In vitro aging of rat lung cells

Available online at www.sciencedirect.com R

Experimental Cell Research 286 (2003) 30 –39

www.elsevier.com/locate/yexcr

In vitro aging of rat lung cells Downregulation of telomerase activity and continuous decrease of telomere length are not incompatible with malignant transformation Fabrice Petitot, Je´roˆme Lebeau, Laurent Dano, Bruno Lectard, Sandrine Altmeyer, Ce´line Levalois, and Sylvie Chevillard* Laboratoire de Cance´rologie Expe´rimentale, De´partement de Radiobiologie et Radiopathologie, Commissariat a` l’Energie Atomique, 92265 Fontenay-aux-Roses, cedex, France Received 19 July 2002, revised version received 25 November 2002

Abstract Most normal mammalian somatic cells cultivated in vitro enter replicative senescence after a finite number of divisions, as a consequence of the progressive shortening of telomeres during proliferation that reflects one aspect of organism/cellular aging. The situation appears more complex in rodent cells due to physiological telomerase expression in most somatic normal tissues, great telomere length, and the difficulties of finding suitable in vitro culture conditions. To study in vitro aging of rat lung epithelial cells, we have developed primary culture conditions adapted to rat fresh lung explants and have studied for 1 year (50 passages) the changes in cellular proliferation and mortality, genetic instability, telomerase activity, telomere length, and tumorigenic potential. We have observed an absence of senescence and/or crisis, a transient genetic instability, the persistence of a differentiated Clara cell phenotype, a steady decrease in telomerase activity followed by a low residual activity together with a continuous decrease in telomere length, a constant rate of proliferation, and the acquisition of tumorigenic potential. The bypass of the growth arrest and the acquisition of long-term growth properties could be explained by the loss of p16INK4a expression, the ARF/p53 pathway not being altered. In conclusion, these results clearly indicate that, in rat lung epithelial cells, in vitro transformation and acquisition of tumorigenic properties can occur even if the telomere length is still decreasing and telomerase activity remains downregulated. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Rat lung cells; Telomere; Telomerase; Transformation; Tumourigenicity

Introduction Most normal mammalian freshly cultivated somatic cells enter replicative senescence after a transient period of proliferation. Recent studies suggest that senescence of cultured cells results from different mechanisms, which depend on the cell’s origin [1]. A well-described mechanism, the “mitotic clock,” depends on an intrinsic cell-division counting mechanism and is characteristic of some human cells such as fibroblasts. This replicative senescence, which de* Corresponding author. CEA, DSV, DRR, LCE, 60 – 68 Avenue du Ge´ne´ral Leclerc, BP6, 92265 Fontenay-aux-Roses cedex, France. Fax: ⫹33-1-46-54-88-86. E-mail address: [email protected] (S. Chevillard).

termines cellular lifespan, is the consequence of the progressive shortening and uncapping of telomeres during proliferation/replication [2] and may reflect aspects of organism aging, as was demonstrated in cells lacking telomerase activity. Telomeric DNA consists of (TTAGGG) sequences that interact with a set of proteins and thus protects chromosome ends from degradation and ligation. Therefore, cellular senescence of cultured cells, which is characterised by a growth arrest, may occur when telomeres become too short to ensure chromosome integrity, thus leading to aberrant fusions and rearrangements. These genetic abnormalities induce typical DNA damage responses, including activation of p53- and Rb-dependent check points that also contribute to senescence [1]. In this context, telo-

0014-4827/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0014-4827(03)00103-4

F. Petitot et al. / Experimental Cell Research 286 (2003) 30 –39

mere shortening could be considered as a tumor suppressor mechanism [3]. Bypass of p14ARF/p53 and p16INK4a/pRb pathways extends cell lifespan but via a “crisis” characterised by massive genetic instability and cell death [1]. Overall, immortalisation of human cells necessitates stabilisation of chromosomes by stopping telomere attrition either by activating telomerase or by an alternative (ALT) recombinational mechanism [4]. The situation appears more complex in rodents, for which senescence is also characterised by a proliferative block but with differences in telomere length, in the dynamics of shortening as well in telomerase regulation, compared with humans. Telomere length ranges from 2 to 20, 20 to 150, and 20 to 100 kb in human [5], mouse [6], and rat [7], respectively. Most rodent normal somatic tissues expressed detectable telomerase activity [8 –10], although telomerase activity is substantially lower than that detected in rodent tumors [8,11–13]. Even if this telomerase activity is not always sufficient to stabilise telomere length in cultured rodent cells, a senescent state occurs after a small number of cell divisions (15–30 population doublings), noticeably before extensive telomere erosion. Different studies, conducted on rodent and human cells, suggest that this growth arrest is not telomere based but could be more related to inappropriate in vitro culture environment [14 –17], such as oxidative damage, inadequate nutrient media, and unrestrained mitogenic stimulation, together causing progressive damage and finally growth arrest. Senescent rodent cells express high levels of negative cell cycle regulators implicated in the p16INK4a/pRB and p19ARF/p53 pathways, but, distinctively from the human cells, alteration of only one of these pathways seems to be sufficient to bypass the growth arrest [1,15,16,18,19]. So, the alterations of the p53 and pRb pathways and of the telomere-shortening pathway are needed to bypass the senescence of human fibroblasts, whereas the abrogation of the p53 or the pRb pathway alone is sufficient to confer unlimited proliferation on mouse embryo fibroblasts. The concept that telomere shortening is not directly implicated in senescence of rodent cells is reinforced by data obtained on telomerase-deficient mice. Probably because telomeres of Mus musculus are very long, after deletion of the mTR gene, which encodes the RNA component of the enzyme [20], mTR⫺/⫺ mice are viable for at least six generations, telomeres shortening at a rate of 4.8 ⫾ 2.4 kb per generation. Moreover, mTR⫺/⫺ cells could be immortalised at the same frequency as that of normal mouse cells, and could be transformed by viral oncogenes and are tumorigenic when xenografted on nude mice. This statement, together with the specificity of telomerase activity and telomerase length in rodents, may explain why rodent cells “immortalise” with such a high frequency compared with human cells. As it is difficult to obtain appropriate culture conditions, the characterisation of in vitro long-term culture of normal rodent cells and more specifically of rat cells is poorly documented. Recently, two studies demonstrated the lack of

