Amplification of Telomerase Reverse Transcriptase Gene in Human Mammary Epithelial Cells with Limiting Telomerase RNA Expression Levels

Research Article

Ying Cao, Lily I. Huschtscha, Amanda S. Nouwens, Hilda A. Pickett, Axel A. Neumann, Andy C-M. Chang, Christian D. Toouli, Tracy M. Bryan, and Roger R. Reddel

Children’s Medical Research Institute, Westmead, and University of Sydney, Sydney, New South Wales, Australia

Abstract any of these components may result in dyskeratosis congenita, a Activation of telomerase is a crucial step during cellular human disease syndrome associated with short telomeres (reviewed by Kirwan et al. in ref. 7). immortalization, and in some tumors this results from amplification of the human telomerase reverse transcriptase Telomerase is not detectable in most normal human somatic (hTERT) gene. Immortalization of normal human cells has cells, whereas it is expressed in >85% of human tumors (8). How been achieved by transduction with hTERT cDNA under the telomerase levels are regulated in these tumors is still poorly control of a strong heterologous enhancer/promoter, but this understood. A major focus of investigation has been the control is sometimes an inefficient process, with periods of poor of hTERT transcription (reviewed by Horikawa et al. in ref. 9). growth or even crisis occurring before immortalization. Here, Amplification of the hTERT gene has been observed in various we showed that normal human mammary epithelial cells human cancer cell lines and tumors, and is also thought to be a expressing exogenous hTERT amplified the transgene exten- mechanism of telomerase activation (10–13). sively and expressed high levels of hTERT mRNA and protein. In view of the observations that hTR is ubiquitously expressed (6, 14), whereas hTERT is expressed only in telomerase-positive Paradoxically, the cells had low levels of telomerase activity cells (5), and that expression of exogenous hTERT alone can and very short telomeres, indicating that telomerase activity immortalize normal human cells (15, 16), abundance of hTERT was did not correlate with hTERT expression. These cells con- previously thought to be the sole limiting factor for telomerase tained only f20 human telomerase RNA (hTR) molecules/cell activity. Although immortalization of many types of normal human (compared with f120 hTR molecules per 293 cell). Expression cells has been achieved by transducing them with exogenous of exogenous hTR caused increased telomerase activity and hTERT, for some strains of normal human cells, this is an ineffi- telomere lengthening. These data indicate that some hTERT- cient process, with periods of poor growth or even crisis being transduced normal cells may express high levels of the experienced before immortalization (17–19). Data are accumulat- transgene but fail to up-regulate endogenous hTR expression ing in support of the notion that hTR levels can also be limiting for sufficiently to enable expression of robust levels of telomerase telomerase activity. For example, hTR mutations have been found activity. [Cancer Res 2008;68(9):3115–23] to cause autosomal dominant dyskeratosis congenita (20) and can reduce telomerase activity via haploinsufficiency (21). Furthermore, Introduction overexpression of both hTERT and hTR elongates telomeres better Human telomeres are nucleoprotein complexes located at the than overexpression of hTERT alone in X-linked and autosomal ends of linear chromosomes (1). They shorten at each cell division, dominant dyskeratosis congenita (22, 23). It has also been shown which limits the number of times a cell can divide (2). Telomerase that concomitant overexpression of hTERT and hTR was necessary is a ribonucleoprotein enzyme complex that adds telomeric repeat to substantially increase telomerase activity in both cancer and sequences to the ends of chromosomes to counteract this primary cells (24). However, none of these studies quantified the shortening (3). It has recently been found that the active human number of hTR molecules in the cells. This is important because a telomerase enzyme is composed of human telomerase reverse previous study has reported that there are 11,000 to 13,500 hTR transcriptase (hTERT), human telomerase RNA (hTR), and dyskerin molecules per cell in the telomerase-negative BJ, IMR90, and SW13 (4): hTERT is the catalytic reverse transcriptase component (5), cell strains (25), which would be inconsistent with a limiting role hTR serves as the RNA template for the addition of telomeric for hTR in light of the 20 to 50 active telomerase molecules per cell repeats (6), and dyskerin is an RNA-binding protein. Mutations in in 293 cells (4). In this study, we explore mechanisms of telomerase activity regulation. Human mammary epithelial cells (HMEC) have been transduced with hTERT in our laboratory, and four independent Note: Supplementary data for this article are available at Cancer Research Online hTERT-immortalized mass cultures (B80-TERT1, 2, 3a, and 3b) (http://cancerres.aacrjournals.org/). were obtained (26). In the present study, we found that the B80- Current address for A.S. Nouwens: School of Molecular and Microbial Sciences, and Australian Institute of Biotechnology and Nanotechnology, The University of TERT1 cell line had undergone extensive amplification of the Queensland, St Lucia, Queensland 4072, Australia; Current address for C.D. Toouli, hTERT transgene and was expressing high levels of hTERT mRNA Bio-Link Partners Ltd., Suite 7/G11, Locomotive Workshop, Australian Technology and protein but had low telomerase activity. Furthermore, we Park, Eveleigh, New South Wales 1430, Australia. Requests for reprints: Roger R. Reddel, Children’s Medical Research Institute, 214 showed that hTR levels limit telomerase activity and telomere Hawkesbury Road, Westmead, Sydney, New South Wales 2145 Australia. Phone: 61-2- length in this cell line. More importantly, our quantitation of hTR 9687-2800; Fax: 61-2-9687-2120; E-mail: [email protected]. I2008 American Association for Cancer Research. molecules per cell provides an explanation for why hTR is limiting: doi:10.1158/0008-5472.CAN-07-6377 the low abundance of hTR (f20 molecules of hTR per cell) is www.aacrjournals.org 3115 Cancer Res 2008; 68: (9). May 1, 2008

