Aneuploidy and Telomere Attrition Are Independent Determinants of Crisis in SV40-Transformed Epithelial Cells1

Home , Ploidy

[CANCER RESEARCH 63, 5813–5820, September 15, 2003] Aneuploidy and Telomere Attrition Are Independent Determinants of Crisis in SV40-transformed Epithelial Cells1

Mihaela Velicescu, Jiamei Yu, Brittney-Shea Herbert, Jerry W. Shay, Eileen Granada, and Louis Dubeau2 University of Southern California/Norris Comprehensive Cancer Center, Department of Pathology, University of Southern California, Keck School of Medicine, Los Angeles, California 90089-9181 [M. V., J. Y., E. G., L. D.], and University of Texas Southwestern Medical Center, Department of Cell Biology, Dallas, Texas 75390-9039 [B-S. H., J. W. S.]

ABSTRACT locations (14), and regional deletions and amplifications (15). Most of these abnormalities have been postulated to be a direct consequence of Replicative immortality is achieved in vitro by overcoming two mortal- the loss of telomeric sequences through the formation of dicentric ity checkpoints, M1 (senescence) and M2 (crisis). Cancer cells are thought chromosomes and breakage-fusion-bridge cycles (16). In contrast, to overcome M2 by activating telomerase, an enzyme believed to confer genomic stability in addition to maintaining telomeric sequences above a numerical chromosomal alterations leading to aneuploidy cannot be critical length. Here we show that a subset of cultured ovarian cystade- simply explained by a simple loss of telomeric sequences (17, 18). noma cells expressing SV40 large T-antigen, which allows bypassing of These alterations are linked to missegregation of the chromosomes, M1, develop a specific type of genomic instability, characterized by nu- possibly due to loss of p53 and RB proteins, which are essential merical (as opposed to structural) chromosomal alterations, that leads to regulators of the mitotic spindle assembly (19) and of normal centro- non-telomere-based premature growth arrest/crisis. Cells recover from some formation during mitosis (20, 21). The ectopic expression of this type of growth arrest and stabilize their ploidy status without telom- telomerase appears to protect against genomic instability (22). It has erase expression. In these cases, telomeres continue to shorten until a been suggested that telomere-based DNA damage may be prevented second, telomere-based growth arrest/crisis event is reached. Transfection by the stabilization of the telomere ends, whereas nontelomere DNA of the catalytic subunit of telomerase does not immortalize cells harboring damage may be more efficiently detected and repaired in cells ex- severe abnormalities in their DNA ploidy but results in immortalization of diploid cells. We conclude that changes in DNA ploidy constitute an pressing telomerase (23). This idea that telomerase expression may important determinant of growth arrest that is independent of telomere confer genetic stability is intriguing, given that this enzyme is ex- attrition in a subset of SV40 large T-antigen-expressing cystadenoma cells. pressed in most cancer cells, which are thought to be genetically more Reestablishment or emergence of ploidy stability, which is not always unstable than their normal counterparts that do not express telomerase. dependent on telomerase activation, is necessary for acquisition of the We sought to better understand the relationship between crisis, potential to achieve replicative immortality. telomere attrition, and development of aneuploidy because such un- derstanding might provide important insights into the mechanism of genetic instability in human cancers. Although our knowledge of in INTRODUCTION vitro mortality checkpoints comes largely from work done with fi- One of the most consistent differences between normal cells and broblasts, we used an epithelial cell model because most human cancer cells cultured in vitro is the fact that cancer cells can divide cancers arise in epithelial cells. We took advantage of established indefinitely, whereas normal cells have a limited life span (1). Most cultures of epithelial cells derived from benign ovarian epithelial normal cells lose telomeric DNA each time they undergo DNA tumors (cystadenomas), which express an adenovirus vector for SV40 synthesis due to their inability to replicate their chromosomal ends (2). large T-antigen that had become stably integrated into the host ge- Such telomere attrition is thought to trigger growth arrest signals that nome (24). This viral oncoprotein, which interferes with RB and p53 limit the life span of normal cells by activating two mortality check- proteins, allows bypassing of M1 but not of M2 (8). Two cell clones points known as senescence [M1] (3, 4) and crisis [M2] (5). M1 is derived from one SV40 large T-antigen-expressing cystadenoma cell characterized by absence of cell division due to inhibition of the cell strain called ML10 (24) recovered from M2 and became immortal cell cycle (6). Cells that overcome or bypass M1 due to loss of cell lines, providing a longitudinal model for studying the mechanisms cycle-inhibitory signals, such as the absence of functional RB or p53 associated with acquisition of ability to bypass crisis. Our observa- proteins, can extend their replicative life span but eventually reach a tions suggest that the form of genetic instability that is associated with second mortality checkpoint, M2, also known as crisis (7, 8). It is numerical chromosomal alterations, in contrast to that associated with thought that cells that overcome this checkpoint, usually through structural chromosomal changes, precedes significant telomere attri- activation of telomerase, an enzyme capable of maintaining telomeric tion in this cell culture model, implying that shortening of the te- sequences above a critical length, acquire the ability to grow indefi- lomeres is unlikely to play any significant role in the induction of this nitely (9–11). In support of these concepts, in vitro replicative im- form of instability. Our results further suggest that this form of mortality is achieved by ectopic expression of the human catalytic genomic instability, which is also prominent in most human cancers, subunit of telomerase (hTERT) in normal cells (12). is an independent mediator of crisis that is not overcome by telom- Crisis is often accompanied by widespread genomic instability (5) erase expression. These findings not only clarify the relationship characterized by both structural and numerical chromosomal alter- between crisis and genomic instability but also provide insights into ations. Structural alterations include fusions (13), nonreciprocal trans- potential mechanisms for the acquisition of aneuploidy during cancer development.