31

replicative senescence in normal rat oligodendrocyte precursor cells and in rat Schwann cells [21,22]. In both cases, cells retain a high level of telomerase activity, are diploid, and the cell cycle check points are maintained. Up to now, in vitro aging of epithelial cells was studied in a unique model of rat hepatic epithelial stem-like cells [23,24] in which telomerase repression and telomere erosion are observed at intermediate passages and reactivation of telomerase occurs after a high number of passages and in tumorigenic lineages. To study in vitro aging of rat lung epithelial cells, we have developed primary culture conditions adapted to rat fresh pulmonary explants. Four distinct cell lineages were maintained in low serum concentration medium and were passaged weekly for 1 year. At regular intervals, we analysed survival, proliferation rate, telomerase activity, telomere length, karyotypic instability, and, at the end, tumorigenic potential. During 1 year, the senescent phase was never detected, a transient genetic instability was observed, and, finally, cells were tumorigenic after injection in nude mice. We observed a downregulation of telomerase activity and a constant decrease in telomere length during the in vitro culture, this diminution continuing in xenografted tumors. The constant decrease in telomere length during the culture could indicate that these cell lines are not fully immortalised. These results clearly indicate that, in rat lung epithelial cells, telomerase reactivation and stabilisation of telomere length are not necessary for in vitro transformation and acquisition of tumorigenic properties.

Materials and methods Isolation of rat lung epithelial cells Three-week-old male Sprague-Dawley OFA rats were from Charles River Laboratories (Iffa Credo, France) and were handled according to the French Legislation and the European Directives regarding the care and use of laboratory animals. Two rats (R1 and R2) were anaesthetised by intraperitoneal injection of pentobarbital sodium (4 mg/kg of body weight) (Sanofi, France). After lung removal, explants from the distal part were prepared, mechanically desegregated in saline buffer, and seeded in low serum concentration medium (see cell culture) in a plastic 60-mm/ collagen-coated dish (Iwaki, France). Two primary cultures were performed for each rat (R1.1, R1.2, and R2.1 R2.2). Two weeks later, when cells spread over the culture dish, explants were discarded. Cell cultures Cultures were grown in 5% CO2 at 95% humidity. Rat lung cells were maintained in a low serum concentration medium composed of F12/DMEM (3/1) medium (Seromed, France) supplemented with 1% (vol/vol) foetal calf serum,

32

F. Petitot et al. / Experimental Cell Research 286 (2003) 30 –39

L-glutamine (0.29 mg/ml), HEPES buffer (10 mM), penicillin (50 IU/ml), streptomycin (50 ␮g/ml), kanamycin (50 ␮g/ml) (all from Life Technologies, France), and amphotericin B (1.25 ␮g/ml), epidermal growth factor (10 ng/ml), cholera toxin (0.1 nM), hydrocortisone (0.4 ␮g/ml), insulin (5 ␮g/ml), adenine (200 ␮M), and all trans-retinol (0.2 ␮g/ml) (all from Sigma, Saint Quentin Fallavier, France). The four rat lung cell cultures were maintained for 1 year with 50 passages (200 –230 population doublings). The total number of population doublings was based on the counting of viable cells performed at each passage. From passage 3 to passage 15, cells were trypsinised and replated each week at a 1:8 split ratio and the split ratio was 1:32 from passage 17 to passage 50. At passage 50, R2.1 cells were cultivated for 5 additional passages in a high serum concentration medium: F12/ DMEM (3/1) medium (Seromed) supplemented with 10% (vol/vol) foetal calf serum, L-glutamine (0.29 mg/ml), HEPES buffer (10 mM), penicillin (50 IU/ml), streptomycin (50 ␮g/ml), kanamycin (50 ␮g/ml), and amphotericin B (1.25 ␮g/ml). The A549 human lung adenocarcinoma cell line was from the American Type Culture Collection (Rockville, MD, USA) and was cultivated in RPMI 1640 medium supplemented with 10% (vol/vol) foetal calf serum, L-glutamine (0.29 mg/ml), HEPES buffer (10 mM), penicillin (50 IU/ml), streptomycin (50 ␮g/ml), kanamycin (50 ␮g/ml), and amphotericin B (1.25 ␮g/ml).

Proliferation and survival Cellular proliferation and survival analyses were performed in two or more separate experiments, by scoring at least 200 cells at each passage. Discrimination between viable and dead cells (including dead cells in the supernatant) was performed after trypan blue staining. Immunocytochemistry and histochemistry Rabbit antisera to rat RSAP and rat RCC10 were a kind gift from Dr. Gurmukh Singh, Pittsburgh, PA [25]. Briefly, for in vitro cultivated cells, 50,000 cells were cytospun, fixed in acetone at ⫺20°C for 20 min and then washed for 5 min at room temperature in phosphate-buffered saline (PBS) containing 0.05% Tween 20. To block endogenous peroxidase, cells were treated in 0.3% H2O2 in methanol for 30 min and washed for 5 min in PBS, 1 mg/ml bovine serum albumin (BSA) (PBS-BSA). To block nonspecific binding, cells were then incubated with a goat serum for 20 min at room temperature. After washing in PBS-BSA, cells were incubated with primary antibody (1:4800) at 4°C overnight and washed four times in PBS-BSA. Reactions were revealed by using the ABC alkaline phosphatase kit (Vectastain, Biosys, Paris, France) with the diaminobenzidine substrate (Sigma). For immunohistochemistry, the same protocol was ap-