Figure 1. hTERT amplification in B80-TERT HMEC lines. A, FISH analysis of B80-TERT1 cells using the full-length hTERT cDNA as a probe. Top, the fluorescein-avidin detection of hTERT (green). Middle, DAPI staining of the nucleus (blue). Bottom, merged picture of hTERT signal and DAPI staining. B, Southern blot analysis of B80, B80-TERT1, 2, 3a, and 3b cells using the full-length hTERT cDNA as a probe. Top, detection of hTERT. Bottom, shows the agarose gel stained with ethidium bromide to indicate DNA loading. Note that only 1 Ag of B80-TERT1 DNA was loaded because of the very high hTERT gene copy number in these cells, whereas 10 Ag of DNA was loaded for the other cell lines. Untransfected B80 cells are the control.

insufficient for significant telomerase activity and telomere fluorescence microscope with appropriate filter sets. Fluorescein and DAPI maintenance even in the presence of a vast excess of hTERT. To images were captured separately with a cooled charge-coupled device our knowledge, our study is the first to report the low abundance camera (SPOT2; Diagnostic Instruments), merged using SPOT2 software, of hTR in telomerase-negative cells (approximately three hTR and further processed using Adobe Photoshop version 6.0 software. Southern blot analysis. Genomic DNA was digested with EcoR I and molecules in each telomerase-negative B80 cell), and that hTR Sal I, separated through a 0.8% agarose gel with Tris-borate-EDTA (TBE) limits telomerase activity and telomere maintenance in the context buffer at pH 7.5, and transferred onto Hybond N+ membrane (Amersham of hTERT amplification. Biosciences) by capillary action in 0.4 mol/L NaOH for 4 h. The membrane was blocked in Church buffer [1% bovine serum albumin, 1 mmol/L EDTA,

0.5 mol/L NaPO4 (pH 7.2), and 7% SDS] at 65jC for 1 h, followed by Materials and Methods hybridization (in Church buffer at 65jC overnight) with the 3.4-kb full- Cells and cell culture. HMECs were cultured in MCDB 170 medium length hTERT cDNA excised from pCIneo-hTERT plasmid and radiolabeled (Invitrogen) as described previously (26). PA317 packaging cells were grown with a-32P-dCTP. The membrane was washed twice in 2ÂSSC/0.1%SDS at in Dulbecco’s Modified Eagles Medium with 10% fetal bovine serum in a 5% 65jC for 20 min, followed by another two washes in 0.2ÂSSC/0.1%SDS j j CO2 incubator at 37 C. at 65 C for 20 min. Washed membrane was exposed to Kodak BioMax MS Fluorescence in situ hybridization analysis. Chromosome prepara- X-ray film (Sigma-Aldrich). tions from colcemid-arrested cells were obtained according to standard Reverse transcriptase-PCR. Total RNA was isolated from cell pellets by cytogenetic methods (27). A 3.4-kb full-length hTERT cDNA excised from using TRI-zol reagent (Invitrogen). One microgram of total RNA was reverse pCIneo-hTERT plasmid (28) or a 650-bp NH2-terminal hTERT fragment transcribed by using Moloney murine leukemia virus reverse transcriptase excised from pJJ101-hTERT1-182aa was labeled with bio-16-dUTP using the (Promega). Taq polymerase (Roche) was used for the amplification. See Biotin-Nick Translation Mix (Roche) according to the manufacturer’s Supplementary Data for PCR conditions and primer sequences. instructions. Approximately 30 ng/AL of the full-length hTERT or the 650-bp Quantitative real-time reverse transcriptase-PCR. The Qiagen RNeasy

NH2-terminal hTERT fragment was hybridized onto separately denatured Mini kit (Qiagen) was used to extract total RNA from cells that had been chromosome preparations for 16 to 18 h in a humidified chamber at 37jC. pelleted by centrifugation. Total RNA (1 Ag) was treated with DNase I The hybridized probe was detected with Fluorescein Avidin DCS (Vector (Invitrogen), followed by reverse transcription [SuperScript III First-Strand Laboratories) followed by signal amplification with biotinylated anti-avidin Synthesis System for reverse transcriptase-PCR (RT-PCR); Invitrogen] antibody (Vector Laboratories) and another layer of Fluorescein Avidin DCS. according to the manufacturer’s instructions. Real-time RT-PCR was Chromosomes were counterstained with diamidino-phenyl-indole-dihydro- performed using SYBR Green PCR Master Mix (Applied Biosystems) in a chloride (DAPI; final concentration, 0.6 Ag/mL; Sigma-Aldrich) for Corbett RotorGene-6000 thermal cycler. Cycle parameters were as follows: chromosome identification, and slides were evaluated on a Leica DMLB incubation at 95jC for 10 min, followed by 40 cycles of 95jC for 15 sec and