Received 8/14/02; revised 6/25/03; accepted 7/8/03. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with MATERIALS AND METHODS 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by National Cancer Institute Grants CA 51167, CA 79750, CA70907, and Cell Lines and Strains and Culture Conditions. ML3, ML5, and ML10 CA14089. cell strains were established from primary cultures of benign ovarian epithelial 2 To whom requests for reprints should be addressed, at Department of Pathology, tumors (cystadenomas) that were infected with an adenovirus vector express- University of Southern California/Norris Comprehensive Cancer Center, Keck School of Medicine of University of Southern California, 1441 Eastlake Avenue, Los Angeles, CA ing SV40 large T-antigen (24). This vector had become stably integrated into 90089-9181. Phone: (323) 865-0720; Fax: (323) 865-0077; E-mail: [email protected]. the host genomes upon multiple reinfections (24). MCV39, MCV500, and 5813

MCV50 cells were derived from ML10 as described in “Results.” HOC-7 cells ethanol and air dried. The slides were incubated with a hybridization mixture ␮ Ј Ј were obtained from Dr. Ronald N. Buick (University of Toronto; Ref. 25). All (20 l) containing 70% formamide, 3 -Cy3-conjugated (CCCTAA)3 2 - Ј Ј cells were grown in MEM (Invitrogen, Carlsbad, CA) supplemented with 10% deoxyoligonucleotide N3 -P5 phosphoramidate telomeric probe and FITC- FBS.3 The origin and characteristics of VA13 cells have been described conjugated centromeric probe (kindly provided by Geron Corp., Menlo Park, previously (26). CA), 0.25% (w/v) blocking reagent (Roche Molecular Biochemicals), and 5%

Telomere Length Determination. Ten-␮g samples of genomic DNA di- MgCl2 in 10 mM Tris (pH 7.2) for 3 min at 78°C. The slides were then gested with RsaI/HinfIII restriction endonucleases were electrophoresed on incubated for2hatroom temperature and washed twice with 70% formamide, 0.8% agarose gels and transferred to Zetabind (Cuno, Inc., Meride, CT) 0.1% BSA, and 10 mM Tris (pH 7.2). After two washes with 0.15 M NaCl, membranes according to the manufacturer’s protocol. The membranes were 0.05% Tween 20, and 0.05 M Tris, the slides were dehydrated by a 2-min hybridized to a 32P-labeled probe consisting of the basic human telomeric incubation in ethanol, air dried in the dark, mounted with Vectashield containing 4Ј,6-diamidino-2-phenylindole (Vector Laboratories), and imaged us- sequence (TTAGGG)3 in Church buffer [500 mM NaPO4 (pH 6.8), 7% SDS, and1mM EDTA (pH 8.0)] at 42°C for 24 h. The membranes were washed ing a Zeiss Axioplan 2 microscope. twice for 10 min at room temperature in 2ϫ SSC, 0.5% SDS, followed by two ϫ 15-min washes in 0.2 SSC, 0.1% SDS. The hybridization signals were RESULTS visualized by phosphorimaging. Assay of Telomerase Activity. Telomerase activity was detected using the Numerical Chromosomal Alterations Precede Telomere Attri- TRAP assay (27) as described previously (28). tion in Ovarian Cystadenoma Cells Expressing SV40 Large T- Analysis of DNA Ploidy by Flow Cytometry. One million cells resus- antigen. Changes in DNA ploidy are a well-documented manifesta- pended in PBS were fixed in 70% ethanol. After centrifugation, the cell pellets tion of chromosomal instability in cells expressing SV40 large were resuspended in 1 ml of PBS, 10 ␮g/ml propidium iodide, and 100 ␮g/ml T-antigen (5, 29–31). The presence and magnitude of such ploidy RNase. Fluorescence was measured on a Coulter Profile II flow cytometer (Beckman Coulter, Hialeah, FL) and analyzed using the MultiCycle software changes in ovarian cystadenoma cells expressing SV40 large T- (Phoenix Flow Systems Inc., San Diego, CA). antigen are illustrated in Fig. 1a, which compares the DNA content of FACS. Cells cultured on plastic dishes were incubated in MEM (Life a strain of ovarian cystadenoma called ML10 after approximately 30 Technologies, Inc., Grand Island, NY) supplemented with 10% FBS in the versus 42 PDs in vitro. Although the cells were predominantly diploid presence of 10 ␮M Hoechst 33342 reagent (Molecular Probes, Inc., Eugene, after 30 PDs, the majority had become either tetraploid or aneuploid OR) for 90 min at 37°C. After dissociation with 0.05% trypsin/0.02% EDTA, after as little as 12 additional PDs. We estimated the percentage of the cells were resuspended in MEM plus 20% FBS and 10 ␮M Hoechst 33342. aneuploid cells in cultured cystadenomas from the ratio of the number The cells were kept at 4°C until sorted based on fluorescence intensity using of nondiploid cells (excluding apoptotic cells) over the total number a FACSar plus cytometer (Becton Dickinson, San Jose, CA). of nonapoptotic cells present based on flow cytometry data. Compar- TUNEL Assay. Cells cultured on Lab-Tek chamber slides (Miles Scien- ing the changes in ploidy status with changes in doubling time at tific) were fixed in freshly prepared 4% methanol-free formaldehyde (Poly- various time points preceding crisis showed that the ploidy changes sciences, Inc., Warrington, PA) in PBS for 25 min at 4°C. After incubation in did not accumulate at a constant rate throughout the life span of the equilibration buffer containing terminal deoxytransferase and fluorescein- labeled dUTP (Apoptosis Detection Kit; Promega, Madison, WI) at 37°C for cells but took place primarily during the 5 PDs that preceded any 60 min inside a humidified chamber, the cells were protected from light and noticeable increase in doubling time due to crisis (Fig. 1b). An treated for 15 min at room temperature with 40 ml of 1 mg/ml propidium abundance of aneuploid cells in this period was confirmed by exam- iodide solution freshly diluted in PBS. Cells with three or more green fluo- ining the number of chromosomes in metaphase spreads prepared rescent nuclear signals seen under a fluorescence microscope were scored as from these SV40-transfected cells (data not shown). These results are positive for apoptosis. similar to what was observed by Romanov et al. (32) in primary Detection of APBs by Immunofluorescence. Cells grown on coverslips cultures of normal breast epithelial cells. overnight were fixed with 4% paraformaldehyde and permeabilized with 0.1% We sought to determine whether the genomic instability resulting in Triton X-100 before blocking with 3% BSA. Staining was performed with a these changes in DNA ploidy, which precede the onset of crisis, PML and either a Rad51 or a TRF2 antibody (Santa Cruz Biotechnology) coincides with changes in telomere length. Although the current diluted in PBS with 0.5% sodium azide. Secondary antibodies included Alexa- evidence suggests an important role for telomere attrition in the Fluor 488-conjugated antimouse and 568-conjugated antirabbit immunoglobu- induction of crisis (17, 18), the role of the type of genetic instability lins (Molecular Probes). The coverslips were mounted with Vectashield containing 4Ј,6-diamidino-2-phenylindole (Vector Laboratories), and cells were that is associated with numerical chromosomal changes leading up to examined using a Zeiss fluorescence microscope. The percentage of APBs crisis is less clear. TRF length was examined by Southern blotting in were scored as the percentage of total cells with PML/Rad51 or PML/TRF2 three strains of ovarian cystadenomas, ML3, ML5, and ML10, after colocalized in the nuclei. 20 PDs and after either 45 PDs (for ML3), 40 PDs (for ML5), or 50 Metaphase Spread Preparation and FISH. Cultured cells were treated PDs (for ML10) in vitro. These cell strains, none of which express with 5 ␮g/ml Colcemid (Invitrogen) for 4 h and then harvested as usual. After telomerase, reach crisis after approximately 40–45, 35–40, and a 30-min incubation in hypotonic 0.075 M KCl at 37°C, the cells were fixed 45–50 PDs, respectively. One telomerase-positive ovarian carcinoma and washed three times with methanol/acetic acid (3:1) and dropped onto cell line (HOC-7) was also included in the analysis for comparison. slides. One- to seven-day-old slides were rehydrated in 1ϫ PBS (pH 7.5) for Although a slight shortening in the average TRF length had indeed 15 min at room temperature. The slides were fixed in 4% formaldehyde in PBS occurred in all three strains when they reached crisis, the magnitude ϫ (pH 7.5) for 2 min and washed in 1 PBS three times (5 min each time). The of the mean telomere shortening (cell to cell variation) was minimal slides were treated with1 mg/ml pepsin (pH 2) at 37°C for 10 min and washed (Fig. 1c), suggesting that this type of crisis was either initiated by a twice for 2 min each time in 1ϫ PBS. The slides were then fixed in the formaldehyde solution for 2 min and washed again in 1ϫ PBS three times. few short telomeres or that crisis was not caused by telomere short- Slides were dehydrated by 2-min serial incubations in 70%, 90%, and 100% ening at all. The average telomere length in ML10 cells at 20 PDs in this experiment was 8.60 Ϯ 0.38 kb, whereas that of ML10 cells entering crisis (50 PDs) was 7.34 Ϯ 0.10 kb. This size was much 3 The abbreviations used are: FBS, fetal bovine serum; TRF, telomere restriction fragment; PD, population doubling; pdM2, ploidy-dependent M2; tdM2, telomere-depen- greater than that seen in HOC-7 cells (Fig. 1c), which had a calculated dent M2; ALT, alternative lengthening of telomeres; TRAP, telomerase repeat amplifi- average telomere length of 2.71 Ϯ 0.23 kb. Examination of the cation protocol; FACS, fluorescence-activated cell sorting; TUNEL, terminal deoxynucle- otidyl transferase-mediated nick end labeling; PML, promyelocytic leukemia; APB, chromosomes in metaphase spreads of cystadenoma cells approaching ALT-associated PML body; FISH, fluorescence in situ hybridization. crisis showed no evidence of end-to-end fusion in 200 cells examined. 5814

