutive telomerase activity in cells with tissue-renewing ... - Nature

Oncogene (1998) 16, 381 ± 385  1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00

Constitutive telomerase activity in cells with tissue-renewing potential from estrogen-regulated rat tissues Andrzej K Bednarek, Yilin Chu and C Marcelo Aldaz The University of Texas M.D. Anderson Cancer Center, Department of Carcinogenesis, Science Park ± Research Division, Smithville, Texas 78957, USA

To investigate the role of telomerase in estrogenregulated rodent tissues, we assayed the activity levels of this enzyme and measured cell proliferation and indicators of cellularity in vagina, mammary gland, and uterus from virgin, pregnant, ovariectomized, and ovariectomized estradiol-treated rats. No association was observed between telomerase activity and increased cell proliferation. Telomerase activity was signi®cantly higher (P=0.003) in vagina obtained from ovariectomized rats (very low proliferation) than in vagina from ovariectomized and estradiol-treated rats (high proliferation, high dierentiation). The high telomerase levels observed in vagina from ovariectomized rats indicates that the same epithelial compartment (i.e., basal layer) that has the potential to reconstitute the epithelium also contains the cells that express telomerase. The lower telomerase activity in the keratinized (dierentiated) vagina was probably due to dilution of the number of telomerase-producing cells by the terminally dierentiated non-telomerase-producing cells. Similar results were observed in uterus from ovariectomized versus ovariectomized and estradiol-treated rats. Telomerase activity was highest in uterus from pregnant rats. Telomerase levels in samples from total mammary gland fat pads varied considerably between groups and appeared to be representative of the amount of epithelium present in the sample. Interestingly, when mammary gland samples from the same animals were obtained from pure epithelial organoid preparations, no dierences in telomerase activity could be distinguished between animals or groups. Overall these data suggest that telomerase activity, particularly in rat vagina and uterus, appears to be associated with a cell subpopulation showing proliferative and tissue reconstitution potential and not directly associated with proliferation status per se.

Telomerase activity was also detected in normal bone marrow and germ line cells (reviewed in Holt et al., 1996a). More recently, however, this enzyme was found to be normally expressed in adult epithelial tissues, apparently in association with the subcompartment that contains cells with self-renewing potential (Yasumoto et al., 1996; HaÈrle-Bachor and Boukamp 1996; Chiu et al., 1996). Studies in in vitro models have suggested that induction of terminal dierentiation is associated with a cessation of telomerase expression (Sharma et al., 1995; Bestilny et al., 1996; Holt et al., 1996b; Albanell et al., 1996). Reports from some laboratories have postulated an association between telomerase expression and cell proliferation in vitro (Chadeneau et al., 1995a). In any case, we know very little about the physiologic regulation of telomerase expression. Recently, however, it has been suggested that telomerase expression may be hormonally regulated in rat prostate (Meeker et al., 1996). Therefore, we chose to investigate the role of telomerase expression in estrogen-regulated tissues. We assayed the activity levels of this enzyme and measured cell proliferation and indicators of cellularity in rat vagina, uterus, and mammary gland. We hypothesized that, after ovariectomy, the cells with tissue-reconstituting ability would retain telomerase activity so that the eects of inducing proliferation and dierentiation by estradiol (E2) treatment could be investigated. To that end, by using a semiquantitative method for determining telomerase levels in the aforementioned tissues, we compared telomerase expression in virgin, pregnant and ovariectomized rats with and without estradiol treatment.

