Synchronization of Tumor and Normal Cells from G, to Multiple Cell Cycles by Lovastatin1

ICANCER RESEARCH 51. 3602-3609. July 1. 1991] Synchronization of Tumor and Normal Cells from G, to Multiple Cell Cycles by Lovastatin1

Khandan Keyomarsi,2 Larue Sandoval, Minia Band, and Arthur B. Pardee

Divisions of Cell Growth and Regulation [K. K., L. S., A. B. P.I and Cancer Genetics [\. B.], Dana Farber Cancer Institute, Boston, Massachusetts 02115

ABSTRACT nization by serum-starvation or contact inhibition is often used to arrest the growth of nontumorigenic diploid fibroblast cells Synchronization of mammalian cells is essential for investigations involving cell proliferation. A simple method for obtaining synchrony in in Go or quiescence. In tumor or transformed cells, however, all types of cells, through several cycles and with minimal overall proliferation is relatively independent of the presence of serum metabolic perturbations, has not yet been available. We describe a or growth factors (5, 6). As a result, neoplastic cells, for the procedure for synchronizing normal as well as tumor cells reversibly in most part, do not cease proliferation upon serum deprivation, the GI phase of the cell cycle using Lovastatin, an inhibitor of 3-hydroxy- nor do they display stringent density-dependent G0 arrest (7). 3-methylglutaryl-coenzyme A reducÃase. This method of synchronization Furthermore, although synchronization by serum starvation was successful with all cell lines tested, including normal and tumor cells may be an appropriate method for studying the events that of mouse, hamster, and human origins. For example, when MCF-7 human occur during the transition from the resting to the growing breast cancer cells were synchronized with Lovastatin and released by state, it also subjects the cultures under study to long periods the addition of mevalonic acid (the product of the reaction catalyzed by 3-hydroxy-3-methylglutaryl-coenzyme A reducÃase), 3 phases of accel of incubation in the absence of essential nutrients and growth erated ih\rumino incorporation into DNA corresponding to 3 S phases of factors. As a result, the cells are metabolically imbalanced and the cell cycle occurred during a 90-h period of cell replication. Thymidine out of cycle, thus inappropriate for analyzing the cell cycle incorporation was decreased to â€¢-4'; during the initial lag of 18 h before events in continually dividing cells. These cells are generally the first S phase, and maximum incorporation was then achieved after poorly synchronized after the initial cycle. only 6 h. The antibody Ki-67, which detects a nuclear antigen associated To develop a procedure effective for synchronizing normal with proliferation, was present in cells arrested with Lovastatin. This and neoplastic populations, the following criteria should be fact, together with the lack of thymidine incorporation during the initial met: (a) both normal and tumor cells should be arrested at the lag time, indicates that the cells were arrested in the d and not in the Co phase of the cell cycle. Furthermore, in synchronized tumor-derived same, specific cell cycle phase; (b) the synchronization must be reversible and noncytotoxic; (c) the metabolic block should human breast epithelial cells, histone H4 RNA was low after Lovastatin release and increased with the onset of DNA synthesis. Concomitant target a specific reaction and must be reversible; (d) large synthesis of DNA and histone H4 RNA expression could be observed quantities of synchronous cells should be obtained; (e) the for 2 cycles. Minimal perturbations of general metabolic functions oc synchronization must be density and medium independent; and curred since the rate of RNA, protein, and initial DNA synthesis were (/) synchrony should be for more than one cycle, rendering it unaffected by Lovastatin, as evidenced by |'H|uridine, ['HJIeiidne, and particularly useful for studying the biochemical processes that initial | "II|ih> micline incorporation. Finally, while the Lovastatin-induced occur in cycling cells as well as immediately after the initial synchronization was overcome by mevalonic acid, addition of squalene or arrest. cholesterol-ethanol had no such effect. Thus, Lovastatin appears to In this paper, we report that these criteria for synchronization prevent formation of an early intermediate in the cholesterol pathway that is essential for progression of cells through early (., phase. are met using Lovastatin to arrest normal and tumor cells of epithelial as well as fibroblast origin reversibly in the GI phase of the cell cycle. Lovastatin, an antihyperlipodemic agent, is INTRODUCTION widely used for the treatment of hypercholesterolemia. This agent competitively inhibits HMG-CoA' reducÃase,the enzyme The study of cellular, biochemical, and molecular mecha nisms that regulate cell division in normal versus tumor cells is required for conversion of HMG-CoA to mevalonic acid (8). In a major topic of cancer research. For this purpose, large quan addition, inhibition of mevalonate synthesis by Lovastatin or tities of highly synchronized populations of both cell types are compactin, a related fungal toxin (9), also inhibits DNA repli necessary to localize sites of proliferation control. Cell prolif cation (10-15). DNA replication is restored by the addition of eration is stringently regulated in normal cells but is deranged mevalonate, the product of the reaction catalyzed by HMG- to varying degrees in cancer cells. Regulation occurs upon CoA reducÃase. escape from quiescence and in the G, phase of the cell cycle We report here that inhibition of mevalonate synthesis by before a special restriction point event, which occurs a few Lovastatin synchronizes a wide array of normal and tumor cell hours before the onset of DNA replication (1,2). lines of mouse, hamster, and human origins in early G,. This Although a number of procedures have been developed for synchronization is noncytotoxic and fully reversible when mev establishing synchronous cultures of mammalian cells (3, 4), alonate is replenished. Rales of RNA and protein synthesis are many of these protocols have not been successfully adapted for not significantly altered by Lovastatin treatment. In addition, use with tumor as well as normal cells. For example, synchro- the synchrony can be followed for at least one full cell cycle and in some cases up to 3 cell cycles. Ki-67 antigen is present Received4/2/91;accepted4/17/91. The costs of publication of this article were defrayed in part by the payment in cells synchronized by Lovastalin bui noi in cells synchronized of page charges. This article must therefore be hereby marked advertisement in by serum slarvalion, suggesling a G, and noi a G0 arrest. This accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This work was supported in part by Grant CA 22427 (A. B. P.) and method of cell synchrony provides a high yield of synchronous S07RR05526-27 (a Basic Research Support Grant to K. K.) from the NIH. K. K. is a recipient of National Research Service Award CA08949-01 from the NIH. 'The abbreviations used are: HMG-CoA, .Vhydroxy-.Vmethylglutaryl coen- 2To whom requests for reprints should be addressed, at Dana Farber Cancer zyme A; PBS. phosphate-buffered saline; FCS, fetal calf serum; MEM, minimal Institute. Division of Cell Growth and Regulation. 44 Binney Street. Room D- essential medium: HEPES, ,V-[2-hydroxyethyl]piperazine-A"-[2-ethanesulfonic acid); DMEM, Dulbecco's modified Eagle's medium. 810A. Boston. MA 02115. 3602