the cells recovered in the diploid fraction (data not shown). Although the median number of chromosomes in the nondiploid cells was around 92 (i.e., near tetraploid), there was marked heterogeneity within the cell population, with the majority of the cells showing varying degrees of aneuploidy. There were no significant differences in the doubling time of the sorted diploid and nondiploid cells immediately after the sorting procedures (data not shown). However, growth curve analyses performed 10 PDs after the sorting procedures revealed a substantial increase in the doubling time of nondiploid cells compared with cells derived from age-matched diploid fractions (Fig. 2b). The sorted diploid cells maintained their logarithmic growth for an additional 10 PDs after the aneuploid cells reached crisis, at which point they also reached crisis. Thus, cells harboring severe ploidy changes reached crisis earlier than the age-matched diploid cells. This result was reproduced in seven independent cell sorting experiments. Growth arrest in the nondiploid cell fraction was not due to loss of SV40 large T-antigen because this antigen continued to be expressed in both cell fractions after the sorting procedures (Fig. 2c). The p53 and p21 proteins, which become elevated during in vitro crisis (data not shown), showed higher expression levels in the nondiploid cells (Fig. 2c). Although this may provide further support to the idea that crisis was more advanced in these cells, the contribution of potential differences in gene dosage due to ploidy changes involving the p53 and p21 loci has not been investigated. The telomeres were of similar

Fig. 1. Changes in DNA content and telomere length in ovarian cystadenoma cells expressing SV40 large T-antigen. a, ML10 cells, derived from an ovarian cystadenoma and transfected with an expression vector for SV40 large T-antigen, were stained with 10 ␮g/ml propidium iodide and examined for DNA content by flow cytometry after 30 and 42 PDs in vitro. b, ML10 cells that had been in culture for approximately 25 PDs were analyzed by flow cytometry to determine the percentage of cells with a nondiploid DNA content. These measurements were repeated every 3.3 PDs until the cells reached crisis. Doubling times were measured at each time point by growth curve analysis. c, telomere length was analyzed by Southern blotting in ML3, ML5, and ML10 ovarian cystadenoma cells after 20 PDs or after either 45, 40, or 50 PDs as indicated. ML3, ML5, and ML10 cells typically reach crisis after 40–45, 35–40, and 45–50 PDs, respectively.