Keywords: telomerase; stem cells; estrogen; proliferation

Macroscopic and histological inspection revealed that ovariectomy, as expected, had produced the atrophy of the three hormonally regulated organs, vagina, uterus, and mammary gland. The comparison of telomerase levels in vagina from the dierent animal groups (virgin, ovariectomized, ovariectomized plus E2-treated, and pregnant) showed activity levels were highest in samples from ovariectomized rats (Figure 1a and b, P=0.003, ovx vs ovx+E2). It is a characteristic of this strati®ed epithelium to respond rapid and dramatically to estradiol treatment. At 48 h after treatment with estradiol, a statistically signi®cant increase in cell proliferation was evident (Figure 1c and e, P=0.02, ovx. vs ovx+E2). Vaginal epithelial thickness was used as an indicator of epithelial cellularity and differentiation. As expected, the increase in dierentiation and cellularity after estradiol treatment produced a much

Introduction The ribonucleoprotein telomerase is the enzyme in charge of synthesizing single-strand telomeric DNA repeats. Since most immortal cell lines and tumor cells express telomerase, it has been speculated that expression of this enzyme is necessary to escape cellular senescence (reviewed in Holt et al., 1996a). Correspondence: CM Aldaz Received 29 May 1997; revised 8 September 1997; accepted 9 September 1997

Results and discussion

Telomerase in estrogen regulated tissues AK Bednarek et al

382

thicker and heavily keratinized epithelium (Figure 1d and e, P50.0001, ovx. vs ovx+E2). The vaginal epithelium from pregnant rats was also thicker, but

a

— ITAS

Virgin

b

OVX

c

OVX + Estradiol

P LB

Pregnant

d

this was due to a mucinouslike dierentiation (i.e., transdierentiation, Horvat et al., 1992) with very low cell proliferation (Figure 1c and d). The high telomerase levels observed in vagina from ovariectomized rats indicates that the same epithelial compartment (i.e., basal layer) that had the potential to reconstitute the epithelium also contains the cells that express telomerase (Figure 1e). Upon induction of proliferation and dierentiation, a tenfold decrease was seen in telomerase levels, mainly due to dilution by the dierentiated non-telomerase-expressing cells of the vaginal epithelium. The epithelium became six times thicker. This appears to indicate that the true telomerase levels, and possibly the population of cells expressing it, remained relatively constant. Similarly, vagina from pregnant rats was also thick and telomerase levels were low. In addition, telomerase appears not to be associated with increased proliferation (Figure 1c and e). Supporting this was the comparison of the vaginal epithelium from ovariectomized rats treated with E2 with that from pregnant rats: although samples from both groups had a similar thickness (Figure 1d), they had very dierent levels of cell proliferation but the same low telomerase levels. In uterus, with its simple epithelium, the dierences between the groups were less evident (Figure 2). The atrophic uterus of ovariectomized animals showed the same telomerase levels as that of virgin rats (Figure 2a and b). Ovariectomized rats treated with estradiol

a — ITAS

e

Virgin Figure 1 (a) Telomerase activity in rat vagina as detected by the TRAP assay. Each lane represents the results from a single animal. The four experimental groups analysed were, as shown, virgin (intact) rats n=5, ovariectomized rats (ovx) n=4, ovx and estradiol-treated n=5 and pregnant rats n=3. P, positive control cell line, 100 cells equivalent, LB, lysis buer, negative control, ITAS, internal standard. Note that highest telomerase activity is observed in samples from ovariectomized animals. (b) Average relative telomerase activity+s.e. from the dierent experimental groups as indicated. Relative telomerase activity is expressed as the percentage of positive control activity after normalization to ITAS and background subtraction. (c) Cell proliferation in rat vagina expressed as the percentage of BrdU labeled cells per total number of basal layer cells in the vaginal epithelium. (d) Epithelial thickness in rat vagina as an indicator of total epithelial cellularity. (e) Representative photomicrographs of tissue sections from rat vagina samples, after anti-BrdU staining, hematoxylin counterstaining6150. Left panel, sample from ovariectomized rat, note the very thin monolayer epithelium with very low cell proliferation, arrows point to the few cells that incorporated BrdU. Right panel, sample from ovariectomized rat 48 h after treatment with estradiol. Note the very high number of BrdU positive nuclei among the basal layer cells and the dramatic increase in dierentiation and epithelial thickness

b

OVX + Estradiol

OVX

P LB

Pregnant

c

Figure 2 (a) Telomerase activity in rat uterus from the dierent experimental groups. Each lane represents the results from a single animal. (b) Average relative telomerase activity+s.e. determined as described in Figure 1b. Cell proliferation in uterine epithelium expressed as the percentage of BrdU labeled cells per total number of basal layer cells


tended to have lower telomerase levels than untreated animals (P=0.1, ovx vs ovx+E2). The highest telomerase levels were observed in uterus from pregnant rats. Again, as in vaginal epithelium, we a

observed no association between telomerase levels and cell proliferation. The endometrium of ovariectomized rats showed an extremely low labeling index (Figure 2c), but the telomerase levels were comparable to those