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1991 American Association for Cancer Research. SYNCHRONIZATION IN G, BY LOVASTATIN populations of cells, so that comparative studies between nor fibroblast cells were obtained from American Type Culture Collection mal and tumor cells can be performed. and grown in Â«-MEMsupplemented with 10% FCS, 2 mM glutamine. 0.1 m.Mnonessential amino acids, and 10 mM HEPES buffer. All cells were grown at 37Â°Cin a humidified incubator containing 6.5% CO^. MATERIALS AND METHODS except A31 and BPA/31, which require 10% COi. All cultures were maintained free of Mycoplasma as determined by the MycoTect Kit Materials (Gibco). (nipf/i.r/-['H]Thymidine (81 Ci/mmol), L-[4,5-'H)leucine (53 Ci/ mmol), [5-'H]uridine (27 Ci/mmol), and [Â«-'2P]CTP(3000 Ci/mmol) Determination of Population Doubling Time were purchased from New England Nuclear, Boston, MA. Lovastatin Approximately 5 x IO4 cells were plated per 35-mm dish (Falcon) was kindly provided by A. W. Alberts of Merck. Sharp and Dohme and grown at 37Â°Cin a humidified incubator. At 24-h intervals, cells Research Pharmaceuticals, Rahway, NJ. Before addition to cultures. from triplicate dishes were washed once with PBS, trypsinized, and Lovastatin was converted from its inactive lactone prodrug form to its counted in a Coulter Counter. The population doubling times were active dihydroxy open acid form by first dissolving 52 mg of the estimated from the linear portion of the growth curve as described compound in 1.04 ml of ethanol (95%) and then adding 813 ftl of l N previously (22). NaOH. The resulting solution was neutralized with l N HC1 to a pH of 7.2 and brought up to a volume of 13 ml with either distilled water or medium, and stored in multiple aliquots at -20Â°C until use. The Cell Synchronization final concentration was 10 HIM. Mevalonic acid lactone, squalene, Cells were plated in 24-well plates (Falcon) at 1 x 10" cells/well, and cholesterol, low density lipoprotein, and serum were purchased from 3 types of protocols were then followed: Sigma Chemical Co., St. Louis, MO. and cell culture medium was from Synchronization by Lovastatin Treatment. This method was applied Gibco Laboratories (Grand Island, NY). All other chemicals used were to all cell lines used in this study. Medium was removed 36-48 h after of reagent grade. The scintillation cocktail used was Budget-Solve from the initial plating and replaced with fresh medium containing Lovas Research Products International (Mount Prospect, IL). tatin at concentrations and for incubation times listed in Table 1. Under these conditions, cells become arrested in G, as determined by [JH]- Cells and Culture Conditions thymidine incorporation and the positive staining of cells (human) with Table 1 lists all the cell lines used in this study. Murine BALB/c Ki-67 antibody (see below). At time 0 h, the medium was removed and 3T3 (A31) cells (16) and benzo[a]pyrene-transformed BALB/c 3T3 replaced with fresh medium containing mevalonic acid at a concentra (BPA31) cells (17) were grown in DMEM supplemented with 4 mM tion 100 times the Lovastatin concentrations used. Cells were harvested glutamine, 10 mM HEPES buffer (pH 7.3), and 10% bovine calf serum. at the indicated times, and DNA synthesis and cell density were The diploid Chinese hamster embryo fibroblast cell line CHEF/18, measured. which is a derivative of the F4224A cell line isolated from a male Synchronization of A31 and WI38 Cells by Serum Deprivation. At 72 Chinese hamster embryo (18), and the ethylmethane sulfonate trans h after plating of cells, medium was removed, cells were washed 3 times formed T30-4 cell line (19) were grown in DMEM supplemented with with medium alone, and A31 cells were incubated with DMEM con 4 mM glutamine, 10 mM HEPES buffer, and 10% fetal calf serum. The taining 8 HIMglutamine and 0.5% calf serum for 72 h; WI38 cells were normal human mammary epithelial cell strain, 70N, was derived from incubated in Â«-MEMcontaining 2 mM glutamine, 0.1 mM nonessential reduction mammoplasty and was grown in DFCI-1 medium as de amino acids, and 0.2% fetal calf serum for 48 h. Under these conditions, cells become arrested in Go as determined by ['H]thymidine incorpo scribed (20). 2INT and 21PT were mastectomy samples from a patient with infiltrating ductal and intraductal carcinoma (21) and were grown ration and lack of Ki-67 antibody staining (see below). At time 0 h, the in the DFCI-1 medium and also in n-MEM supplemented with 10% low serum medium was removed and the cells were stimulated by the FCS, 1 Mg/ml of insulin, 12.5 ng/ml epidermal growth factor, 2.8 /Â¿M addition of medium containing 10% serum. Cells were harvested at the hydrocortisone, 2 mM glutamine, 1 mM sodium pyruvate, 0.1 mM indicated time intervals after addition of complete medium. nonessential amino acids, and 10 mM HEPES buffer. MCF-7 cells were Synchronization of 70N Cells by Growth Factor Deprivation. At 48 h obtained from Michigan Cancer Foundation, Michigan, IL, and cul after plating of cells, medium was removed, and cells were washed 3 tured in Â«-MEMsupplemented with 10% FCS, 1 ng/m\ insulin, 2 mM times with and incubated in DFCI-3 medium for 48 h. DFCI-3 medium glutamine, 1 mM sodium pyruvate, 0.1 mM nonessential amino acids, is DFCI-1 without essential growth factors (21). Under these condi and 10 mM HEPES buffer. HeLa cells and WI38 human diploid tions. 70N cells become quiescent (G0) as determined by [3H]thymidine