plied to acetone-fixed frozen sections of either normal lung or tumors that have developed on nude mice. Cell cycle analysis Cells were trypsinised and harvested by centrifugation at 1400 rpm for 10 min, washed in PBS, fixed in 70% cold ethanol, and stored at ⫺20°C. Before analysis, cells were washed in PBS and stained in PBS containing 25 ␮g/ml propidium iodide (Sigma) and 50 ␮g/ml RNase A (Roche Diagnostics, Meylan, France) for 30 min at 37°C. Samples were analysed by using a FACScalibur flow cytometer (Becton Dickinson, Le pont de claix, France) on at least 20,000 cells. Cellular debris, fixation artefacts, and doublets were gated out with FL2 area and FL2 width parameters. The cell cycle was analysed by using Modfit software (Verity, Becton Dickinson). Tumorigenicity in athymic mice To test the tumorigenic potential, 5 ⫻ 106 cells were suspended in 0.2 ml of PBS and injected subcutaneously into 4-week-old athymic female Swiss nude (nu/nu) mice. Each week for at least 3 months, animals were carefully examined for tumor growth. RNA extraction and quantitative real-time reverse transcription–polymerase chain reaction (RT-PCR) For the first-strand cDNA synthesis, 1 ␮g of total RNA, extracted with RNA Plus (Quantum Biotechnologies, Meylan, France), was mixed with 1 ␮M of random primer p(dN)6 (Roche Diagnostic) and 250 ␮M dNTP (Life Technologies, France) in a 1⫻ reverse buffer [6.7 mM MgCl2, 67 mM Tris-HCl, pH 8.8, 16.6 mM (NH4)2SO4], then incubated for 5 min at 65°C and placed on ice before adding 200 units of M-MLV (Life Technologies) in a final volume of 20 ␮l and further incubated for 30 min at 42°C, followed by incubation for 3 min at 72°C to inactivate reverse transcriptase. PCR reactions were performed by using SYBR Green PCR Core Reagent (Perkin Elmer, France), on an ABI PRISM 7700 Sequence Detector apparatus, and analysed with the dedicated Gene Amp software (Perkin Elmer). For a given gene, quantitation was performed by reference to a standard curve obtained by serial dilutions of R1.1 (passage 1) cDNA handled separately but concomitantly with experimental samples. This standard curve is used to check the linearity of the PCR reaction and any potential variations in the yield of amplification between each experiment. For each sample, measures of gene expression were normalised as a function of the expression of the 18S ribosomal RNA. Primers [forward (F) and reverse (R)] and PCR conditions are for (1) p16INK4a: CTTCACCAAACGCCCCGAACAC (F) and GCAGAGCATGGGTCGCAGGTTC (R), primer concentration: 0.04 pmol/␮l, MgCl2 concentration: 3 mM, PCR product: 154 bp; (2) p19ARF: GCAGAGCATGGGTCG-

F. Petitot et al. / Experimental Cell Research 286 (2003) 30 –39 Table 1 Immunocytochemical analysis of the four cultures using rat antibodies directed against either lung surfactant apoprotein (RSAP) or Clara cell 10-kDa protein (RCC10)a Cell cultures

R1.1 R1.2 R2.1

R2.2

Passage no.

3 50 3 50 3 12 50 3 12 50

Positive immunostaining (%) RSAP

RCC10

74 90 76 80 83 87 88 75 81 93

74 86 81 89 90 89 83 79 93

a

RSAP antibody reacts with both pneumocyte II and Clara cells, whereas RCC10 is specific to Clara cells. Cytospun cells were analysed at successive passages and at least 500 cells were scored at each point.

CAGGTTC (F) and GCAGAGCATGGGTCGCAGGTTC (R), primer concentration: 0.04 pmol/␮l, MgCl2 concentration: 3 mM, PCR product: 286 bp; (3) p21Cip1: AGCAAAGTATGCCGTCGTCT (F) and CGAAGTCAAAGTTCCACCGT (R), primer concentration: 0.2 pmol/␮l, MgCl2 concentration: 3 mM, PCR product: 188 bp; and (4) 18S: CTCAACACGGGAAACCTCAC (F) and ATGCCAGAGTCTCGTTCGTT (R), primer concentration: 0.2 pmol/ ␮l, MgCl2 concentration: 3 mM, size of the amplified product: 151 bp. Telomerase activity assay Telomerase activity was measured by means of the telomere repeat amplification protocol using the Trapeze telomerase detection kit (Intergen, Gaithersburg, MD, USA). For protein extraction, 106 cells were treated with 200 ␮l of CHAPS lysis buffer (supplied in the kit). PCR amplifications were performed on extracts corresponding to 500 – 10,000 cells (around 0.05–1 ␮g of protein, determined by Bradford assay), as recommended by the manufacturer. For quantitation, serial dilutions of extracts from 50 to 10,000 reference cells were simultaneously amplified, allowing us to check the linearity of the reaction. The relative telomerase activity was set at 1 as the reference. Experiments were performed at least twice on independent cell pellets or tumor aliquots.