Total Mammary Gland Fat Pad — ITAS

Virgin b

OVX

OVX + Estradiol

P LB

Pregnant

Mammary Gland Epithelial Organoids — ITAS

Virgin

c

OVX

d

OVX + Estradiol

P LB

Pregnant

e

Figure 3 (a) Telomerase activity in samples from total mammary gland fat pads. Each lane represents the results from a single animal. Note that the highest activity is observed in samples from pregnant animals and as such appears representative of the amount of mammary epithelium present in the sample, see also Figure 3e (b) Telomerase activity in samples from mammary epithelial organoids obtained after enzymatic dissociation of the same mammary fat pads shown in Figure 3a. Note that no dierence in telomerase activity levels is observed among the dierent groups. Representative microphotographs from tissue sections of mammary gland samples obtained from animals representing the dierent experimental groups, H&E stained6150. (c) sample from normal virgin rat: (d) mammary gland from ovariectomized animal, (e) mammary gland from pregnant rat, note the dramatic increase in the number of mammary epithelial cells

383


384

of intact animals. In the uterine epithelium, the telomerase-producing cells are not diluted by layers of dierentiated cells like in the strati®ed epithelium of vagina. Findings in uterus again agree with the notion that there is a relatively constant number of cells expressing telomerase and that to this tissue compartment belongs the characteristic potential for regenerating the epithelium. Similarly to our ®ndings in vagina and uterus from ovariectomized rat females, high telomerase activities were previously described in atrophic prostate and seminal vesicles from castrated rat males (Meeker et al., 1996). Telomerase activity ¯uctuations were also previously observed in human normal endometrium associated with epithelial cell proliferation changes during menstrual cycle (Kyo et al., 1997). In contrast to that study, however, we found no correlation between proliferation and telomerase expression. Furthermore, we found high levels of telomerase activity in samples from ovariectomized animals, whose status could be comparable with postmenopausal status in humans (Kyo et al., 1997). Finally, we also analysed telomerase levels in mammary glands. In normal rats, the mammary epithelial component represents only a limited percentage of the total protein content of a mammary fat pad sample. To address the signi®cance of this problem in telomerase detection, in our study we compared mammary gland fat pad samples obtained from the same animals in two dierent ways: (a) from total mammary gland fat pad and (b) from pure epithelial organoids obtained from enzymatically dissociated mammary fat pads. The comparison of telomerase levels using samples from total mammary fat pads shows considerable variability between the dierent groups. Mammary glands from ovariectomized rats showed very weak telomerase ladders (Figure 3a). On the other hand, mammary fat pad samples from pregnant rats showed the strongest telomerase ladders (Figure 3a). Interestingly, however, when the mammary gland samples from the same animals were obtained from epithelial organoid preparations, telomerase activity was much higher and no dierences could be distinguished between the dierent groups. As can be observed in Figure 3b, virgin, ovariectomized, and pregnant rats showed approximately the same levels of telomerase activity in mammary gland epithelial cells. In addition, no signi®cant dierences in telomerase activity were observed after estradiol treatment in ovariectomized rats. Thus, the pattern of telomerase expression in the mammary epithelium appears, up to certain extent, dierent from that observed in epithelium from vagina and uterus. The epithelial component of the mammary gland undergoes a dramatic development and specialized dierentiation during pregnancy (Figure 3e). However, in contrast to what we observed in vagina and more similarly to what we observed in uterus, we did not detect in the simple columnar epithelium of mammary gland a dilution of the telomerase levels at the more dierentiated pregnant stage. In addition, the telomerase analysis of samples from total mammary fat pads (Figure 3a), appears to indicate that telomerase levels are proportional to the representation of the epithelial component in the sample, since samples from