Table 1 Conditions of Lovastatin synchronization in a variety of cell lines Cells were subjected to synchronization by Lovastatin. serum deprivation, or growth factor deprivation as described in text. At different times after the addition of fresh medium containing mcvalonate or growth factors, cells were harvested and rates of DNA synthesis and cell density were determined. The pattern of DNA synthesis obtained for each cell line was initially expressed as those in Figs. 1, 2, and 6, and are presented below. DNA to reach of cell of synthesis time(h)IS15151312121215g81210812Timepeak DNA cycles Cells"MCF-721PT21NT70NW138HeLaA-31BP/A-31CHEF-time*(h)27.734.033.433.824.525.5N synchronyLovastatinLovastatinLovastatinLovastatinLovastatinGrowthof incubation20/33 inhibition9696929398999698929091969695Lagsynthesis(h)2422221920241824161423IS1418No.followed322>12<12<123<1221 h20/36 h20/36 h20/36 h15/24h10/24h60/33

factordeprivationLovastatinSerum

scrumÂ«-MEMDMEMLowstarvationLovastatinLovastatinSerum D19.321.022.525.1Median-MEMÂ«-MEMDFCI-1Â«-MEMDFCI-1DFCI-3tr-MEMLowh40/30 h60/36 serumDMEMDMEMDMEMMethodstarvationLovastatinLovastatinLovastatinflMLovastatin/h h20/30 18T-30-4Doubling h40/36 h% 1For a description of these cell lines see "Materials and Methods." The doubling times for each cell line are expressed as an average of 4 values. 3603

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1991 American Association for Cancer Research. SYNCHRONIZATION IN G, BY LOVASTATIN incorporation and lack of Ki-67 antibody staining (see below). At time 0 h, cells were stimulated by the addition of DFCI-1 medium. Cells were harvested at the indicated times after addition of DFCI-1 medium.