33

mean fluorescence signal in G0/G1 cells after subtraction of the background fluorescence signal. The relative telomere length value was calculated as the ratio between the telomere signal of each sample and the signal detected in the control cells (R2.1 cells), with compensation corresponding to the DNA index of G0/G1 cells estimated by flow cytometry for cell cycle analysis. Results Cellular characteristics of rat lung epithelial cell cultures To obtain primary rat lung cell cultures and to follow their transformation in vitro, fresh explants from the distal part of the lung were withdrawn from 3-week-old SpragueDawley rats and were cultivated for 1 year in low serum concentration medium. This medium is known to allow the persistence of epithelial differentiation in cell cultures. The presence in the culture medium of cholera toxin, which specifically inhibits fibroblast proliferation, permits the selective growth of epithelial cells. We have established, through 50 passages, four cell lines (R1.1, R1.2, R2.1, and R2.2, originating from two rats, R1 and R2), which were characterised throughout the transformation process. At early passages, light microscopic examination of the four cell cultures showed the presence of nonciliated cells containing numerous secretory granules and an evident production of surfactant in the culture medium. Fibroblasts, which were present before the first passage, completely disappeared from the culture after the second passage. To characterise these epithelial cells and specifically to determine whether these cells originated from alveolar type II or bronchiolar Clara cells, we performed immunostaining with either a rabbit anti-rat RSAP reacting with both type II pneumocytes and Clara cells or a mouse anti-rat Clara cell secretory protein or RCC10 reacting specifically with Clara cells. The specificity of these antibodies was checked on rat lung paraffin-embedded sections and both alveolar and bronchiolar positive staining was observed with RSAP antibody, while with RCC10 the immunostaining was only observed in

Telomere length analysis The mean length of the telomere was determined by using the Q-FISHFCM method, as recently described [26], by using a fluorescein isothiocyanate (FITC)-labelled (CCCTAA)3 PNA probe (Perkin Elmer). The analysis was performed in a FACScalibur flow cytometer (Becton Dickinson). The telomere fluorescence signal was defined as the

Fig. 1. Growth rate, measured by the mean population doubling, of the four cell cultures (R1.1, R1.2, R2.1, and R2.2) as a function of the passage number.

34

F. Petitot et al. / Experimental Cell Research 286 (2003) 30 –39

Fig. 2. Cell cycle and change in ploidy of the four cell cultures (R1.1, R1.2, R2.1, and R2.2) as a function of the passage number. DNA content was measured on freshly isolated cells at passage 0 (p0) and at different successive passages up to the 50th passage (p50).

bronchiolar cells, alveolar cells remaining totally negative. At passage 3, in the four cultures, 74 – 83% and 74 –90% of cells were positively stained with the RSAP and RCC10 antibodies (Table 1), respectively, thus indicating that the four cell cultures originated mainly from bronchiolar epithelial Clara cells. Interestingly, throughout the year of culture, the global size of the cells progressively decreased, mainly due to a decrease in the nucleus/cytoplasm ratio, but cells retained morphological characteristics of Clara cells such as nonciliated cells, secretory granules, and surfactant production, and, at passage 50, specific Clara cell immunostaining was still positive (Table 1). These data clearly indicated that after long-term in vitro culture, the four cell lines retained at least a few characteristics of differentiated epithelial Clara cells.

Growth, cell cycle analysis, and genetic instability To characterise these cell lines further, growth properties were studied for 1 year of culture. These cell cultures displayed a high proliferative rate, with approximately 200 – 230 population doublings in 50 passages, without significant change in proliferation capacity during the year of culture (Fig. 1). At the earliest passages (roughly from 1 to 4), a transient high mortality was observed in the cultures corresponding to damaged cells and/or cells ill-adapted to the in vitro culture conditions, but thereafter mortality was no longer detectable. So, under our culture conditions, we observed neither a typical period of senescence, characterised by a slow growth rate, nor a period of cell crisis corresponding to marked mortality. Fig. 2 illustrates the cell cycle and changes in ploidy as

F. Petitot et al. / Experimental Cell Research 286 (2003) 30 –39

35

a function of the passage number. Initially, at passage 0 (p0), cells from fresh explants were mainly in G0/G1 phase (98%), but after in vitro expansion, cell cultures were highly proliferative since at least 30% of the cells were in S or in G2/M phases. A genetic instability, characterised by the appearance of a tetraploid population, occurred around the 20 –25th passages, but with slight differences between the four cell lines. For three of them, R1.1, R1.2, and R2.2, the diploid cell population progressively disappeared and after a few passages only pseudo-tetraploid and higher than tetraploid cells remained in the cultures (Fig. 2). For the R2.1 culture, tetraploidy appeared at the 20th passage, but it was transitory since after the 30th passage we only observed pseudo-diploid cells within the culture. As depicted above, the trend in ploidy, which is characteristic of genetic instability, was not correlated with cell viability or with growth rate. Tumorigenicity in nude mice In spite of an absence of evident senescence and crisis, we observed genetic instability and cellular immortalisation during successive passages. Since these modifications are classically depicted as the first events of cell transformation, we analysed whether these cells had acquired tumorigenic properties. We performed subcutaneous injections of 5 ⫻ 106 cells of the four cell lines at the 50th passage (p50) in athymic nude mice. After a period of 3 months, none of the injected mice had developed tumors. To check if the absence of tumor development could be related to the in vitro low serum concentration medium, we transferred R2.1 cells at p50 in the same medium but supplemented with 10% foetal calf serum. Cells cultivated for 5 additional passages under these new conditions appeared homogeneous and their morphology, proliferation rate, and ploidy were unchanged compared to those of the cells analysed at p50. After this short adaptation to higher serum concentration, similar injections were performed in athymic nude mice and four of five injections gave rise to tumors (⬎8 mm in size) (R2.1–T1 to R2.1–T4) within 3 weeks. Histopathological diagnosis indicated that all tumors were squamous cell (epidermoid) carcinomas and these tumors stained positively with the specific Clara cell antibody.