Cancer Res 2008; 68: (9). May 1, 2008 3116 www.aacrjournals.org

60jC for 1 min. Housekeeping gene glyceraldehyde-3-phosphate dehydroge- See Supplementary Data for Preparation of hTR standards and in vitro nase (GAPDH) was used as the reference gene. Primer pairs were matched hTERT translation. for PCR efficiencies. See Supplementary Data for primer sequences. Â 6 Â 7 Northern blot analysis. Total RNA from 5 10 or 1 10 cells was Results prepared using an RNeasy Mini kit (Qiagen), TRI-zol reagent (Invitrogen), or phenol/chloroform extraction from whole cell lysate to compare potential Amplification of exogenous hTERT in HMECs. Four indepen- yield differences. Total RNA extracted by any of the above methods was dent hTERT-immortalized mass cultures B80-TERT1, 2, 3a, and 3b precipitated, resuspended in 10 AL of water and 10 AL of formamide loading were previously obtained in our laboratory by transfecting B80 buffer, followed by heating at 70jC for 5 min, loaded on a 1.5-mm thick HMECs with pCIneo-hTERT (26). In the present study, we detected 4% polyacrylamide, 8M urea gel, and electrophoresed at 25 W for 1 h in extensive amplification of hTERT DNA in B80-TERT1 cells by Â 1 TBE buffer. RNA was transferred onto Hybond N+ membrane fluorescence in situ hybridization (FISH) analysis of metaphase (Amersham Biosciences) by electroblotting at 1.5A for 2 h in 0.5Â TBE buffer, with the buffer cooled to 4jC. The membrane was crosslinked with 254 nm UV light (Stratalinker; Stratagene) and prehybridized in Church buffer (see above) at 55jC for 1 h, followed by hybridization in Church buffer at 55jC overnight with 32P5¶-end–labeled hTR probe (5¶-CGG TGG AAG GCG GCA GGC CGA GGC-3¶). The membrane was washed three times in 0.1ÂSSC/0.1%SDS at 55jC for 10 min, exposed to a phosphor screen, and scanned with a Storm 860 optical scanner with ImageQuant software (Molecular Dynamics). Western blot analysis. Cell pellets were lysed in freshly made radioim- munoprecipitation assay buffer [10 mmol/L Tris-HCl (pH8.0), 140 mmol/L NaCl, 1% Triton X-100, 0.1% SDS, 1Â complete protease inhibitors (Roche), 10 mmol/L phenylmethylsulfonylfluoride, and 0.1% sodium deoxycholate]. Lysates equivalent to 50 Ag of protein were separated on an 8% polyacrylamide/SDS gel and transferred to Hybond enhanced chemiluminescence (ECL) nitrocellulose (Amersham Biosciences) by electroblotting. The membrane was blocked with 1Â PBS containing 2% skim milk and 0.05% Tween 20 at room temperature for 1 h. The membrane was then incubated with the primary antibody (goat polyclonal anti-hTERT; sc-7215; 1:1,000 dilution; Santa Cruz) in 1ÂPBS containing 1% skim milk and 0.05% Tween 20 at 4jC overnight. The membrane was washed in 1ÂPBS containing 0.05% Tween 20, followed by the incubation with the secondary antibody (rabbit anti-goat immunoglobulin horseradish perox- idase; 1:10,000 dilution; DAKO Cytomation) in 1ÂPBS containing 1% skim milk and 0.05% Tween 20 at room temperature for 1 h. Amersham ECL plus Western blotting detection system (Amersham Biosciences) and high performance chemiluminescence film (Amersham Biosciences) were used for detection. Direct telomerase activity assay. A direct telomerase activity assay recently developed in our laboratory (29) was used. Cells (1 Â 107) were lysed and the lysate was incubated with purified anti-hTERT antibody (4). Protein G/Agarose beads were added and the immunoprecipitate was collected into a microspin column. The beads were washed, and antigenic peptide (4) was mixed with the immunoprecipitate to elute telomerase. The activity of immunopurified telomerase was analyzed as described in Supplementary Data. Production of retroviruses and infections. hTR construct pBABE- puro-U3-hTR (kindly provided by Dr. Kathleen Collins, Department of Molecular and Cell Biology, University of California, Berkeley, CA) and vector alone (pBABE-puro) were transfected into PA317 packaging cells separately using siPORT XP-1 transfection reagent (Ambion). After 48 h, medium was harvested and filtered through a 0.45-Am membrane, and polybrene (Sigma-Aldrich) was added to a final concentration of 4 Ag/mL. B80-TERT1 cells were grown to 50% confluence and then incubated with retroviral supernatant at 37jC for 8 h, followed by incubation with fresh medium overnight. Forty-eight hours after infection, cells were passaged and grown in selection medium containing 0.6 Ag/mL puromycin (Sigma- Aldrich). Clones were obtained and maintained in medium containing 0.6 Ag/mL puromycin. Figure 2. hTERT mRNA and protein expression correlates with hTERT gene Terminal restriction fragment analysis. Hinf I- and Rsa I-digested copy number. hTERT mRNA expression was detected by RT-PCR (A) and A quantitative real-time RT-PCR (B; using GAPDH as the reference gene). Error genomic DNA (1.5 g) was fractionated on a pulsed-field electrophoresis bars, ranges for duplicate PCR reactions. hTERT protein expression was apparatus (Bio-Rad) as previously described (30). The gel was dried, detected by Western blot using hTERT antibody sc7215 (C). In vitro translated denatured, and hybridized to a [g-32P] dATP 5¶-end–labeled telomeric oligo- hTERT was used as a positive control in the hTERT Western; it migrates a nucleotide probe, (TTAGGG) . The gel was exposed to a phosphor screen little more slowly in the gel due to the presence of NH2-terminal tags on the 3 protein. hTERT was not detectable in untransfected B80 cells by either RT-PCR and scanned with a Storm 860 optical scanner with ImageQuant software. or Western blot. B80 refers to untransfected B80, and 1, 2, 3a, and 3b refer to Median telomere length was obtained with Telometric software (31). B80-TERT1, B80-TERT2, B80-TERT3a, and B80-TERT3b cells, respectively. www.aacrjournals.org 3117 Cancer Res 2008; 68: (9). May 1, 2008

Figure 3. Telomerase activity measured by a direct telomerase assay correlates with hTR levels. Telomerase reaction products were analyzed on a 10% denaturing sequencing gel (A). Note only 2.5 Â 106 293 cells were assayed, whereas 1 Â 107 cells were used for the others. Results from four independent telomerase assays were quantified with ImageQuant software and are shown in B. Columns, mean; bars, SE. No telomerase activity was detected in B80 cells. hTR (C) and DKC1 mRNA (D) levels were measured by real-time RT-PCR. Columns, mean; bars, SE from three separate PCR runs (C). Columns, mean; bars, ranges for duplicate PCR reactions from the same PCR run (D). B80, 1, 2, 3a, and 3b are defined in Fig. 2 legend.