Synthesis Rates of DNA, RNA, and Protein

Relative synthesis rates of DNA, RNA, or protein were estimated by measuring the incorporations of [3H]thymidine. [3H]uridine, or [3H]- leucine, respectively (these measurements will include transport rates through cell membranes as well as rates of incorporation). For each determination, monolayers of cells were pulsed with 5 MCi/ml of the radioisotope for l h at 37Â°Cin medium lacking thymidine and hypo- xanthine. Incorporation was stopped by the addition of ascorbic acid (final concentration: 67 ITIM),and cells were washed twice with PBS and incubated in the presence of 5% trichloroacetic acid for l h at 37Â°C.Monolayers were dissolved in l N NaOH and neutralized with 0 12 24 36 48 60 72 84 96 glacial acetic acid. A clear counting mixture was obtained by the Time, hours addition of 10 ml of scintillation cocktail, and radioactivity was deter Fig. 1. Synchronization of MCF-7 cells by Lovastatin can be followed for mined with correction for quenching by the H-number method (23). multiple cycles. Exponentially growing MCF-7 cells were plated on 24-well plates For each experimental condition, parallel wells were used for determi for 36 h. Cells were then cultured in medium containing 20 >iMLovastatin for 33 h. Fresh medium containing 2 mM mevalonic acid was then added at zero time nation of cell number using a Coulter Counter (model ZF; Coulter (abscissa). Cells were harvested at the indicated time intervals and DNA synthesis Electronics, Inc., Hialeah, FL). Rates of synthesis are expressed in dpm and cell density were measured. Rates of DNA synthesis are expressed in dpm/ per cell under the conditions of the assay. cell under the conditions of the assay. Each data point was performed in duplicate, and the experiment was repeated 4 times. Ki-67 Immunohistochemistry Cells (6 x 10') were plated per chamber of 4 chamber slides. After synchronization, the cells were washed twice with PBS and fixed with Turn methanohacetone (1:1) for 1 min at room temperature. Fixed cells were "starvation then rinsed 3 times in PBS and incubated with fluorescein isothiocyanate-conjugated monoclonal mouse anti-human Ki-67 nuclear an tigen diluted 1:10 with PBS (Dako, Irvine, CA) for l h at room temperature. After 3 washes with PBS, the slides were mounted with p-phenylenediamine as an anti-fader as described (24). Microscopic evaluation of stained cells was carried out using a Zeiss fluorescence microscope.

Northern Blot Analysis

Total cellular RNA was extracted from 70N and 21PT cells by 12 24 36 guanidinium isothiocyanate and subjected to cesium chloride gradient Time, hours purification as described (25). For Northern blot analysis, 20 ^g of total Fig. 2. Synchronization by Lovastatin versus serum starvation. Exponentially RNA were fractionated under denaturing conditions on a 1.2% agarose/ growing mouse fibroblast A31 cells were plated on duplicate 24-well plates for 36 0.66 M formaldehyde gel and transferred to a Nytran filter (Schleicher h. Cells on one set of plates (â€¢)werethen cultured in medium containing 40 i;\i & Schnell) for subsequent hybridization as described (26). Lovastatin for 30 h. Fresh medium containing 4 m\i mevalonic acid was then added at zero time (abscissa). Cells on the second set of plates (D) were cultured Probe Preparation in medium containing 0.49J serum for 72 h. Fresh medium containing 10?c serum was then added at time zero. Cells from each plate were harvested at the indicated The DNA probes were prepared by random-primed labeling (27) time intervals and DNA synthesis and cell density were measured. Rates of DNA (Boehringer Mannheim). The Clal/EcoRl 1.4-kilobase fragment from synthesis are expressed in dpm per cell under the conditions of the assay. Each data point was performed in duplicate, and the experiment was repeated 3 times. pmycB122 (28) corresponding to the third exon of c-myc, the Â£c0RI/ Pstl 1.5-kilobase fragment from pH4 cDNA clone (29). and the Pstl 800-base pair fragment from p36B4 cDNA clone (30) were all labeled arrested cultures followed by the addition of 2 mM mevalonate with [Â«-"PjdCTP to a specific activity of 1 x IO9dpm/^g of DNA. resulted in the synchronous recovery of DNA synthesis. DNA synthesis resumed after a 15-18-h lag period. During the 90-h RESULTS period of cell replication, 3 phases of accelerated thymidine incorporation occurred at 24-26-h intervals corresponding to Synchronization of Tumor and Normal Cells by Lovastatin 3 successive S phases of the cell cycle. The periodic DNA Can Be Followed for Multiple Cycles. In view of the involvement synthesis that occurred at 24, 48, and 72 h suggested no extra of Lovastatin in inhibition of DNA replication and cell growth, delay in the first cycle, implying that growth arrest occurred we explored the possibility of inducing a GI synchrony in several early in the G, phase of the cell cycle. DNA synthesis in this cell types by blockage of the mevalonate production in these cell line reached a maximum (100%) at 24 h after mevalonate cells, followed by its replenishment. Figs. 1 and 2 demonstrate was replenished, decreased to 21% at 39 h, and increased again that Lovastatin could indeed synchronize tumor and normal at a time corresponding to a second S-phase peak, 48 h after cells in G,. Furthermore, this synchrony was reversible and removal of the Lovastatin block. could be followed for multiple cycles. With MCF-7 cells (and mouse fibroblast A31 cells; see Fig. Exposure of MCF-7, human breast epithelial cancer cell line, 2), ['H]thymidine incorporation was progressively reduced with to 20 MMLovastatin for 33 h resulted in over 95% inhibition of each cycle, suggesting that the cells gradually escape synchrony DNA synthesis as measured by incorporation of [3H]thymidine over time. This loss of synchrony may be due to large variations into DNA (Fig. 1). Removal of Lovastatin from the growth- in the rate of cycle traverse by individual cells during the 3604