Fig. 3. Real-time quantitative reverse transcription–polymerase chain reaction of p16INK4a, p19ARF, and p21Cip1 during cellular aging. For the four cell lines, mRNA gene expression was analysed every 10 passages. For each sample, measures of gene expression were normalised as a function of the expression of the 18S ribosomal RNA.

p16INK4a displays an altered expression. In the four cell lines, p16INK4a mRNA is down-expressed after 10 passages and becomes undetectable at the 23rd passage. Tp53 mRNA is also constantly expressed in the four cell lines and the direct sequencing of its coding cDNA sequence, at the 50th passage, indicated a wild-type sequence in all cases (data not shown). Tumorigenic cells (tumors that have been developed on nude mice) do not express 16INK4a and retain the same level of expression for the other genes, as at the 45th passage just before xenografting. Telomerase activity

Expression of tumor suppressor genes during cellular aging Different studies showed that an increased level of negative cell cycle regulators, implicated in the p16INK4a/pRb and p19ARF/p53 pathways, appear to be involved in triggering a premature growth arrest in rodent cells. We investigated whether the absence of senescence could be due to the alteration of expression of the cyclin-dependent kinase inhibitors p16INK4a, p19ARF, and p21Cip1 and/or an alteration of the p53 status. As shown in Fig. 3, p19ARF and p21Cip1 are constantly expressed during cellular aging whereas

To determine if changes of telomerase activity occurred during the long-term culture, we measured the change in telomerase enzymatic activity by TRAP assay at passages ranging from 0 to 50 (Fig. 4). In freshly isolated cells (p0), telomerase activity was low but significant, and was very similar in the four cultures. In each cell culture, the initial level of activity was fixed at 1 as a reference for the serial analyses. In the four cultures, we first observed a significant and transitory increase in telomerase activity (up to 4.5fold) during the first 5 passages. Thereafter, the activity progressively decreased 2.5- to 5-fold, depending on the cell

36

F. Petitot et al. / Experimental Cell Research 286 (2003) 30 –39

Fig. 4. Change in telomerase activity of the four cell cultures (R1.1, R1.2, R2.1, and R2.2) as a function of the passage number. Telomerase activity was determined by the TRAP assay, as described in Materials and Methods. For each cell culture, the initial telomerase activity, determined in freshly isolated cells (passage 0), was fixed at 1 as the reference for the serial analyses. Each cell pellet (106 cells) was used for assay of both protein and telomerase activity.

lines, by the 25th passage. Subsequently, telomerase activity did not change significantly and remained low up to the 50th passage in all cases. The change in telomerase activity was also analysed in R2.1 cells seeded at the 50th passage and cultivated for 5 additional passages (passage 55) in medium supplemented with 10% foetal calf serum as well as in the four derived tumors that developed on nude mice (Fig. 5). Telomerase activity was not modified compared with that detected at p50, either in cells cultivated for 5 passages in “normal” or higher serum concentration nor in tumors grown on athymic nude mice.

Fig. 5. Change in telomerase activity of R2.1 cells seeded at passage 50 for 5 additional passages either in “normal” serum concentration (p55) or in 10% foetal calf serum (p55 10% FCS). Telomerase activity was also measured in the four tumors (R2.1–T1 to R2.1–T4) that developed after injection of R2.1 (p55 10% FCS) cells in athymic nude mice. Telomerase activity was determined by the TRAP assay as described in Materials and Methods. The initial telomerase activity, determined in freshly isolated R2.1 cells (p0), was fixed at 1 as the reference for the serial analyses. Each cell pellet (106 cells) or tumor aliquot was used for assay of both protein and telomerase activity. Results correspond to the mean of at least two independent measurements and error bars correspond to standard deviation of the mean.

around one-third and was highly significant (P ⬍ 10⫺3, nonparametric Mann-Whitney test), the mean values being 1.85 and 1.11 for p0 and p50, respectively, which corresponds to a large decrease in mean telomere length. Interestingly, telomere length continued to decrease after transplantation on nude mice, since signal intensity decreased 2-fold in developed tumors, and this decrease was not related to the short adaptation to 5 passages in 10% foetal calf

Telomere length To ascertain the effect of telomerase activity on telomere length, we measured in the four rat lung-derived cultures the change in mean telomere length by using the Q-FISHFCM method. In the four cultures, telomere length was estimated at passages 10 and 50 and was expressed in relative units compared to the value obtained for R2.1 p50, which was fixed at 1 (Fig. 6). The four freshly isolated lung cell cultures were included in the analysis as p0 references. In all cell lines, at p10 the fluorescence signal intensity was equivalent to or slightly lower than that observed at p0: 1.72 and 1.85 (P ⫽ 0.04, nonparametric Mann-Whitney test), respectively. The results obtained at p50 showed, whatever the cell line, a clear decrease in signal intensity and thus in mean telomere length (Fig. 6). By calculating the mean value for the four cell lines, the decrease in signal intensity was

Fig. 6. Change in mean telomere length of the four cell cultures (R1.1, R1.2, R2.1, and R2.2). For each cell culture, telomere length was measured in freshly isolated cells at passage 0 (p0) and at passages 10 and 50 (p10 and p50). Telomere length was measured in R2.1 cells cultivated for 5 additional passages at higher serum concentration (R2.1 p55 10% FCS) as well as in tumors that developed after injection of these cells in athymic nude mice (R2.1–T1 and R2.1–T2). For comparison, we also included the A549 human lung carcinoma cell line. Results are expressed in relative units, compared to the value obtained for R2.1 at p50, which are fixed at 1.

F. Petitot et al. / Experimental Cell Research 286 (2003) 30 –39

serum performed before cell injection in nude mice (Fig. 6). For comparison, we also included the A549 human lung carcinoma cell line and it should be noted that after the year of culture the mean telomerase length in rat cells remained higher than that of this cell line.