spreads using both a 3.4-kb full-length hTERT cDNA (Fig. 1A) and a Correlation between hTERT RNA/protein expression and hTERT 650-bp NH2-terminal hTERT probe (data not shown). Amplification gene copy number. Expression of hTERT mRNA was of hTERT was observed in every metaphase (Fig. 1A)and measured by conventional and real-time RT-PCR. As shown in interphase nucleus (Supplementary Fig. S1), which indicates that Fig. 2A, the highest level of hTERT mRNA was detected in B80- this immortalized cell line has become clonal, although it was TERT1 cells, less hTERT mRNA was detected in B80-TERT3b, and originally established as a mass culture. A similar pattern of the least amount of hTERT mRNA was detected in B80-TERT2 and hybridization in B80-TERT1 cells was also observed by FISH B80-TERT3a cells. No hTERT mRNA could be detected in untrans- analysis when pCIneo vector alone was used as a probe (data not fected B80 cells. Similar amounts of GAPDH mRNA were detected shown). These observations indicate that the hTERT transgene and in all five HMEC cultures (Fig. 2A). Quantitation by real-time its accompanying vector sequence were extensively amplified in RT-PCR is shown in Fig. 2B. No hTERT mRNA was detected in B80-TERT1 cells. untransfected B80 cells (data not shown). All hTERT immortalized Furthermore, hTERT amplification was also detected in B80- lines showed a significantly higher level of hTERT mRNA compared TERT3b cells by FISH analysis (Supplementary Fig. S2), although with telomerase-positive 293 cells, with the highest level of hTERT to a much lesser extent than in B80-TERT1 cells. No hTERT ampli- mRNA detected in B80-TERT1 (>30-fold greater than in 293 cells). fication was detectable by FISH in B80-TERT2 or 3a cells (data not To determine whether hTERT protein is being expressed and shown). whether the protein level correlates with the hTERT gene copy To confirm the hTERT amplification seen by FISH in B80-TERT1 number and hTERT mRNA level, Western blot analysis using an and B80-TERT3b cells, Southern blot analysis was performed using antibody against hTERT was performed, with in vitro translated the full-length hTERT cDNA as a probe. A very high hTERT copy hTERT as a positive control (Fig. 2C). No hTERT could be detected number was detected in only 1 Ag of B80-TERT1 genomic DNA, in untransfected B80 cells (data not shown) and the highest hTERT whereas only a small amount of hTERT DNA was detected in 10 Ag protein level was found in B80-TERT1 cells. Therefore, we conclude of genomic DNA from B80-TERT2 and B80-TERT3a. An interme- that hTERT copy number in this panel of HMEC cell lines diate TERT copy number was observed in B80-TERT3b (Fig. 1B). correlates with expression levels of hTERT mRNA and protein. The genomic DNA was digested with EcoR I and Sal I restriction Lowest telomerase activity and lowest hTR in the highest enzymes, which excise the 3.4-kb hTERT cDNA from plasmid hTERT-expressing cells. To determine whether hTERT expression pCIneo-hTERT with which the B80-TERT lines were generated. The correlates with telomerase activity, a direct quantitative telomerase presence of a dominant 3.4-kb band (Fig. 1B, top) showed that it activity assay was performed in all four hTERT-immortalized lines was the exogenous hTERT that was amplified. This was confirmed (Fig. 3A and B). Surprisingly, B80-TERT1 cells had the lowest by the observation that a high copy number of the vector backbone telomerase activity (32% of the activity in B80-TERT3b cells). was detected in B80-TERT1 cells when the pCIneo vector alone was hTR levels have previously been shown to limit telomerase used as a probe in Southern blot analysis (data not shown). activity in hTERT-overexpressing HeLa, HT1080, and normal

Cancer Res 2008; 68: (9). May 1, 2008 3118 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 2008 American Association for Cancer Research. hTERT Amplification in Cells with Limiting hTR Levels human lung fibroblasts (24). To determine whether telomerase addition to any form of precursor, incompletely synthesized and activity is limited by hTR level in hTERT-immortalized B80 cells, even partially degraded hTR, provided the target 150-bp amplicon relative hTR level was examined using real-time RT-PCR. As shown was intact. There are also some technical reasons that may con- in Fig. 3C, the hTERT-transduced B80 cells had increased hTR tribute to the differences between the two methods. For example, expression (15- to 51-fold) compared with the untransfected con- the signal to noise ratio may confound quantitation by Northern trols. The lowest level of hTR was detected in B80-TERT1 (15-fold more than by real-time RT-PCR. In addition, this difference is increase), and the highest level of hTR was detected in B80-TERT3b similar to the 3-fold difference in TLC1 (yeast telomerase RNA) (51-fold). Different primer sets gave similar real-time RT-PCR copy number examined by Northern and real-time RT-PCR analysis results (Supplementary Fig. S3). (32). We consider that the copy number of functional hTR mole- To determine the number of hTR molecules in each cell, we cules is more likely to correspond to the number obtained by performed both Northern blot analysis and real-time RT-PCR using Northern because this technique detects the mature full-length serially diluted in vitro transcribed hTR as standards. Both assays hTR. Regardless of which technique is more accurate, the data showed excellent linearity of the standards over the range used indicate that telomerase activity correlates with hTR, rather than (Fig. 4, bottom). Three Northern blot analyses with 293 cells (using with hTERT expression levels in these hTERT-overexpressing three different methods to extract total RNA to compare potential HMECs. RNA yield difference; see Materials and Methods section) gave very As dyskerin (encoded by the DKC1 gene) is part of the active reproducible results, and the number of hTR molecules in each telomerase enzyme in addition to hTERT and hTR (4), we examined 293 cell was 123 F 23 (mean F SE; data not shown). Northern blot DKC1 mRNA expression to determine whether DKC1 level analysis showed that the number of hTR molecules per cell was 23, correlates with levels of hTERT, hTR, or telomerase activity. As 28, 43, and 90 in B80-TERT1, 2, 3a, and 3b, respectively (Fig. 4A). shown in Fig. 3D, hTERT overexpression resulted in increased A repeat Northern gave a 2-fold lower estimate of hTR molecules DKC1 expression from 2- to 4-fold in the four B80-TERT lines. per cell (data not shown). Three separate real-time RT-PCR However, no consistent correlation was found between expression analyses with duplicates in each PCR run gave a slightly higher of DKC1 mRNA and either hTERT or hTR expression, or telomerase estimate of hTR copy number as follows: 45, 76, 79, and 148 activity. molecules of hTR in B80-TERT1, 2, 3a, and 3b, respectively, and Increased telomerase activity and telomere lengthening f300 molecules of hTR in each 293 cell (Fig. 4B). resulted from hTR overexpression. We next addressed whether The f2-fold difference in hTR molecules per cell quantified by overexpression of hTR would increase telomerase activity in B80- our Northern and real-time RT-PCR could be due to the fact that TERT1 cells. An hTR-containing retroviral vector, pBABE-U3-hTR, Northern blot analysis only detects the mature 451 nucleotide hTR, was used to induce hTR expression, with pBABE vector alone as a whereas real-time RT-PCR would detect the mature hTR in control. B80-TERT1 cells were infected with the retrovirus and