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1991 American Association for Cancer Research. SYNCHRONIZATION IN G, BY LOVASTATIN postsynchronization period. This phenomenon was more ap in Fig. 3. During the 33-h Lovastatin treatment (indicated by parent in cells that were synchronized by serum starvation (Fig. negative time values; â€”33h to 0 h). there was no effect of 2, Table 1) (31.32). Lovastatin on the rate of RNA and protein synthesis or initial Generality of Lovastatin Synchronization. This type of syn DNA synthesis as compared with the final DNA synthesis, chrony was not limited to human tumor cells. For example, which was substantially (96^) inhibited after the first S phase after mouse nontransformed fibroblast cell lines (A31) were was complete. When mevalonate was restored to the culture incubated with 40 /uMLovastatin for 30 h. over 90cc of DNA medium, there was the expected lag time of 12-15 h before replication was inhibited (Fig. 2). After the removal of Lovas initiation of DNA synthesis, which proceeded synchronously as tatin and addition of 4 HIMmevalonate, there was a lag time of before. On the other hand, there were only moderate increases 8 h before DNA synthesis was initiated synchronously and in the rates of RNA and protein synthesis (indicated by positive maximum ['H]thymidine incorporation, corresponding to S values; 0-44 h). which eventually returned to initial levels when phase, was reached at 16 h. The synchronization in these cells the cells divided. To account for any alterations in rates of could also be followed for 3 cycles. Table 1 summarizes the uptake of ['HJuridine or ['HJleucine by the cells, total protein conditions of Lovastatin synchronization for the 10 cell lines and RNA were also measured spectrophotometrically at the used in this study. These cell lines were from a variety of species indicated times; the patterns obtained were very similar to those and tissues of origin including normal human diploid lung in Fig. 3, bottom panel (data not shown). From these results, fibroblast cells (VVT38), normal and primary tumor-derived we conclude that the loss in cell cycle progression during the human breast epithelial cells (70N, 21PT and 2INT), normal Lovastatin treatment is independent of inhibited protein. DNA, and transformed mouse (A-31 and BP/A-31) and hamster or RNA biosynthesis, and that general metabolism is minimally (CHEF 18 and T-30-4) fibroblast cells, human cervical carci perturbed. noma cell line (HeLa), and a pleural effusion-derived breast Ki-67 Antibody Staining in GÂ«versusG, Cells. The monoclo cancer cell line (MCF-7). All cell lines showed marked DNA nal antibody, Ki-67, reacts with a human nuclear antigen pres synthesis inhibition (90-98%) by Lovastatin treatment ranging ent in all cycling (G,, S, G2, and M) cells, but is absent in serum in concentration from 10 to 60 /J.Mand time of incubation from starved quiescent (G0) cells (33, 34). Thus, with this antibody 24 to 36 h. In all cases, DNA synthesis was completely restored we can determine whether cells that were synchronized by by the addition of 100-fold excess mevalonate (1-6 m\i) as Lovastatin were arrested in G0 (no Ki-67 staining) or G, (Ki- compared with Lovastatin concentrations used (10-60 //M)- 67 staining), as compared with cells that were synchronized in Lovastatin arrest was therefore reversible. The pattern of [3H]- Go by growth factor deprivation. Immunostaining with Ki-67 thymidine incorporation obtained was indicative of similar antibody revealed that cells were arrested in G,>by growth factor synchrony among all cells tested, and therefore confirmed the deprivation since they were not stained with this antibody (Fig. generality of synchronization by Lovastatin. 4A). However. 24 h after the growth factor stimulation, bright One of the salient features of Lovastatin synchrony is that it fluorescence was detected in these cells suggesting reentry into was independent of tissue culture medium. For example, when the cell cycle (Fig. 4B). On the other hand, bright nuclear 21PT cells were grown in either a-MEM or DFCI-1 medium fluorescence was seen in cells arrested by Lovastatin at time 0 and subjected to Lovastatin treatment, identical patterns of and 24 h after mevalonate addition (Fig. 4. C and D) and in ['Hjthymidine incorporation were obtained, as indicated by the length of the lag period before DNA synthesis and length of [3HHhymldlnt time to reach peak DNA replication (Table 1). k/ Synchronization by Lovastatin versus Serum Deprivation. To compare the kinetics of reversal of Lovastatin arrest versus 200- serum arrest, A-31 cells were subjected to serum starvation for 72 h (Fig. 2). DNA synthesis was inhibited by over 90/t and 100 these cells are arrested at quiescence or G0. After serum stim S IO ulation there was a 12-h lag before the initiation of DNA O -40 -30 -20-10 O 10 20 30 40 synthesis. Lovastatin arrest, on the other hand, was followed by an 8-h lag period before DNA synthesis, suggesting that cells were arrested in G, rather than G0. Furthermore, the length of (3H].|tucln* time to reach maximum ['H]thymidine incorporation after the 200 lag period was much longer in cells arrested by serum starvation (more than 24 h) than in cells arrested by Lovastatin (16 h). In 100 addition, synchronization by serum starvation could be followed for less than one full cycle, whereas Lovastatin synchrony could -40 -30 -20 -10 0 10 20 30 40 be followed for multiple cycles. This phenomenon was also observed with human diploid fibroblast (\Vi38) cells, which Time, hours were synchronized by serum starvation or Lovastatin treatment, Fig. 3. Lovastatin does not affect RNA. protein, or initial DNA synthesis. and with normal human breast epithelial (70N) cells, which Exponentially growing MCF-7 cells were plated on triplicate 24-well plates for 36 h. Cells were then cultured in medium containing 20 MMLovastatin for 33 h. were synchronized by growth factor deprivation or Lovastatin At indicated times during Lovastatin treatment, shown by negative time values treatment (Table 1). Together, these results suggest that Lovas (i.e., -33 h to 0 h). cells from each plate were harvested and DNA synthesis (['Hjthymidine incorporation. â€¢).RNA synthesis (|3H]uridine incorporation. O), tatin synchronization, unlike serum starvation, can be used for protein synthesis (|'H]leucine incorporation. â€¢).and cell density were measured. studying events in cycling cells during the second or third cycles. Fresh medium containing 2 mM mevalonic acid was then added at zero time Rate of Macromolecule Biosynthesis. The effect of Lovastatin (abscissa). Rates of macromolecule synthesis and cell density Â»eremeasured as before and expressed as percentage of control (time. 0 h). Horizontal line on the treatment on the rates of DNA, RNA. and protein synthesis panel, basal level of ['H]uridine and ['Hjleucine incorporation. Each data point (determined by isotope incorporation) in MCF-7 cells is shown Â»asperformed in duplicate, and the experiment was repeated 3 times. 3605