Genetic Instability Resulting in Numerical Chromosomal Al- terations Is a Determinant of Crisis Independent of Telomere Shortening. The results of Fig. 1b strongly suggest the existence of a close relationship between the development of alterations in DNA ploidy and the establishment of crisis. We investigated whether such alterations could by themselves play a causative role in the development of crisis. ML10 cells that had undergone approximately 25 PDs in vitro were stained with Hoechst 33342 dye, a fluorescent DNA binding compound able to diffuse intracellularly into living cells. The Fig. 2. Isolation and characterization of ML10 cell subpopulations fractionated based intensity of fluorescence emitted reflected cellular DNA content, on differences in their DNA content. a, ML10 ovarian cystadenoma cells that had ␮ allowing separation of different cell populations based on their ploidy completed an average of 25 PDs in vitro were treated with 10 M Hoechst 33342 reagent and separated into diploid and nondiploid fractions by FACS. R1 and R2 indicate the status by FACS (Fig. 2a). Metaphase spreads prepared from cells that position of the sorting gates. b, the sorted diploid and nondiploid fractions were analyzed were put back in culture after having been subjected to such sorting by growth curve analysis performed 10 PDs after the sorting procedures. The sorted diploid and nondiploid fractions were analyzed for expression of SV40 large T-antigen, procedures showed that 24% of the cells recovered in the nondiploid p53, p21, and actin (used as protein loading control) by Western blotting (c) and for fraction had a normal number of chromosomes compared with 76% of telomere length by Southern blotting (d). 5815

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2003 American Association for Cancer Research. DETERMINANTS OF IN VITRO CRISIS sizes in both cellular fractions (Fig. 2d), suggesting that these differences in the timing of initiation of crisis were not necessarily due to differences in telomere attrition. Further evidence that the changes in in vitro kinetics represented crisis as opposed to senescence comes from lack of expression of senescence-associated ␤-galactosidase, a well-established senescence-associated marker (33), in cells undergo- ing growth arrest (data not shown). Recovery from Crisis-based Chromosomal Instability Is Not Dependent on Telomerase Activation. ML10 ovarian cystadenoma cells sorted based on their DNA content were kept in culture through their crisis period. Of 17 separate experiments with either sorted or unsorted cells, spontaneous recovery from crisis occurred twice, resulting in the establishment of two cell lines called MCV39 and MCV500. Both cell lines were derived from nondiploid fractions of sorted ML10 cells. Microscopic examination of metaphase spreads revealed that MCV39 had 57 chromosomes, whereas MCV500 had 67 chromosomes (data not shown). MCV39 cells were cultured for over 200 PDs and showed no change in their in vitro kinetics over this time period, attesting to their true immortality. In contrast, MCV500 cells showed logarithmic growth for only 35 PDs, at which point they underwent a second crisis. Recovery from this second crisis was observed only once in five separate attempts (using MCV500 cells that had been stored frozen in liquid nitrogen soon after their recovery from the first crisis), resulting in the establishment of a second immortal cell line, derived from MCV500, called MCV50. The earliest time at which we were able to analyze TRF length in MCV39 and MCV500 cells was 20 PDs after they had recovered from crisis. Compared with the parental ML10 cells, MCV39 cells had considerably shorter telomeres at this time point (20 PDs) and ex- Fig. 3. Examination of telomerase activity and genetic stability after recovery from crisis. Sorted nondiploid ML10 cells were kept in culture approximately 4 weeks after pressed telomerase (Fig. 3). Telomere length and telomerase activity reaching crisis, at which time two cell clones, MCV39 and MCV500, recovered from were essentially unchanged in this cell line 30 PDs later (Fig. 3). The crisis. After 35 further PDs, MCV500 cells underwent a second crisis event. An immortal cell clone called MCV50 recovered from this second crisis. Top panel, Southern blot severe imbalances in DNA ploidy that characterized the parental analysis of telomere length in the parental ML10 cells and in MCV39, MCV500, and ML10 cells were no longer apparent in MCV39 cells, at least based on MCV50 cells at various time points after their isolation (expressed as number of PDs). analysis of cellular DNA content by flow cytometry (Fig. 3). Middle panel, telomerase activity was examined by TRAP assay at selected time points in the various cell lines and strains. The characteristic ladders in the second, third, and fifth MCV500 cells differed from age-matched MCV39 cells in that they lanes indicate the presence of such activity. Bottom panel, examination of DNA content showed no measurable telomerase activity, and their average TRF in the parental ML10 cells and in MCV39 and MCV500 cells by flow cytometry. G1 length was substantially larger after 20 PDs (Fig. 3). Similarly to indicates the position of the peaks corresponding to cells in the G1 phase of the cell cycle. MCV39, however, they showed a relatively stable DNA profile compared with the parental ML10 cells (Fig. 3). Thus, telomerase reacti- vation was not required for overcoming the first crisis event in experiments on VA13 cells using a probe for telomeric/centromeric MCV500, nor was it required for conferring stability in DNA ploidy. DNA. These cells, which were derived from normal human lung The average TRF length decreased rapidly between the first and fibroblasts transfected with SV40 large T-antigen, were previously second crisis (Fig. 3). Recovery from the second crisis event occurred shown to use ALT mechanisms (26). In contrast, very few, if any, in cells with short TRF lengths and was accompanied by the appear- extrachromosomal telomeric repeats were observed and there was no ance of telomerase activity (Fig. 3). This activity persisted at later evidence of long and heterogeneous telomeres in ML10 cells (Fig. passages, which showed no further changes in TRF length (Fig. 3). 4c). We conclude that the lack of significant telomere attrition in Absence of ALT. We were intrigued by the fact that whereas little cystadenoma cells approaching crisis was not due to the presence of telomere attrition took place before the initiation of ploidy-dependent ALT mechanisms. crisis (pdM2) in all three cystadenoma cell strains examined, the rate Consequences of Ectopic Expression of hTERT in Diploid ver- of attrition appeared to be accelerated after recovery from pdM2 in sus Aneuploid Cells. We further examined the role of telomerase on MCV500 cells. We investigated the possibility that our cystadenoma chromosomal stability in our cell culture model by introducing an cell strains developed ALT (34) in the absence of telomerase before expression vector for the catalytic subunit of the human telomerase the first crisis. Indeed, ALT mechanisms are common in SV40- cDNA (hTERT) into ML3, ML5, and ML10 cystadenoma cells. transformed cell lines (35). We therefore performed immunofluores- Several immortal clones of transfected cells were isolated that were cence staining of our cultured cells with antibodies against PML essentially diploid and showed a normal and stable DNA profile when bodies and either Rad51 or a TRF2, which are components of APBs analyzed by flow cytometry (data not shown). Only a small proportion (36). The percentage of cells containing such bodies, which was of aneuploid cells were detectable immediately after the DNA trans- scored as the percentage of total cells with PML/Rad51 or PML/TRF2 fection procedures. These transfected aneuploid cells were separated colocalized in the nuclei (Fig. 4, a and b), was less than 1% for most from the more abundant diploid cells by FACS. Although we were of the cells examined. Another and perhaps more specific character- able to obtain cell populations that were highly enriched for telomer- istic of ALT is the presence of marked heterogeneity in telomere ase-positive aneuploid cells using such sorting procedures, the small length and extrachromosomal telomeric repeats (35). These changes number of diploid cells present in the sorted cell population outgrew are readily apparent in Fig. 4d, which shows the results of FISH the nondiploid cells in as little as 35 PDs (Fig. 5, a and b). The cells 5816