Figure 4. hTR molecules per cell determined by both quantitative Northern blot analysis and real-time RT-PCR. A, a representative Northern blot analysis using in vitro transcribed hTR (i.v. hTR) as standards (top). Total RNA from 5 Â 106 293 cells and 1 Â 107 B80-TERT1, 2, 3a, and 3b cells was loaded on the gel. RC, a recovery control added before RNA precipitation (32P-labeled 100-mer DNA oligonucleotide). About 40% of the RNA was recovered after RNA precipitation (data not shown), which was taken into account when calculating hTR molecules per cell (given beneath the gel) by interpolation from the standard curve shown at bottom. B, hTR ratio (to 293) from a representative real-time RT-PCR analysis (top). The standard curve (bottom) was from the same real-time RT-PCR analysis. hTR molecules per cell calculated from three independent experiments are shown in the right column (mean F SE). Each RT-PCR reaction used 1 Ag total cellular RNA; the number of hTR molecules per cell was adjusted for the proportion of total RNA yield represented by this 1 Ag. Serially diluted in vivo hTR was included to make the standard curve (bottom). To achieve similar reverse transcription and PCR efficiency, in vivo hTR was reverse transcribed together with 1 Ag of VA13 RNA (hTR-negative; ref. 49), followed by PCR amplification.

www.aacrjournals.org 3119 Cancer Res 2008; 68: (9). May 1, 2008

vector control clones, hTR levels varied from 0.8- to 1.9-fold compared with the parental cells (Fig. 5A). The difference in hTR level between hTR-infected and vector control clones was highly significant (P < 0.0001). hTERT expression varied from 1.2- to 4.3-fold in the three vector control clones and from 1- to 6.1-fold in the seven hTR-overexpressing clones compared with parental B80-TERT1 cells (Fig. 5B). DKC1 RNA expression varied from 1.5- to 2.7-fold in the three vector control clones and from 1.7- to 3-fold in the seven hTR-overexpressing clones (Fig. 5C). No significant difference was observed in either hTERT level (P = 0.68) or DKC1 level (P = 0.25) between hTR-infected and vector control clones (Fig. 5). To measure telomerase activity levels, we used a direct telomerase activity assay. Telomerase activity was increased from 33- to 63-fold in hTR-overexpressing clones in comparison with 1.1- to 1.6-fold in vector-alone controls (Fig. 6A). We then determined whether the increased telomerase activity resulting from hTR overexpression in B80-TERT1 cells would cause telomere lengthening. As shown in Fig. 6B, B80-TERT1 cells had very short telomeres (median telomere length, 5 kb), whereas B80-TERT3b cells had very long telomeres (median telomere length, 14 kb), correlating with the telomerase activities in those cells (Fig. 3A). As expected, all hTR-overexpressing clones had extensively elongated telomeres (median telomere length varied from 29–36 kb in 7 hTR-overexpressing clones; Fig. 6B), compared with the vector controls that exhibited short telomeres (median telomere length varied from 4.5–5.2 kb in three vector control clones), similar to the parental B80-TERT1 cells. We therefore conclude that hTR overexpression in B80-TERT1 cells resulted in increased telomerase activity and consequent telomere lengthening.

Discussion hTERT amplification has previously been observed in human cancer cell lines (10), neuroblastomas, and tumors of lung, cervix, and breast (10–13). There seems to be an inconsistent relationship between amplification of the hTERT gene and the levels of hTERT mRNA, protein expression, and telomerase activity (10–13). Our study has identified a reason for lack of correlation between hTERT expression levels and telomerase activity, namely inadequate up- regulation of hTR expression. There was a very low abundance of hTR in telomerase-negative B80 cells before hTERT overexpression, approximately three molecules per cell as measured by real-time RT-PCR (and undetectable by Northern blot; data not shown). Therefore, even a 15-fold up- regulation of hTR in B80-TERT1 (Fig. 3C) was associated with a telomerase activity level that resulted only in maintenance of very Figure 5. hTR, hTERT mRNA, and DKC1 mRNA expression in short telomeres. That hTR levels were limiting in the B80-TERT1 hTR-overexpressing B80-TERT1 clones quantitated by real-time RT-PCR. cells was confirmed by showing that expression of exogenous The log of ratios to parental B80-TERT1 cells is presented on the Y-axis. Columns, mean; bars, ranges for duplicate PCR reactions. Unpaired t test hTR resulted in high levels of telomerase activity and very long was used to compare BABE empty vector control clones (n = 3) and telomeres. hTR-transduced clones (n = 7). P values are shown in each panel. Although until recently it has been assumed that telomerase activity is controlled solely by the availability of hTERT, data are selected with puromycin. Individual colonies were picked and accumulating in support of the notion that hTR and hTERT are cultured for further analysis. Seven hTR-overexpressing clones both limiting. Notably, overexpression of both hTR and hTERT in (hTR15, hTR17, and hTR19-23), three pBABE vector-alone clones HeLa cells and fibroblasts resulted in high levels of telomerase (BABE2, BABE6, and BABE8), and parental B80-TERT1 cells were activity and massive telomere elongation (24). Moreover, the examined for hTR, hTERT, and DKC1 expression using real-time important contribution of hTR levels to telomerase activity is RT-PCR. indicated by the long-standing observation that hTR levels are In the hTR-infected cells, hTR was overexpressed from 35- to substantially elevated in a wide variety of human tumors (33–37) 125-fold compared with parental B80-TERT1 cells, whereas in the and immortal human cell lines (5). hTR amplification has also been