Fig. 4. Ki-67 stains cells arrested by Lovastatin (d) but not cells arrested by growth factor deprivation (G0). Exponentially growing human normal breast epithelial 70N cells were plated on duplicate 4 chamber slides. Cells on one slide were synchronized by treatment with 15 JIMLovastatin for 24 h. Fresh medium containing 1.5 mivi mevalonic acid was added at time 0 h. Cells on the second slide were cultured in DFCI-3 medium (medium without essential growth factors) for 48 h. Fresh DFC1-1 medium was added at time 0 h. Immunohistochemistry was performed as described in text. Ki-67 was visualized with a fluorescein-conjugated antibody. The width of each photomicrograph represents 230 MITI.A. O hour after growth factor deprivation arrest: B. 24 h after growth factor deprivation arrest; C, 0 h after Lovastatin arrest; D, 24 h after Lovastatin arrest; Â£-//. phase-contrast photomicrographs of the same fields as in A-D, respectively. exponentially growing breast epithelial cells (data not shown). synchrony will be very useful for comparative studies of the These results demonstrate that Lovastatin synchronized cells molecular and biochemical events that occur, respectively, in in the G i and not in the G(1phase of the cell cycle. mid- to late-cycle of both normal and tumor cells. Expression of Cell Cycle Specific mRNAs in Cells Synchro Blockage of Lovastatin Synchrony. To determine which inter nized by Lovastatin. To detect the expression of cell cycle- mediates in the cholesterol biosynthesis pathway can block the regulated genes such as the histone H4 and c-myc mRNAs, synchronization obtained by Lovastatin, exponentially growing Northern blot analysis was performed in cells synchronized by MCF-7 cells were incubated in the presence of Lovastatin alone Lovastatin or growth factor deprivation (Fig. 5). Total RNA and Lovastatin plus different intermediates (Fig. 6). Only mev was isolated from tumor-derived breast epithelial cells (21PT) alonate, the end product of the reaction catalyzed by HMG- at the indicated times after mevalonate supplementation and CoA reducÃase,was capable of blocking the Lovastatin-induced also from growth factor-deprived quiescent normal cells (70N) synchrony when present in conjunction with Lovastatin. Cho that had been restimulated to proliferate by the addition of lesterol, either in the form of cholesterol-ethanol or as low growth factors. In 21PT cells, the c-myc mRNA, an immediate density lipoprotein-cholesterol, did not affect the synchroniza early gene product, was low at time 0 h after the replenishment tion process when added simultaneously with Lovastatin. Also, of mevalonate, and the levels increased rapidly thereafter and squalene, a precursor of cholesterol biosynthesis following far- remained constitutively high during the rest of the cell cycle. nesyl pyrophosphate, did not block the synchronization. Based On the other hand, the histone H4 mRNA, which is an S phase- on these observations, it is conceivable that one or a number of regulated gene product, was not detected until the onset of the the early phosphorylated intermediates in the mevalonate path first and the second S phases in these cells (Fig. 5/4). These way are required to overcome the effects of Lovastatin. times of appearance of the histone H4 message coincided with peaks of DNA synthesis as measured by ['H]thymidine incor DISCUSSION poration and the levels decreased rapidly between the 2 S phases in these cells (Fig. 5, A and B). Cells that were synchronized by One of the major difficulties in studying the cell cycle-related growth factor deprivation, however, could be followed for only events in normal versus malignant cells has been the unavaila one cell cycle; c-myc mRNA increased shortly after growth bility of methods for synchronization of tumor cells that can factor stimulation, whereas histone H4 mRNA was detected at also be used with proliferating normal cells. Highly synchro the onset of DNA synthesis and remained constant thereafter nous cell populations are required for the study of cell cycle (Fig. 5, C and D). These observations indicate that Lovastatin events that regulate proliferation in these cells. In this report. 3606