pregnant rats, which have a greatly expanded epithelial compartment (Figure 3e), showed the highest activity. In addition, the comparison between the two methods for obtaining of samples also indicates that caution should be exercised when normal breast samples from either humans or animals are cataloged as negative for telomerase activity, since samples are usually obtained from the total tissue without any speci®cation of the total contribution of the epithelial component to the sample. In summary, our data indicate that telomerase activity is not associated with increase proliferation in rat vagina, uterus, and mammary gland. The decrease in telomerase activity observed after estradiol induction of proliferation and dierentiation, in particular in vagina, appears to be the result of dilution by the dierentiated non-telomerase-expressing cells. The very high telomerase levels observed in tissues from ovariectomized animals indicates that the same epithelial compartment that has the potential to reconstitute the epithelium (i.e., stem cell containing) also contains the cells that express telomerase. These studies suggest that telomerase activity can be considered a bio marker of proliferative potential rather than a marker of cell proliferation. Further experiments are needed to determine whether all the cells in the epithelial basal layer of the estrogenregulated tissues can express telomerase or whether this ability is only limited to a subset of cells. Materials and methods Animals and their treatment Normal mammary glands, uterus, and vaginas were obtained from BUF/NCI female rats. Two groups of rats were ovariectomized at 8 weeks of age. At 21 weeks of age, one group (®ve rats) was injected with estradiol (E2) (5 mg per rat) given as an estradiol benzoate suspension in cotton seed oil. Rats were euthanized 48 h after estradiol injection. Total number of animals per group were normal virgin n=5, ovx n=5, ovx+E2 n=5 and pregnant n=3. Tissues were then aseptically removed. Part of each sample was analysed by conventional histologic techniques, and the remainder was snap-frozen in liquid nitrogen and stored at 7808C until further analysed. Mammary gland organoids were prepared by dispersion in a solution containing collagenase (800 m) and hyaluronidase (200 m) as previously described (Aldaz et al., 1992). Organoids were recovered from a retention ®lter, ¯ash frozen, and stored at 7808C. Telomerase activity assay For telomerase detection, we used the PCR-mediated telomere repeat ampli®cation protocol (TRAP) (Kim et al., 1994; Piatyszek et al., 1995). As a positive control, an extract from cells with known telomerase activity (human breast cancer line MDA-MB-157, 100 cells equivalent) was used. As a negative control, cell extract was substituted for lysis buer (Wright et al., 1995). As an additional control for the TRAP assay, we used an internal standard (ITAS) (a gift of Dr J Shay) that ampli®es from the same primers (Wright et al., 1995). This internal standard, which consists of a 150 bp DNA product, allows identi®cation of falsenegative tumor samples that might contain Taq polymerase inhibitors. Cell extracts were obtained and TRAP was performed as previously described (Kim et al., 1994; Piatyszek et al., 1995)


but with minor modi®cations (Wright et al., 1995; Bednarek et al., 1995). Two microliters of tissue extract (protein concentration 0.5 mg/ml) were used per assay. The CX primer, ITAS and Taq DNA polymerase (7 m per assay) were added to each sample after 5 min incubation at 908C to make a `hot start'. Aliquots (10 ml) of the PCR mixture were analysed on 0.4 mm, 8% nondenaturing, acrylamide gels (20640 cm) run in 0.56TBE buer until the xylene cyanol had migrated 17 cm from the origin. The gels were then dried and exposed for 20 h to hyper®lm MP ®lms (Amersham Corp, Arlington Heights, IL). Following autoradiography, gels were analysed and overnight exposure with a Molecular Dynamics PhosphorImager (Sunnyvale, CA). Area integration of all peaks (except the ®rst ampli®cation product from the bottom) were normalized to the signal from the internal standard and then, after background subtraction, expressed relative to the positive control signal that was run with each experiment. The ®rst PCR product from the bottom was not included because it usually incorporates background from primer-dimer formation (Chadeneau et al., 1995b; Sommerfeld et al., 1996). The method described is only semiquantitative, but it is sucient for the comparative analysis of tumors relative to the same positive control cell extract.