Fig. 4. Lack of ALT mechanisms in cultured ovarian cystadenomas. a and b, ML10 cells were reacted with primary antibodies against PML bodies and Rad51 (a) or TRF2 (b), followed by appro- priate fluorescent secondary antibodies. Immuno- reactivity for PML resulted in green fluorescence (long arrows), whereas reactivity for either Rad51 or TRF2 resulted in red fluorescence (short arrows). The figure shows superimposed photographs of the two immunostains, revealing very little nuclear colocalization of either PML/Rad51 or PML/ TRF2 and, thus, no evidence of APBs. c and d, analysis of ML10 ovarian cystadenoma cells (c) and VA13 cells (d) by FISH using probes for telomeric/centromeric DNA. The results show a large number of extrachromosomal telomeric repeats (arrows; Ref. 26) as well as marked heterogeneity in telomeric lengths in VA13 cells, which use ALT mechanisms (26), whereas such changes are absent in ML10 cells.

remained diploid thereafter, as attested by flow cytometry studies as DISCUSSION well as by performing chromosome counts in metaphase spreads (data not shown). No appreciable differences in the doubling time could be Our results clearly show that a specific type of genetic instability detected in the two fractions based on growth curve analyses per- associated with numerical chromosomal alterations is an independent formed immediately after the cell sorting procedure (data not shown). determinant of crisis in ovarian cystadenoma cells expressing SV40 TUNEL assay (data not shown) showed evidence of apoptosis in large T-antigen. Such cells underwent crisis after accumulation of 34 Ϯ 5% of the nondiploid cell fraction compared with 15 Ϯ 4% of severe changes in their DNA ploidy and before extensive telomere the diploid cell fraction (measured by determining the number of attrition had taken place. This lack of significant telomere attrition in apoptotic cells/100 cells/field in four separate fields), suggesting that the period preceding crisis was not due to ALT mechanisms. Sepa- the selective disadvantage of nondiploid cells was due, at least in part, ration of cells that had undergone severe ploidy changes from those to increased apoptosis, which is one of the hallmarks of crisis. that had remained diploid using FACS showed that crisis occurred The fact that telomerase-transfected aneuploid cells were outgrown earlier in the nondiploid fraction, despite similar telomere lengths in by the initially small proportion of diploid cells suggests that ectopic all fractions. Recovery from ploidy-dependent crisis as well as from expression of telomerase could not immortalize cells harboring severe this specific type of genetic instability did not depend on telomerase changes in DNA ploidy, although it was sufficient to immortalize cells expression because MCV500 cells, which underwent two separate with normal DNA content or with less severe ploidy changes. Sub- crisis events, did not express this enzyme after recovery from the first clonal populations of sorted telomerase-transfected cells were ob- event, although their ploidy status had become stable. It is only after tained and examined for their in vitro longevity to verify this hypoth- recovery from the second crisis event and after further telomere esis. Most of the subclones recovered from the sorted aneuploid shortening, 35 PDs later, that induction of telomerase expression was fraction were diploid, further attesting to the selective advantage of observed in those cells. Expression of telomerase in ovarian cystade- such cells over the more abundant aneuploid cells. Two aneuploid noma cells expressing SV40 large T-antigen could confer in vitro subclones, containing 61 and 109 chromosomes, respectively, were immortality only to cells that had recovered or escaped from ploidy- cultured in parallel with an age-matched diploid subclone. Telomerase dependent crisis. We conclude that there are at least two determinants expression was verified in all three subclones by the TRAP assay of crisis in this epithelial cell culture system: one that is ploidy (data not shown). The two aneuploid subclones started showing signs dependent (pdM2); and another that is telomere dependent (tdM2). of crisis after an estimated 20–25 PDs following the cloning proce- These results are summarized in the diagram shown in Fig. 6, which dures (Fig. 5c). Such changes were noticed 3–5 PDs earlier in the illustrates the independent contributions of alterations in DNA ploidy subclone with 109 chromosomes compared with the one with 61 and telomere attrition to the crisis phenomenon. chromosomes (data not shown). The aneuploid subclones could not be Our results emphasize the distinction between genetic instability propagated further after this experiment, attesting to the fact that they associated with structural versus numerical chromosomal alterations, had not been immortalized by transfection with the catalytic subunit both of which are hallmarks of cancer. An association between struc- of telomerase. tural abnormalities and telomere attrition and a protective role of 5817