c-myc **â€¢*â€¢â€¢**â€¢Â»â€¢â€¢â€¢â€¢â€¢â€¢

Histone H4 â€¢â€¢â€¢â€¢â€¢.*â€¢â€¢â€¢* Fig. 5. Northern blot analysis of tumor cells synchronized by Lovastatin versus normal cells synchronized by growth factor depriva tion. Exponentially growing human tumor-de rived breast epithelial 2IPT cells (A, B) were 36B4 synchronized by treatment with 20 pM Lovas tatin for 36 h. Fresh medium containing 2 mM mevalonate was added at time 0 h. Normal 0 4 e 12 16 20 24 28 32 36 40 44 48 52 56 60 8 16 24 32 40 48 56 64 human breast epithelial 70N cells (C, D) were synchronized by growth factor deprivation as Hours Following Lovastatin Removal described for Fig. 4. At indicated times, total RNA was extracted from cells and analyzed by D Northern blot analysis, using 20 MgRNA/lane. c-myc . ... Â«...*..* Blots were hybridized with c-myc, histone H4. O or 36B4 (30) (used for equal loading). Arrows Q. (A and C), locations of 28S and 18S ribosomal O u 3- RNAs. For each experimental condition, DNA Â£ synthesis and cell density were also measured (B, D) as described in text. Histone H4 II Z'

36B4

036 9 12 15 18 21 24 27 30 33 36 Hours Following Growth Factor Stimulation we have described a procedure for synchronizing normal and tumor cells in the G, phase of the cell cycle using Lovastatin. This procedure has been used successfully to synchronize the growth of many types of normal and tumor-derived fibroblast as well as epithelial cells. A number of approaches have been used in the past to investigate events in cycling cultured cells including synchro nization by mitotic detachment (35), centrifugal elutriation (36), and flow cytometry (37). These physical methods of syn chrony do not use any toxic agents, and the cells are not deprived of essential nutrients. However, only a modest degree of synchrony can be achieved due to considerable overlapping of different phases of the cell cycle resulting in a low yield of a synchronous population. Synchronization by isoleucine depri vation (38), double thymidine block (39), and other chemical Time, hours methods such as treatment with methotrexate (40, 41), hydrox- Fig. 6. Reversal of Lovastatin synchronization. Exponentially growing MCF- yurea (3, 42), or aphidicolin (43, 44) induce a high degree of 7 cells were plated on 5 x 24-well plates. Lovastatin, 20 ^M, was added to the culture medium alone (â€¢)orin the presence of 30 /ug/ml cholesterol-ethanol (D), metabolic imbalance, perturbing subsequent progression of cells 100 fig/ml low density lipoprotein-cholesterol (O), 80 fiM squalene (A), or 2 mM through the cell cycle, and can have severe cytotoxic effects. mevalonic acid (â€¢),for33 h. Fresh medium containing 2 mM mevalonic acid was Synchronization by Lovastatin overcomes all the above lim then given at zero time (abscissa). Cells were harvested at the indicated time intervals, and DNA synthesis and cell density were measured as described in text. itations with the following advantages: (a) tumor as well as normal cells are synchronized in the same phase of the cell cycle, i.e., early G,; (Â¿>)Lovastatin arrests cells reversibly and The mechanism by which Lovastatin synchronizes cells is without any cytotoxicity and is reversed by the addition of unknown. Lovastatin suppresses the synthesis of mevalonate mevalonate; (c) the general metabolism of RNA, protein, and by inhibiting HMG-CoA reducÃase,the rate-limiting enzyme in initial DNA synthesis are not disturbed, and the cells are not the sterol-biosynthetic pathway. Mevalonate is the precursor of nutriently or growth factor deprived with this treatment; (d) a number of isoprenoid products: these include cholesterol, large quantities of synchronous cell populations are obtained; involved in membrane structure; dolichol phosphates required (e) Lovastatin treatment is not dependent on cell density nor for glycoprotein synthesis; polyisoprenoid side chains of ubi- the type of tissue culture medium required for cell proliferation; quinone and heme-a, involved in electron transport; and iso- (/) this treatment can be used with many different types of pentenyladenine, which is present in some types of transfer cells; (g) the Lovastatin-induced synchrony can be followed for RNAs (45-48). Furthermore, Lovastatin or compactin also several cycles, allowing the study of molecular and biochemical inhibits DNA replication, which is restored by the addition of events that occur in cycling versus G0 released cells. mevalonate, the product of the reaction catalyzed by HMG- 3607