Cell proliferation assay Cell proliferation was determined by establishing the frequency of 5-bromo-2'-deoxyuridine (BrdU) incorporation. To that end, animals received an i.p. dose of BrdU (100 mg/g rat weight in 1.0 ± 1.5 ml of PBS) 30 min prior to being euthanized. Paran sections were stained for antiBrdU according to standard, commonly used procedures. Labeling index represents the percentage of BrdU labeled nuclei per total number of basal layer nuclei. To that end a minimum of 500 nuclei were counted per sample, counting was performed from random microscopic ®elds of view (606). Statistical analysis Statistical analysis was performed using an unpaired t-test algorithm (StatView, Abacus).

Acknowledgements We thank Dr S Sukumar for critical reading of this manuscript and J Ing for art work. This was supported in part by US Public Health Service grant CA59967.

References Aldaz CM, Chen A, Gollahon L, Russo J and Zappler K. (1992). Cancer Res., 52, 4791 ± 4798. Albanell J, Han W, Mellado B, Gunawardane R, Scher HI, Dmitrovsky E and Moore MA. (1996). Cancer Res., 56, 1503 ± 1508. Bednarek A, Budunova I, Slaga TJ and Aldaz CM. (1995). Cancer Res., 55, 4566 ± 4569. Bestilny LJ, Brown CB, Miura Y, Robertson LD and Riabowol KT. (1996). Cancer Res., 56, 3796 ± 3802. Chadeneau C, Siegel P, Harley CB, Muller WJ and Bacchetti S. (1995a). Oncogene, 11, 893 ± 898. Chadeneau C, Hay K, Hirte HW, Gallinger S and Bacchetti S. (1995b). Cancer Res., 55, 2533 ± 2536. Chiu CP, Dragowska W, Kim NW, Vaziri HM, Harley CB and Landsorp PM. (1996). Stem Cells, 14, 239 ± 243. HaÈrle-Bachor C and Boukamp P. (1996). Proc. Natl. Acad. Sci. USA, 93, 6476 ± 6481. Holt SE, Shay JW and Wright WE. (1996a). Nat. Biotec., 14, 836 ± 839. Holt SE, Wright WE and Shay JW. (1996b). Mol. Cell. Biol., 16, 2932 ± 2939. Horvat B, Vrcic H and Damjanov I. (1992). Exp. Cell Res., 199, 234 ± 239.

Kim NW, Piatyszek MA, Prowse KR, Harley CB, West MD, Ho PLC, Coviello GM, Wright WE, Weinrich SL and Shay JW. (1994). Science, 266, 2011 ± 2015. Kyo S, Takakura M, Kohama T and Ionue M. (1997). Cancer Res., 57, 610 ± 614. Meeker AK, Sommerfeld HJ and Coey DS. (1996). Endocrinology, 137, 5743 ± 5746. Piatyszek MA, Kim NW, Weinrich SL, Hiyama K, Hiyama E, Wright WE and Shay JW. (1995). Meth. Cell Sci., 17, 1 ± 15. Sharma HW, Sokolowski JA, Perez JR, Maltese JY, Sartorelli AC, Stein CA, Nichols G, Khaled Z, Telang NT and Narayanan R. (1995). Proc. Natl. Acad. Sci. USA, 92, 12343 ± 12346. Sommerfeld H-J, Meeker AK, Piatyszek MA, Bova GS and Shay JW. (1996). Cancer Res., 56, 218 ± 222. Wright WE, Shay JW and Piatyszek MA. (1995). Nucl. Acids Res., 23, 3794 ± 3795. Yasumoto S, Kunimura C, Kikuchi K, Tahara H, Ohji H, Yamamoto H, Ide T and Utakoji T. (1996). Oncogene, 13, 433 ± 439.

385