Discussion To study in vitro aging of rat lung epithelial cells, we have developed culture conditions adapted to rat fresh pulmonary explants and to the derived adherent cell lines. Four cell lines, derived from four independent fresh explants, were cultivated for 50 passages during 1 year. Morphological examination and immunocytochemistry indicated that, from the earlier passages to the 50th passage, in vitro adherent cultivated cells retained specific characteristics of bronchiolar Clara cells. Moreover, squamous carcinomas, which developed after injection of one of the cell lines in nude mice, exhibited similar Clara cell differentiation markers. The histogenesis of spontaneous and carcinogen-induced pulmonary neoplasms of rodents is controversial, Clara cells [27–31] and/or alveolar type II pneumocytes [25,32,33] each being a potential target of carcinogenesis. Consequently, sequential analysis of these cell lines provides a suitable system for studying in vitro the regulation and the role of telomerase and telomere length in cellular aging, transformation, and carcinogenesis of rat lung epithelial cells. Here, we show that long-term culture of rat lung epithelial cells is possible under appropriate culture conditions, with no apparent period of senescence. Expression of telomerase activity is not necessary for either in vitro longterm culture or tumor development in athymic mice, since its level steadily decreased up to passage 25 and then remained constant and low. At the earliest passages (0 to 5), telomerase activity increased transitorily but significantly. Telomerase activity could be associated with cell proliferation in normal cells [9,34,35], and proliferation and cell cycle data (Figs. 1 and 2) correlated well with this transitory increase in telomerase activity. At passage 0, more than 99% of cells were in G0/G1 phase of the cell cycle, while soon after explantation cells entered the cell cycle, as indicated by the proliferation rate (Fig. 1) and by the high percentage of cells in S and G2/M phases (Fig. 2). So, this early and transient increase in telomerase activity appeared mainly related to cell proliferation. Inversely, the subsequent downregulation observed at passages 5 through 25, and the low and constant activity detected up to the 50th passage, were unrelated to the cell cycle since, during the year of culture, the proliferative rate of the four cell lines remained high (Fig. 1). Telomerase was therefore repressed during in vitro aging, and reexpression was not observed in cancer cells since its activity remained at the same low level in squamous carcinomas that developed after injection of one of the cell lines in nude mice. As previously shown

37

[8,11–13], initial telomerase activity (passage 0) was lower (10- to 100-fold lower) than in a variety of immortal cell lines. As telomerase activity in the four cell cultures was lower after passage 25 than the initial value, we can classify these cells as telomerase-negative. This does not mean a complete absence of activity but rather, as discussed below, an activity that does not permit maintenance of telomere length. Although an upregulation of telomerase is generally characteristic of cellular immortalisation and is associated with tumorigenesis, under our conditions telomerase activation was not necessary for long-term culture or for the acquisition of tumorigenic potential by rat lung epithelial cells. This tumorigenic potential seems unrelated to a cell selection during the short adaptation period because morphology, proliferation rate, ploidy, and telomerase activity remained unchanged. Even though we observed a senescent state and/or crisis phase during the 1-year culture of rat lung cells, a period of chromosomal instability occurred, characterised by a tetraploid cell population, leading to the appearance of “pseudo-transformed” cells with abnormal karyotypes and acquisition of tumorigenic potential. Recent data suggest that if the telomere-independent senescent state is bypassed, normal rodent cells may proliferate indefinitely without genetic alterations [36]. This hypothesis was illustrated by two studies showing that, under suitable culture conditions, rat oligodendrocyte precursor cells and rat Schwann cells are able to divide indefinitely while remaining diploid and retaining cell cycle check points [21,22]. Moreover, Tang et al. [21] showed that rodent cells may have unlimited proliferative capacity if cultured in conditions that avoid both differentiation and the activation of check point responses that arrest the cell cycle. In these models, cells retained a high level of telomerase activity, in contrast to our results, since, without ever detecting senescence, we observed transient genetic instability, acquisition of tumorigenic potential, and downregulation of telomerase activity, together with characteristics of bronchiolar Clara cells and an unlimited proliferative capacity. Our results are very similar to those obtained with mTR⫺/⫺ mouse cells, in which chromosomal abnormalities linked to telomere shortening were observed in primary cells derived from aging mice (6th generation) or from cell lines derived from mice of different generations but after a long, continuous in vitro proliferation [20,37,38]. The main difference between the two models is that chromosomal instability appeared in cells derived from mTR⫺/⫺ mice after a higher number of cell divisions than in our rat cell cultures. These differences could be explained by (1) a greater mean telomere length in mouse compared to rat, (2) less intracellular heterogeneity of telomere length in mouse compared to rat, shorter telomeres leading to premature genetic instability [39], and (3) a decrease in the amount of hTERT protein, according to the observed decrease in telomerase activity and/or modification of the composition of the set of proteins interacting with telomeric DNA that may destabilise telomere capping

38

F. Petitot et al. / Experimental Cell Research 286 (2003) 30 –39

and thus lead to end-to-end fusions and chromosomal instability. In line with this transgenic model, we observed that continuous growth of rodent cells presenting a downregulation of telomerase activity could be a potent inducer of chromosomal instability. Recently, a correlation between high telomerase activity and the stabilities of genome and DNA ploidy was also described in renal cell carcinoma [40]. And Hande et al. [37] observed that a chromosome-specific stabilisation of telomere length could occur at various points during mouse mTR⫺/⫺ cell growth. This indicates that, even in the absence of telomerase, a chromosome-specific telomere maintenance could take place in rodent cells, thus permitting indefinite growth. In theory, the observed continuous telomere attrition could be prolonged since a lower signal is obtained for the human lung cancer cell line (A549) with a known TRF length of 5–7 kb (Fig. 6). However, an alternative recombination mechanism may contribute to prevention of senescence by stabilising shorter telomeres. In any case, although these cells are finally highly tumorigenic, since squamous cell carcinomas developed in nude mice within 3 weeks, we can hypothesise that these cells may further proliferate before the mean telomere length shortens to the critical length. At this point, genetic adaptation of these cells will be essential for induction of telomerase or alternative mechanisms and leads to stabilisation of the telomere and acquisition of a fully immortalised phenotype. In our culture conditions, the four cell lines do not undergo typical replicative senescence or crisis. Because different studies indicate that alteration of the p16INK4a/pRB pathway alone is sufficient for immortalisation of embryonic stem cells [18,19], the loss of p16INK4a expression in these cell lines could in itself explain the bypass of growth arrest. The ARF/p53 pathway does not seem to be implicated in the process since the p53 status of the four cell lines remained wild-type and the expression of p19ARF and p21Cip1 remained constant. Moreover, cell cycle check point responses were conserved, because after gamma-irradiation we observed both G1 and G2 blockages (data not shown). So, alteration of the p16INK4a/pRB pathway seems to be sufficient to escape senescence, thus leading to the acquisition of a partially transformed phenotype with a transient period of genetic instability and the acquisition of tumoral potential. It is important that during long-term culture, rat lung epithelial cells retained a few characteristics of untransformed cells such as (1) absence of senescence and/or crisis, (2) persistence of cell cycle check points, (3) downregulation of telomerase activity and continuous decrease in telomere length, and (4) persistence of a differentiated phenotype. Although, the continuous decrease in telomere length (even in xenografted tumors) indicated that these cells are not fully immortalised, these characteristics are not incompatible with in vitro transformation and acquisition of tumorigenic potential. Finally, it would be interesting to understand why rodent