Cancer Res 2008; 68: (9). May 1, 2008 3120 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 2008 American Association for Cancer Research. hTERT Amplification in Cells with Limiting hTR Levels observed in various tumors (38–40), and it may be one mechanism hTERT per cell than 20 to 50 because hTERT is readily detectable of increased hTR expression in the tumors. More recently, a viral by our Western conditions in this cell line (Fig. 2C) and not in 293 form of telomerase RNA (vTR) with 88% homology to chicken cells (data not shown). Thus, our quantification of hTR molecules telomerase RNA was found to be encoded by oncogenic strains of per cell explains the very short telomeres maintained by this very Marek’s disease virus (MDV) but not by nononcogenic MDV strains large number of hTERT molecules: f20 molecules of hTR are (41). Furthermore, the wild-type vTR was shown to facilitate avian insufficient for significant telomere elongation (Fig. 6B), even in the tumorigenesis (42). More importantly, the decreased hTR levels presence of a vast excess of hTERT. observed in human dyskeratosis congenita patients due either to A previous study has reported that there are 11,000 to 13,500 hTR hTR or DKC1 mutations are associated with abnormally short molecules per cell in the telomerase-negative BJ, IMR90, and SW13 telomeres (reviewed by Kirwan et al. in ref. 7), which can be cell strains (25). Assuming that other normal human somatic cells rescued by overexpression of both hTERT and hTR (22, 23). It have hTR levels within an order of magnitude of the levels in seems that the hTR mutations associated with dyskeratosis conge- HMECs, our estimate of approximately three molecules of hTR per nita reduce telomerase activity via haploinsufficiency (21). Studies B80 cell is more consistent than the previous study (25) with the in mice also suggest telomerase RNA can be limiting (43, 44). observation that a 2-fold change of hTR abundance (hTR haplo- However, none of the above mentioned studies quantitated the insufficiency) can lead to dyskeratosis congenita (reviewed by number of hTR molecules per cell. It has been suggested that there Marrone et al. in ref. 45) because a 2-fold decrease in a vast excess are 20 to 50 active telomerase molecules per cell in 293 cells (4). of hTR would not be expected to affect telomere length. Our study B80-TERT1 cells are predicted to have many more molecules of is also in agreement with the observation of a low TLC1 abundance

Figure 6. hTR overexpression in B80-TERT1 results in increased telomerase activity and elongated telomeres. Telomerase assay products were analyzed on a 10% denaturing sequencing gel (A). Note that only 2.5 Â 106 293 cells were assayed, whereas 1 Â 107 cells were used for the others. Telomere length was measured by terminal restriction fragment (TRF) analysis using pulsed field gel electrophoresis (B). HeLa and GM847 are examples of telomerase-positive and Alternative Lengthening of Telomeres–positive cells, respectively. The correlation between telomerase activity and median telomere restriction fragment length is shown in C (the correlation value R was obtained using MS Excel software); points, mean values for three BABE clones and seven hTR clones; bars, SE.

www.aacrjournals.org 3121 Cancer Res 2008; 68: (9). May 1, 2008

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 2008 American Association for Cancer Research. Cancer Research in yeast (32). The reason for the discrepancy between our data and sensitivity of the assay (e.g., telomerase activity was readily those of Yi et al. (25) is likely to be technical rather than an intrinsic detectable in 2.5 Â 106 293 cells without overexpression of hTR difference between B80 cells on the one hand and telomerase- or hTERT). In the series of cell lines generated from B80 cells, the negative BJ, IMR90, and SW13 cells on the other because hTR levels rank order according to telomerase activity was the same as that were quantitated in 293 cells in both studies and reported to be for telomere length, and in the hTR-transduced B80-TERT1 cells, 120 to 300 molecules per cell (Northern and real-time RT-PCR; this up-regulation of telomerase activity was accompanied by a study) and f69,000 molecules per cell (competitor RNA RT-PCR corresponding increase in telomere length. In agreement with method; ref. 25). The number of RNA molecules per cell reported data from a previous study (24), this indicates that in human cells, here may be an underestimate due to the yield of RNA during cell telomerase activity level is a major determinant of telomere length. extraction being <100%, but this is unlikely to account for a 200- to Our observation that hTERT amplification correlates with hTERT 600-fold difference; furthermore, three different methods of RNA mRNA/protein expression but not telomerase activity in hTERT- extraction gave similar results (see Materials and Methods). immortalized HMECs indicates that hTERT amplification and The limiting hTR levels in the B80-TERT1 cells may have been hTERT expression are not good surrogate markers for telomerase the selection pressure that resulted in a growth advantage for a activity levels. Insufficient hTR up-regulation could be one of the clone of cells within the original mass culture that had acquired explanations for the previously observed nonassociation between amplification of hTERT. Before amplification of hTERT, it is hTERT amplification and telomerase activity in some tumors (10– possible that the cellular or local concentrations of both molecules 13). Our study is the first to examine the expression of other were below the Kd for the interaction, in which case, a higher telomerase components, hTR and dyskerin, in the context of hTERT concentration would favor the interaction between the hTERT-immortalized HMECs, and provides further direct evidence molecules and lead to increased telomerase activity, regardless of that hTR is limiting for telomerase activity and telomere length stoichiometry. Another possibility is that before amplification, maintenance. Furthermore, our quantitation of hTR levels hTERT may have been poorly expressed, as it is in B80-TERT3a (approximately three hTR molecules per cell in telomerase-negative (Fig. 2C), such that the number of hTERT molecules was lower than B80 cells) provides an explanation for this observation, which was hTR, and this limiting hTERT expression may have selected for at odds with the previously reported vast excess of hTR in amplification of hTERT. Alternatively, cells with increased hTERT telomerase-negative cells. may have a growth advantage due to putative telomere length- independent functions of hTERT (46, 47). Acknowledgments Up-regulation of hTR has previously been observed in cells Received 11/25/2007; revised 2/7/2008; accepted 2/21/2008. expressing exogenous hTERT (22, 23, 48). The increased hTR was Grant support: National Health and Medical Research Council (NHMRC) Peter suggested to be due to an increased half-life as a result of the Doherty Postdoctoral Fellowships (228413, Y. Cao; 228412, A.S. Nouwens), NHMRC association of hTR with hTERT (48). Our results indicate that this Healthy Ageing Research Grant (219306, L.I. Huschtscha and R.R. Reddel), and a Wellcome Trust Senior Research Fellowship (GRO66727MA, T.M. Bryan). cannot be the whole explanation because the cells with the highest The costs of publication of this article were defrayed in part by the payment of page hTERT protein levels had the lowest hTR levels. charges. This article must therefore be hereby marked advertisement in accordance We found a very consistent correlation between telomere length with 18 U.S.C. Section 1734 solely to indicate this fact. We thank Kathleen Collins for the construct pBABE-puro-U3-hTR; Karen and telomerase activity (Fig. 6C) as measured by the direct MacKenzie and Laura Veas for the hTR2 and B2M primers; Julie Jurczyluk for the telomerase activity assay. This assay includes a step in which construct pJJ101-hTERT1-182aa, in vitro transcribed hTR, and technical assistance; Scott Cohen for the construct pUC-hTR and the direct telomerase activity assay; and telomerase is partially purified by immunoprecipitation from cell Jeremy Henson, Akira Nguyen, Ze Huai Zhong, Jane Noble, and Elizabeth Collins for extracts with an hTERT antibody, which greatly increases the technical assistance and advice.