with abnormally replicated DNA or missegregated chromosomes (19, 38–40). Defects in both p53 (20, 21) and RB (20) could also lead to aneuploidy due to the role of these proteins in monitoring normal centrosome duplication during mitosis. Although we have not sys- tematically investigated the molecular determinants of ploidy-based crisis, it is possible that the increase in the levels of the p21 protein, which parallels changes in DNA ploidy, may have been important in mediating this phenomenon. Although the existence of p53-independent mechanisms for p21 induction is well established, the elevated levels of p53 in cells carrying severe ploidy changes may have been sufficient to overwhelm the SV40 large T-antigen, resulting in p21 protein induction. Ploidy changes may be perceived as a type of DNA damage by the cellular machinery, resulting in growth arrest through telomere-independent mechanisms. A role for the p53 and p21 proteins in crisis was suggested recently by Romanov et al. (32). This conclusion is also supported by earlier results from Kiyono et al. (41), who showed that expression of telomerase could not immortalize cells transfected with E6, which neutralizes p53 and often results in ploidy changes, but led to immortalization in cells transfected with E7, in which p53 function is intact, and ploidy status usually remains stable. Although coexpression of both E6 and E7 in the presence of telomerase resulted in immortalization, the ploidy status of the cells was not examined in this study (41), raising the possibility that immortalization had taken place only in cells that had remained diploid or near diploid. Premature p16-mediated growth arrest independent of telomere attrition has been documented in presenescent epithelial cells and keratinocytes lacking essential nutrients (42, 43), as well as in cells treated with agents that cause DNA double-strand breaks (44). It Fig. 5. Consequences of telomerase expression in diploid versus nondiploid cells. is unlikely that these mechanisms played an important role in medi- ML10 cells transfected with an expression vector for the catalytic subunit of telomerase were separated into diploid and nondiploid fractions by FACS. a and b, the DNA content ating ploidy-dependent growth arrest in our cell culture system be- of the nondiploid fraction was analyzed by flow cytometry 3 (a) and 35 (b) PDs after the cause the absence of a functional RB protein deprived p16 of its cell sorting procedures. The results show that the aneuploid cells had been completely downstream substrate. In addition, the fact that our cultured SV40 overgrown by the diploid cells by the time this mixed cell population had reached 35 PDs. c, one diploid and two aneuploid subclones, the latter containing 61 and 109 chromo- large T-antigen-expressing cystadenomas can be propagated 40–60 somes, respectively, were isolated from the sorted aneuploid fraction by plating the cells PDs in vitro and that a subset of cells can be immortalized by onto 96-well microtiter plates at a density of 0.5 cell/well. All three clones expressed telomerase, as evidenced by positive TRAP assay results (data not shown). The figure expression of telomerase argues against a growth arrest due to deple- shows growth curve analyses of each clone performed in parallel and initiated approxi- tion of essential nutrients. mately 20–25 PDs after the cloning procedures, at which time the aneuploid clones were Although the exact relevance of the various mortality checkpoints showing signs of crisis. The number of cells/dish was determined using a Coulter counter. Each point represents the average of triplicate dishes. The vertical bars represent SEs. observed in vitro to the development of human cancers is not clear, evidence that human cancers arise from cells that have recovered from the equivalent of in vitro crisis includes the facts that (a) cancer cells telomerase against such changes have been demonstrated in previous studies using normal cells (12), cells carrying p53 mutations (37), and cells transformed with SV40 large T-antigen (22). It has also been suggested that any form of non-telomere-based DNA damage may be more efficiently detected and repaired in cells expressing telomerase (23). Our results suggest that the concept of a role for telomerase in controlling genetic instability should not be generalized to all forms of instability. The apparent increase in ploidy stability seen in cells expressing telomerase may simply reflect the fact that telomerase expression only confers immortality to cells that already acquired such stability by recovering from pdM2, providing them with a selective advantage over genetically more unstable subpopulations. This idea is in complete agreement with earlier results (17), in which human mammary epithelial cells expressing SV40 large T-antigen were observed to contain heterogeneous populations of cells, some of which were chromosomally diploid, and others that were grossly Fig. 6. Model: telomere attrition and abnormalities in DNA ploidy are independent determinants of crisis. We propose, based on our results obtained with SV40 large aneuploid. In this previous study (17), the epithelial cells that escaped T-antigen-expressing ML10 cystadenoma cells, that after bypassing M1 due to overcom- crisis were almost universally chromosomally diploid, suggesting that ing of the action of cell cycle-regulatory proteins such as p53 and p16/RB, cells resume escape from crisis only occurred in this subpopulation of cells. their logarithmic growth until they show manifestations of chromosomal instability leading to alterations in DNA ploidy. When severe, these alterations induce pdM2 The mechanisms responsible for the development of either tetra- characterized by increased apoptosis. Cells that adapt to these ploidy changes or those that ploidy or aneuploidy in cells harboring abnormalities in the p53 never develop changes of high enough severity to induce crisis can resume or sustain logarithmic growth until further telomere attrition leads to a second crisis event, inde- pathway may come from impairment in a postmitotic cell cycle pendent of pdM2, called tdM2. Recovery from tdM2 requires the establishment of a checkpoint that normally prevents endoreduplication and arrests cells mechanism for telomere maintenance such as telomerase activation. 5818