cells expressing telomerase in vivo either maintain (nerve cells) or do not maintain (Clara or hepatic [23] cells) high telomerase activity in vitro. Since the state of differentiation could be of major importance for the modulation of telomerase activity, sequential analysis of long-term rat epithelial Clara cell cultures may provide a unique model to investigate the effect of culture components on telomerase activity with regard to cellular immortalisation and tumorigenesis. Acknowledgments This study was supported by grants from COGEMA (PICD14 and its continuity), EU F14P-CT95-0008 and F14P-CT95-0025, and EDF. F. Petitot was supported by a fellowship from COGEMA. References [1] C.J. Sherr, R.A. DePinho, Cellular senescence: mitotic clock or culture shock?, Cell 102 (2000) 407– 410. [2] E.H. Blackburn, Telomere states and cell fates, Nature 408 (2000) 53–56. [3] T. de Lange, T. Jacks, For better or worse? Telomerase inhibition and cancer, Cell 98 (1999) 273–275. [4] S.E. Artandi, R.A. DePinho, A critical role for telomeres in suppressing and facilitating carcinogenesis, Curr. Opin. Genet. Dev. 10 (2000) 39 – 46. [5] R.K. Moyzis, J.M. Buckingham, L.S. Cram, M. Dani, L.L. Deaven, M.D. Jones, J. Meyne, R.L. Ratliff, J.R. Wu, A highly conserved repetitive DNA sequence, (TTAGGG)n, present at the telomeres of human chromosomes, Proc. Natl. Acad. Sci. USA 85 (1988) 6622– 6626. [6] D. Kipling, H.J. Cooke, Hypervariable ultra-long telomeres in mice, Nature 347 (1990) 400 – 402. [7] V.L. Makarov, S. Lejnine, J. Bedoyan, J.P. Langmore, Nucleosomal organization of telomere-specific chromatin in rat, Cell 73 (1993) 775–787. [8] C. Chadeneau, P. Siegel, C.B. Harley, W.J. Muller, S. Bacchetti, Telomerase activity in normal and malignant murine tissues, Oncogene 11 (1995) 893– 898. [9] R.A. Greenberg, R.C. Allsopp, L. Chin, G.B. Morin, R.A. DePinho, Expression of mouse telomerase reverse transcriptase during development, differentiation and proliferation, Oncogene 16 (1998) 1723– 1730. [10] L. Martin-Rivera, E. Herrera, J.P. Albar, M.A. Blasco, Expression of mouse telomerase catalytic subunit in embryos and adult tissues, Proc. Natl. Acad. Sci. USA 95 (1998) 10471–10476. [11] M.A. Blasco, M. Rizen, C.W. Greider, D. Hanahan, Differential regulation of telomerase activity and telomerase RNA during multistage tumorigenesis, Nat. Genet. 12 (1996) 200 –204. [12] D. Broccoli, L.A. Godley, L.A. Donehower, H.E. Varmus, T. de Lange, Telomerase activation in mouse mammary tumors: lack of detectable telomere shortening and evidence for regulation of telomerase RNA with cell proliferation, Mol. Cell. Biol. 16 (1996) 3765– 3772. [13] A.M. Burger, M.C. Bibby, J.A. Double, Telomerase activity in normal and malignant mammalian tissues: feasibility of telomerase as a target for cancer chemotherapy, Br. J. Cancer 75 (1997) 516 –522. [14] T. Kiyono, S.A. Foster, J.I. Koop, J.K. McDougall, D.A. Galloway, A.J. Klingelhutz, Both Rb/p16INK4a inactivation and telomerase activity are required to immortalize human epithelial cells, Nature 396 (1998) 84 – 88.