References 9. Horikawa I, Barrett JC. Transcriptional regulation of activity in normal human cells leads to elongation of the telomerase hTERT gene as a target for cellular and telomeres and extended replicative life span. Curr Biol 1. Moyzis RK, Buckingham JM, Cram LS, et al. A highly viral oncogenic mechanisms. Carcinogenesis 2003;24: 1998;8:279–82. conserved repetitive DNA sequence, (TTAGGG)n, pres- 1167–76. 17. MacKenzie KL, Franco S, May C, Sadelain M, Moore ent at the telomeres of human chromosomes. Proc Natl 10. Zhang A, Zheng C, Lindvall C, et al. Frequent MA. Mass cultured human fibroblasts overexpressing Acad Sci U S A 1988;85:6622–6. amplification of the telomerase reverse transcriptase hTERT encounter a growth crisis following an extended 2. Harley CB, Futcher AB, Greider CW. Telomeres gene in human tumors. Cancer Res 2000;60:6230–5. period of proliferation. Exp Cell Res 2000;259:336–50. shorten during ageing of human fibroblasts. Nature 11. Fan X, Wang Y, Kratz J, et al. hTERT gene 18. Noble JR, Zhong ZH, Neumann AA, Melki JR, Clark SJ, 1990;345:458–60. amplification and increased mRNA expression in central Reddel RR. Alterations in the p16(INK4a) and p53 tumor 3. Greider CW, Blackburn EH. Identification of a specific nervous system embryonal tumors. Am J Pathol 2003; suppressor genes of hTERT-immortalized human fibro- telomere terminal transferase activity in Tetrahymena 162:1763–9. blasts. Oncogene 2004;23:3116–21. extracts. Cell 1985;43:405–13. 12. Palmqvist R, Zhang A, Xu D, et al. hTERT gene copy 19. Wen VW, Wu K, Baksh S, et al. Telomere-driven 4. Cohen SB, Graham ME, Lovrecz GO, Bache N, number is not associated with hTERT RNA expression karyotypic complexity concurs with p16INK4a inactiva- Robinson PJ, Reddel RR. Protein composition of or telomerase activity in colorectal cancer. Int J Cancer tion in TP53-competent immortal endothelial cells. catalytically active human telomerase from immortal 2005;116:395–400. Cancer Res 2006;66:10691–700. cells. Science 2007;315:1850–3. 13. Zhu CQ, Cutz JC, Liu N, et al. Amplification of 20. Vulliamy T, Marrone A, Goldman F, et al. The RNA 5. NakamuraTM,MorinGB,ChapmanKB,etal. telomerase (hTERT) gene is a poor prognostic marker in component of telomerase is mutated in autosomal do- Telomerase catalytic subunit homologs from fission non-small-cell lung cancer. Br J Cancer 2006;94:1452–9. minant dyskeratosis congenita. Nature 2001;413:432–5. yeast and human. Science 1997;277:955–9. 14. Avilion AA, Piatyszek MA, Gupta J, Shay JW, Bacchetti 21. Marrone A, Stevens D, Vulliamy T, Dokal I, Mason PJ. 6. Feng J, Funk WD, Wang SS, et al. The RNA component S, Greider CW. Human telomerase RNA and telomerase Heterozygous telomerase RNA mutations found in of human telomerase. Science 1995;269:1236–41. activity in immortal cell lines and tumor tissues. Cancer dyskeratosis congenita and aplastic anemia reduce 7. Kirwan M, Dokal I. Dyskeratosis congenita: a genetic Res 1996;56:645–50. telomerase activity via haploinsufficiency. Blood 2004; disorder of many faces. Clin Genet 2008;73:103–12. 15. Bodnar AG, Ouellette M, Frolkis M, et al. Extension of 104:3936–42. 8. Kim NW, Piatyszek MA, Prowse KR, et al. Specific life-span by introduction of telomerase into normal 22. Wong JM, Collins K. Telomerase RNA level limits association of human telomerase activity with immortal human cells. Science 1998;279:349–52. telomere maintenance in X-linked dyskeratosis conge- cells and cancer. Science 1994;266:2011–5. 16. Vaziri H, Benchimol S. Reconstitution of telomerase nita. Genes Dev 2006;20:2848–58.