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2003 American Association for Cancer Research. DETERMINANTS OF IN VITRO CRISIS characteristically show shorter telomeres than normal tissues and 14. Artandi, S. E., Chang, S., Lee, S. L., Alson, S., Gottlieb, G. J., Chin, L., and DePinho, almost invariably express telomerase (27, 45, 46); (b) genetically R. A. Telomere dysfunction promotes non-reciprocal translocations and epithelial cancers in mice. Nature (Lond.), 406: 641–645, 2000. engineered mice used as models to study crisis in vivo are the only 15. O’Hagan, R. C., Chang, S., Maser, R. S., Mohan, R., Artandi, S. E., Chin, L., and transgenic mice known to develop a spectrum of tumors similar to that DePinho, R. A. Telomere dysfunction provokes regional amplification and deletion in found in the adult human population (14); and (c) chromosomal cancer genomes. Cancer Cell, 2: 149–155, 2002. 16. Maser, R. S., and DePinho, R. A. Connecting chromosomes, crisis, and cancer. abnormalities associated with early stages of breast cancer are similar Science (Wash. DC), 297: 565–569, 2002. to those present in late-passage human mammary epithelial cells that 17. Shay, J. W., Van Der Haegen, B. A., Ying, Y., and Wright, W. E. The frequency of spontaneously overcome senescence in culture (32). Our results, immortalization of human fibroblasts and mammary epithelial cells transfected with SV40 large T-antigen. Exp. Cell Res., 209: 45–52, 1993. which emphasize the role of numerical chromosomal alterations in the 18. Wei, W., and Sedivy, J. M. Differentiation between senescence (M1) and crisis (M2) crisis phenomenon, may therefore have implications for our under- in human fibroblast cultures. Exp. Cell Res., 253: 519–522, 1999. standing of the mechanism of acquisition of such numerical alter- 19. Cross, S. M., Sanchez, C. A., Morgan, C. A., Schimke, M. K., Ramel, S., Idzerda, R. L., Raskind, W. H., and Reid, B. J. A p53-dependent mouse spindle checkpoint. ations in human cancers. According to the classical theories of tumor Science (Wash. DC), 267: 1353–1356, 1995. progression first advanced by Foulds (47) and Nowell (48), the 20. Duensing, S., and Munger, K. Human papillomaviruses and centrosome duplication profound genomic alterations that characterize cancer cells occur errors: modeling the origins of genomic instability. Oncogene, 21: 6241–6248, 2002. 21. Fukasawa, K., Choi, T., Kuriyama, R., Rulong, S., and Vande Woude, G. F. Abnor- randomly, followed by elimination of changes that are detrimental to mal centrosome amplification in the absence of p53. Science (Wash. DC), 271: cell survival and selection of those associated with a growth advan- 1744–1747, 1996. tage. Our results illustrate the fact, supported by a number of studies 22. Zhu, J., Wang, H., Bishop, J. M., and Blackburn, E. H. Telomerase extends the lifespan of virus-transformed human cells without net telomere lengthening. Proc. (5, 29–31, 49) as well as from conclusions based on computer Natl. Acad. Sci. USA, 96: 3723–3728, 1999. modeling (50), that aneuploidy, although normally associated with 23. Sherr, C. J., and DePinho, R. A. Cellular senescence: mitotic clock or culture shock? cells that have undergone malignant transformation, can occur before Cell, 102: 407–410, 2000. 24. Luo, M. P., Gomperts, B., Imren, S., DeClerck, Y. A., Ito, M., Velicescu, M., Felix, such transformation and even before acquisition of replicative immor- J. C., and Dubeau, L. Establishment of long-term in vitro cultures of human ovarian tality, at least in the presence of a permissive genetic background such cystadenomas and LMP tumors and examination of their spectrum of expression of as in cells harboring a p53 mutation. The contribution and importance matrix-degrading proteinases. Gynecol. Oncol., 67: 277–284, 1997. of chromosomal changes acquired through these mechanisms to the 25. Buick, R. N., Pullano, R., and Trent, J. M. Comparative properties of five human ovarian adenocarcinoma cell lines. Cancer Res., 45: 3668–3676, 1985. establishment of the cancer genome as well as to the cancer phenotype 26. Ford, L. P., Zou, Y., Pongracz, K., Gryaznov, S. M., Shay, J. W., and Wright, W. E. itself remain to be investigated. Telomerase can inhibit the recombination-based pathway of telomere maintenance in human cells. J. Biol. Chem., 276: 32198–32203, 2001. 27. Kim, N. W., Piatyszek, M. A., Prowse, K. R., Harley, C. B., West, M. D., Ho, P. L., ACKNOWLEDGMENTS Coviello, G. M., Wright, W. E., Weinrich, S. L., and Shay, J. W. Specific association of human telomerase activity with immortal cells and cancer. Science (Wash. DC), We thank Dr. R. A. Weinberg from the Whitehead Institute of Biomedical 266: 2011–2015, 1994. Research for providing us with an expression vector for the telomerase cata- 28. Wan, M., Li, W-Z., Duggan, B., Felix, J., Zhao, Y., and Dubeau, L. Telomerase lytic subunit, Sergei Gryaznov of Geron Corporation for FISH probes, Dr. activity in benign and malignant epithelial ovarian tumors. J. Natl. Cancer Inst. (Bethesda), 89: 437–441, 1997. Axel Scho¨nthal from University of Southern California for critical review of 29. Chang, P. L., Gunby, J. L., Tomkins, D. J., Mak, I., Rosa, N. E., and Mak, S. the manuscript, and Dr. Daniel Weisenberger for assistance in preparing the Transformation of human cultured fibroblasts with plasmids carrying dominant se- manuscript. We also thank the FACS Core of the USC/Norris Comprehensive lection markers and immortalizing potential. Exp. Cell Res., 167: 407–416, 1986. Cancer Center for help with the cell sorting procedures. 30. Ray, F. A., Peabody, D. S., Cooper, J. L., Cram, L. S., and Kraemer, P. M. SV40 T antigen alone drives karyotype instability that precedes neoplastic transformation of human diploid fibroblasts. J. Cell. Biochem., 42: 13–31, 1990. REFERENCES 31. Stewart, N., and Bacchetti, S. Expression of SV40 large T antigen, but not small t 1. Hayflick, L., and Moorhead, P. S. The serial cultivation of human diploid cell strains. antigen, is required for the induction of chromosomal aberrations in transformed Exp. Cell Res., 25: 585–621, 1961. human cells. Virology, 180: 49–57, 1991. 2. Harley, C. B., Futcher, A. B., and Greider, C. W. Telomeres shorten during ageing of 32. Romanov, S. R., Kozakiewicz, B. K., Holst, C. R., Stampfer, M. R., Haupt, L. M., and human fibroblasts. Nature (Lond.), 345: 458–460, 1990. Tlsty, T. D. Normal human mammary epithelial cells spontaneously escape senes- 3. Levy, M. Z., Allsopp, R. C., Futcher, A. B., Greider, C. W., and Harley, C. B. cence and acquire genomic changes. Nature (Lond.), 409: 633–637, 2001. Telomere end-replication problem and cell aging. J. Mol. Biol., 225: 951–960, 1992. 33. Dimri, G. P., Lee, X., Basile, G., Acosta, M., Scott, G., Roskelley, C., Medrano, E. E., 4. Olovnikov, A. M. A theory of marginotomy. The incomplete copying of template Linskens, M., Rubelj, I., Pereira-Smith, O., Peacocke, M., and Campisi, J. A bio- margin in enzymic synthesis of polynucleotides and biological significance of the marker that identifies senescent human cells in culture and in aging skin in vivo. Proc. phenomenon. J. Theor. Biol., 41: 181–190, 1973. Natl. Acad. Sci. USA, 92: 9363–9367, 1995. 5. Counter, C. M., Avilion, A. A., LeFeuvre, C. E., Stewart, N. G., Greider, C. W., 34. Bryan, T. M., and Reddel, R. R. Telomere dynamics and telomerase activity in in Harley, C. B., and Bacchetti, S. Telomere shortening associated with chromosome vitro immortalized human cells. Eur. J. Cancer, 33: 67–73, 1997. instability is arrested in immortal cells which express telomerase activity. EMBO J., 35. Henson, J. D., Neumann, A. A., Yeager, T. R., and Reddel, R. R. Alternative 11: 1921–1929, 1992. lengthening of telomeres in mammalian cells. Oncogene, 21: 598–610, 2002. 6. Itahana, K., Dimri, G. P., Hara, E., Itahana, Y., Zou, Y., Desprez, P. Y., and Campisi, 36. Yeager, T. R., Neumann, A. A., Englezou, A., Huschtscha, L. I., Noble, J. R., and J. A role for p53 in maintaining and establishing the quiescence growth arrest in Reddel, R. R. Telomerase-negative immortalized human cells contain a novel type of human cells. J. Biol. Chem., 277: 18206–18214, 2002. promyelocytic leukemia (PML) body. Cancer Res., 59: 4175–4179, 1999. 7. Hara, E., Tsurui, H., Shinozaki, A., Nakada, S., and Oda, K. Cooperative effect of 37. Elmore, L. W., Turner, K. C., Gollahon, L. S., Landon, M. R., Jackson-Cook, C. K., antisense-Rb and antisense-p53 oligomers on the extension of life span in human and Holt, S. E. Telomerase protects cancer-prone human cells from chromosomal diploid fibroblasts, TIG-1. Biochem. Biophys. Res. Commun., 179: 528–534, 1991. instability and spontaneous immortalization. Cancer Biol. Ther., 1: 391–397, 2002. 8. Shay, J. W., Pereira-Smith, O. M., and Wright, W. A role for both RB and p53 in the 38. Khan, S. H., and Wahl, G. M. p53 and pRb prevent rereplication in response to regulation of human cellular senescence. Exp. Cell Res., 196: 33–39, 1991. microtubule inhibitors by mediating a reversible G arrest. Cancer Res., 58: 396–401, 9. de Lange, T. Activation of telomerase in a human tumor. Proc. Natl. Acad. Sci. USA, 1 1998. 91: 2882–2885, 1994. 39. Lanni, J. S., and Jacks, T. Characterization of the p53-dependent postmitotic check- 10. Greider, C. W., and Blackburn, E. H. Telomeres, telomerase and cancer. Sci. Am., 274: 92–97, 1996. point following spindle disruption. Mol. Cell. Biol., 18: 1055–1064, 1998. 11. Lustig, A. J. Crisis intervention: the role of telomerase. Proc. Natl. Acad. Sci. USA, 40. Notterman, D., Young, S., Wainger, B., and Levine, A. J. Prevention of mammalian 96: 3339–3341, 1999. DNA reduplication, following the release from the mitotic spindle checkpoint, re- 12. Bodnar, A. G., Ouellette, M., Frolkis, M., Holt, S. E., Chiu, C. P., Morin, G. B., quires p53 protein, but not p53-mediated transcriptional activity. Oncogene, 17: Harley, C. B., Shay, J. W., Lichtsteiner, S., and Wright, W. E. Extension of life-span 43–51, 1998. by introduction of telomerase into normal human cells. Science (Wash. DC), 279: 41. Kiyono, T., Foster, S. A., Koop, J. I., McDougall, J. K., Galloway, D. A., and 349–352, 1998. Klingelhutz, A. J. Both Rb/p16INK4a inactivation and telomerase activity are re- 13. Chin, L., Artandi, S. E., Shen, Q., Tam, A., Lee, S. L., Gottlieb, G. J., Greider, C. W., quired to immortalize human epithelial cells. Nature (Lond.), 396: 84–88, 1998. and DePinho, R. A. p53 deficiency rescues the adverse effects of telomere loss and 42. Ramirez, R. D., Morales, C. P., Herbert, B. S., Rohde, J. M., Passons, C., Shay, J. W., cooperates with telomere dysfunction to accelerate carcinogenesis. Cell, 97: 527–538, and Wright, W. E. Putative telomere-independent mechanisms of replicative aging 1999. reflect inadequate growth conditions. Genes Dev., 15: 398–403, 2001. 5819