F. Petitot et al. / Experimental Cell Research 286 (2003) 30 –39 [15] S.E. Artandi, R.A. DePinho, Mice without telomerase: what can they teach us about human cancer?, Nat. Med. 6 (2000) 852– 855. [16] W.E. Wright, J.W. Shay, Telomere dynamics in cancer progression and prevention: fundamental differences in human and mouse telomere biology, Nat. Med. 6 (2000) 849 – 851. [17] R.D. Ramirez, C.P. Morales, B.S. Herbert, J.M. Rohde, C. Passons, J.W. Shay, W.E. Wright, Putative telomere-independent mechanisms of replicative aging reflect inadequate growth conditions, Genes Dev. 15 (2001) 398 – 403. [18] J. Sage, G.J. Mulligan, L.D. Attardi, A. Miller, S. Chen, B. Williams, E. Theodorou, T. Jacks, Targeted disruption of the three Rb-related genes leads to loss of G(1) control and immortalization, Genes Dev. 14 (2000) 3037–3050. [19] J.H. Dannenberg, A. van Rossum, L. Schuijff, H. te Riele, Ablation of the retinoblastoma gene family deregulates G(1) control causing immortalization and increased cell turnover under growth-restricting conditions, Genes Dev. 14 (2000) 3051–3064. [20] M.A. Blasco, H. W. Lee, M. P. Hande, E. Samper, P.M. Lansdorp, R.A. DePinho, C.W. Greider, Telomere shortening and tumor formation by mouse cells lacking telomerase RNA, Cell 91 (1997) 25–34. [21] D.G. Tang, Y.M. Tokumoto, J.A. Apperly, A.C. Lloyd, M.C. Raff, Lack of replicative senescence in cultured rat oligodendrocyte precursor cells, Science 291 (2001) 868 – 871. [22] N.F. Mathon, D.S. Malcolm, M.C. Harrisingh, L. Cheng, A.C. Lloyd, Lack of replicative senescence in normal rodent glia, Science 291 (2001) 872– 875. [23] V.M. Golubovskaya, S.C. Presnell, M.J. Hooth, G.J. Smith, W.K. Kaufmann, Expression of telomerase in normal and malignant rat hepatic epithelia, Oncogene 15 (1997) 1233–1240. [24] V.M. Golubovskaya, L.V. Filatov, C.I. Behe, S.C. Presnell, M.J. Hooth, G.J. Smith, W.K. Kaufmann, Telomere shortening, telomerase expression, and chromosome instability in rat hepatic epithelial stemlike cells, Mol. Carcinog. 24 (1999) 209 –217. [25] R.A. Herbert, B.S. Stegelmeier, N.A. Gillett, A.H. Rebar, W.W. Carlton, G. Singh, F.F. Hahn, Plutonium-induced proliferative lesions and pulmonary epithelial neoplasms in the rat: immunohistochemical and ultrastructural evidence for their origin from type II pneumocytes, Vet. Pathol. 31 (1994) 366 –374. [26] M. Hultdin, E. Gronlund, K. Norrback, E. Eriksson-Lindstrom, T. Just, G. Roos, Telomere analysis by fluorescence in situ hybridization and flow cytometry, Nucleic Acids Res. 26 (1998) 3651–3656. [27] A.R. Kennedy, R.B. McGandy, J.B. Little, Histochemical, light and electron microscopic study of polonium-210 induced peripheral tu-

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35] [36] [37]

[38]

[39]

[40]

39

mors in hamster lungs: evidence implicating the Clara cell as the cell of origin, Eur. J. Cancer 13 (1977) 1325–1340. S.L. Kauffman, L. Alexander, L. Sass, Histologic and ultrastructural features of the clara cell adenoma of the mouse lung, Lab. Invest. 40 (1979) 708 –716. C.L. Sanders, K.E. McDonald, K.E. Lauhala, Promotion of pulmonary carcinogenesis by plutonium particle aggregation following inhalation of 239PuO2, Radiat. Res. 116 (1988) 393– 405. S. Rehm, M. Takahashi, J.M. Ward, G. Singh, S.L. Katyal, J.R. Henneman, Immunohistochemical demonstration of Clara cell antigen in lung tumors of bronchiolar origin induced by N-nitrosodiethylamine in Syrian golden hamsters, Am. J. Pathol. 134 (1989) 79 – 87. S. Rehm, W. Lijinsky, G. Singh, S.L. Katyal, Mouse bronchiolar cell carcinogenesis. Histologic characterization and expression of Clara cell antigen in lesions induced by N-nitrosobis-(2-chloroethyl) ureas, Am. J. Pathol. 139 (1991) 413– 422. M. Ohshima, J.M. Ward, G. Singh, S.L. Katyal, Immunocytochemical and morphological evidence for the origin of N-nitrosomethylurea-induced and naturally occurring primary lung tumors in F344/ NCr rats, Cancer Res. 45 (1985) 2785–2792. S. Rehm, J.M. Ward, A.A. ten Have-Opbroek, L.M. Anderson, G. Singh, S.L. Katyal, J.M. Rice, Mouse papillary lung tumors transplacentally induced by N-nitrosoethylurea: evidence for alveolar type II cell origin by comparative light microscopic, ultrastructural, and immunohistochemical studies, Cancer Res. 48 (1988) 148 –160. C.D. Belair, T.R. Yeager, P.M. Lopez, C.A. Reznikoff, Telomerase activity: a biomarker of cell proliferation, not malignant transformation, Proc. Natl. Acad. Sci. USA 94 (1997) 13677–13682. S. Kyo, M. Takakura, T. Kohama, M. Inoue, Telomerase activity in human endometrium, Cancer Res. 57 (1997) 610 – 614. J.W. Shay, W.E. Wright, Aging. When do telomeres matter?, Science 291 (2001) 839 – 840. M.P. Hande, E. Samper, P. Lansdorp, M.A. Blasco, Telomere length dynamics and chromosomal instability in cells derived from telomerase null mice, J. Cell Biol. 144 (1999) 589 – 601. K.L. Rudolph, S. Chang, H.W. Lee, M. Blasco, G.J. Gottlieb, C. Greider, R.A. DePinho, Longevity, stress response, and cancer in aging telomerase-deficient mice, Cell 96 (1999) 701–712. M.T. Hemann, M.A. Strong, L.Y. Hao, C.W. Greider, The shortest telomere, not average telomere length, is critical for cell viability and chromosome stability, Cell 107 (2001) 67–77. H. Izumi, T. Hara, A. Oga, K. Matsuda, Y. Sato, K. Naito, K. Sasaki, High telomerase activity correlates with the stabilities of genome and DNA ploidy in renal cell carcinoma, Neoplasia 4 (2002) 103–111.