Cancer Res 2008; 68: (9). May 1, 2008 3122 www.aacrjournals.org

23. Westin ER, Chavez E, Lee KM, et al. Telomere in yeast: implications for telomerase haploinsufficiency. 41. Fragnet L, Blasco MA, Klapper W, Rasschaert D. The restoration and extension of proliferative lifespan in RNA 2006;12:1721–37. RNA subunit of telomerase is encoded by Marek’s dyskeratosis congenita fibroblasts. Aging Cell 2007;6: 33. Soder AI, Going JJ, Kaye SB, Keith WN. Tumour disease virus. J Virol 2003;77:5985–96. 383–94. specific regulation of telomerase RNA gene expression 42. Trapp S, Parcells MS, Kamil JP, et al. A virus-encoded 24. Cristofari G, Lingner J. Telomere length homeostasis visualized by in situ hybridization. Oncogene 1998;16: telomerase RNA promotes malignant T cell lymphoma- requires that telomerase levels are limiting. EMBO J 979–83. genesis. J Exp Med 2006;203:1307–17. 2006;25:565–74. 34. Heine B, Hummel M, Demel G, Stein H. Demonstra- 43. Hathcock KS, Hemann MT, Opperman KK, Strong 25. Yi X, Shay JW, Wright WE. Quantitation of telomerase tion of constant upregulation of the telomerase RNA MA, Greider CW, Hodes RJ. Haploinsufficiency of mTR components and hTERT mRNA splicing patterns in component in human gastric carcinomas using in situ results in defects in telomere elongation. Proc Natl Acad immortal human cells. Nucleic Acids Res 2001;29: hybridization. J Pathol 1998;185:139–44. Sci USA 2002;99:3591–6. 4818–25. 35. Paradis V, Dargere D, Laurendeau I, et al. Expression 44. Chiang YJ, Hemann MT, Hathcock KS, et al. 26. Toouli CD, Huschtscha LI, Neumann AA, et al. of the RNA component of human telomerase (hTR) in Expression of telomerase RNA template, but not telo- Comparison of human mammary epithelial cells im- prostate cancer, prostatic intraepithelial neoplasia, and merase reverse transcriptase, is limiting for telomere mortalized by simian virus 40 T-Antigen or by the normal prostate tissue. J Pathol 1999;189:213–8. length maintenance in vivo. Mol Cell Biol 2004;24: telomerase catalytic subunit. Oncogene 2002;21:128–39. 36. Yashima K, Litzky LA, Kaiser L, et al. Telomerase 7024–31. 27. Perrem K, Colgin LM, Neumann AA, Yeager TR, expression in respiratory epithelium during the multi- 45. Marrone A, Walne A, Dokal I. Dyskeratosis congenita: Reddel RR. Coexistence of alternative lengthening of stage pathogenesis of lung carcinomas. Cancer Res 1997; telomerase, telomeres and anticipation. Curr Opin telomeres and telomerase in hTERT-transfected GM847 57:2373–7. Genet Dev 2005;15:249–57. cells. Mol Cell Biol 2001;21:3862–75. 37. Morales CP, Lee EL, Shay JW. In situ hybridization for 46. Stewart SA, Hahn WC, O’Connor BF, et al. Telomer- 28. Colgin LM, Wilkinson C, Englezou A, Kilian A, the detection of telomerase RNA in the progression ase contributes to tumorigenesis by a telomere length- Robinson MO, Reddel RR. The hTERTa splice variant from Barrett’s esophagus to esophageal adenocarcino- independent mechanism. Proc Natl Acad Sci U S A 2002; is a dominant negative inhibitor of telomerase activity. ma. Cancer 1998;83:652–9. 99:12606–11. Neoplasia 2000;2:426–32. 38. Soder AI, Hoare SF, Muir S, Going JJ, Parkinson EK, 47. Cao Y, Li H, Deb S, Liu JP. TERT regulates cell 29. Cohen SB, Reddel RR. A sensitive direct human Keith WN. Amplification, increased dosage and in situ survival independent of telomerase enzymatic activity. telomerase activity assay. Nat Methods. In press 2008. expression of the telomerase RNA gene in human Oncogene 2002;21:3130–8. 30. Bryan TM, Englezou A, Gupta J, Bacchetti S, Reddel cancer. Oncogene 1997;14:1013–21. 48. Yi X, Tesmer VM, Savre-Train I, Shay JW, Wright WE. RR. Telomere elongation in immortal human cells 39. Heselmeyer-Haddad K, Janz V, Castle PE, et al. Both transcriptional and posttranscriptional mecha- without detectable telomerase activity. EMBO J 1995; Detection of genomic amplification of the human nisms regulate human telomerase template RNA levels. 14:4240–8. telomerase gene (TERC) in cytologic specimens as a Mol Cell Biol 1999;19:3989–97. 31. Grant JD, Broccoli D, Muquit M, Manion FJ, Tisdall J, genetic test for the diagnosis of cervical dysplasia. Am J 49. Bryan TM, Marusic L, Bacchetti S, Namba M, Ochs MF. Telometric: a tool providing simplified, repro- Pathol 2003;163:1405–16. Reddel RR. The telomere lengthening mechanism in ducible measurements of telomeric DNA from constant 40. Andersson S, Wallin KL, Hellstrom AC, et al. Frequent telomerase-negative immortal human cells does not field agarose gels. Biotechniques 2001;31:1314–8. gain of the human telomerase gene TERC at 3q26 in involve the telomerase RNA subunit. Hum Mol Genet 32. Mozdy AD, Cech TR. Low abundance of telomerase cervical adenocarcinomas. Br J Cancer 2006;95:331–8. 1997;6:921–6.

www.aacrjournals.org 3123 Cancer Res 2008; 68: (9). May 1, 2008

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 2008 American Association for Cancer Research. Amplification of Telomerase Reverse Transcriptase Gene in Human Mammary Epithelial Cells with Limiting Telomerase RNA Expression Levels

Ying Cao, Lily I. Huschtscha, Amanda S. Nouwens, et al.

Cancer Res 2008;68:3115-3123.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/68/9/3115

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2008/04/30/68.9.3115.DC1

Cited articles This article cites 48 articles, 20 of which you can access for free at: http://cancerres.aacrjournals.org/content/68/9/3115.full#ref-list-1

Citing articles This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/68/9/3115.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/68/9/3115. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.