43. Ramirez, R. D., Herbert, B. S., Vaughan, M. B., Zou, Y., Gandia, K., Morales, C. P., Lingner, J., Harley, C. B., and Cech, T. R. Telomerase catalytic subunit homologs Wright, W. E., and Shay, J. W. Bypass of telomere-dependent replicative senescence from fission yeast and human. Science (Wash. DC), 277: 955–959, 1997. (M1) upon overexpression of Cdk4 in normal human epithelial cells. Oncogene, 22: 47. Foulds, L. The histologic analysis of mammary tumors in mice. II. The histology of 433–444, 2003. responsiveness and progression. J. Natl. Cancer Inst. (Bethesda), 17: 713–752, 1956. 44. Robles, S. J., and Adami, G. R. Agents that cause DNA double strand breaks lead to 48. Nowell, P. C. The clonal evolution of tumor cell populations. Science (Wash. DC), p16INK4a enrichment and the premature senescence of normal fibroblasts. Onco- 194: 23–28, 1976. gene, 16: 1113–1123, 1998. 49. Barrett, M. T., Sanchez, C. A., Prevo, L. J., Wong, D. J., Galipeau, P. C., Paulson, 45. Meyerson, M., Counter, C. M., Eaton, E. N., Ellisen, L. W., Steiner, P., Caddle, S. D., T. G., Rabinovitch, P. S., and Reid, B. J. Evolution of neoplastic cell lineages in Ziaugra, L., Beijersbergen, R. L., Davidoff, M. J., Liu, Q., Bacchetti, S., Haber, D. A., Barrett oesophagus. Nat. Genet., 22: 106–109, 1999. and Weinberg, R. A. hEST2, the putative human telomerase catalytic subunit gene, is 50. Tsao, J. L., Tavare, S., Salovaara, R., Jass, J. R., Aaltonen, L. A., and Shibata, D. up-regulated in tumor cells and during immortalization. Cell, 90: 785–795, 1997. Colorectal adenoma and cancer divergence. Evidence of multilineage progression. 46. Nakamura, T. M., Morin, G. B., Chapman, K. B., Weinrich, S. L., Andrews, W. H., Am. J. Pathol., 154: 1815–1824, 1999.

5820

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2003 American Association for Cancer Research. Aneuploidy and Telomere Attrition Are Independent Determinants of Crisis in SV40-transformed Epithelial Cells

Mihaela Velicescu, Jiamei Yu, Brittney-Shea Herbert, et al.

Cancer Res 2003;63:5813-5820.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/63/18/5813

Cited articles This article cites 48 articles, 18 of which you can access for free at: http://cancerres.aacrjournals.org/content/63/18/5813.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/63/18/5813.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/63/18/5813. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.