CoA reducÃase, but not by the addition of intermediates of and protein synthesis during interphase in L-cells. Exp. Cell Res., 57: 114- 118, 1969. nonsterol and sterol end products of the mevalonate pathway 6 Medrano, E. E.. and Pardee, A. B. Prevalent deficiency in tumor cells of (11,49-51). These studies suggest that mevalonate production, cycloheximide induced cell cycle arrest. Proc. Nati. Acad. Sci. USA, 77: 4123-4126, 1980. independent of its role as a precursor of cholesterol synthesis, 7 Scher, C. D., Stone, M. E., and Stiles, C. D. Platelet-derived growth factor plays an essential role in DNA replication. Mevalonic acid is prevents Go growth arrest. Nature (Lond.). 281: 390-392, 1979. also required for the post-translation modification of a number 8 Alberts, A. W., Chen, J., Kuron, G., Hunt, V.. Hoffman, C, Rothrock, J., Lopez, M., Joshua, H., Harris, E., Patched, A., Monaghan, R., Currie, S.. of proteins including raj-related GTP-binding proteins, nuclear Stapley, E., Albers-Schonberg, G., Hensens, O., Hirshfield, G., Hoogsteen, lamin B and prelamin A, and the g subunits of heterotrimeric K., Liesch, J., and Springer, J. Mevinolin: a highly potent competitive G proteins (reviewed in Ref. 48). inhibitor of hydroxymethylglutaryl-coenzyme A reducÃaseand a cholesterol- lowering agent. Proc. Nati. Acad. Sci. USA, 77: 3957-3961, 1980. A number of reports have suggested that inactivation of ras 9 Endo, A., Kuroda. M.. and Tansawa, K. Competitive inhibition of 3-hydroxy- by Lovastatin is involved in inhibition of DNA synthesis (48). 3-methylglutaryl coenzyme A reducÃaseby ML-236A and ML-236B. fungal metabolites having hypocholesterolemic activity. FEBS Lett., 72: 323-326, To be active, ras must associate with the cell plasma membrane. 1976. This association is achieved by the addition of a farnesyl moiety 10 Larson, R. A., Chung, J., Scanu, A. M., and Vachnin, S. Neutrophils are to the carboxyl terminus of the protein. Lovastatin has been required for the DNA synthetic response of human lymphocytes to mevalonic acid: evidence suggesting that a nonsterol product of mevalonate is involved. shown to inhibit the biological activity of the ras protein by Proc. Nati. Acad. Sci. USA. 79: 3028-3032, 1982. interfering with this post-translational modification, which is n Habenicht, A. J. R., Glomset, J. A., and Ross, R. Relation of cholesterol and mevalonic acid to the cell cycle in smooth muscle and Swiss 3T3 cells derived from mevalonate (52, 53). However, recently it has been stimulated to divide by platelet-derived growth factor. J. Biol. Chem., 255: reported that inhibition of cell growth by Lovastatin is inde 5134-5140, 1980. pendent of ras function. Lovastatin had a similar growth- 12 Maltese, W. A., and Sheridan. K. M. Isoprenylated proteins in cultured cells: inhibitory activity against ras-dependent as well as ras-inde- subcellular distribution and changes related to altered morphology and growth arrest induced by mevalonate deprivation. J. Cell. Physiol., 133:471- pendent cell lines (54). 481, 1987. We show that only mevalonate, of the compounds tested, can '3 Langan, T. J., and Volpe. J. J. Cell cycle-specific requirement for mevalonate, but not for cholesterol, for DNA synthesis in glial primary cultures. J. reverse the Lovastatin-induced synchrony; late intermediates in Neurochem., 49: 513-521, 1987. the sterol biosynthesis pathway were ineffective. It is probable 14. Doyle, J. W., and Kandutsch. A. A. Requirement for mevalonate in cycling that in early GI phase the action of one or more proteins, which cells: quantitative and temporal aspects. J. Cell. Physiol., 137: 133-140, 1988. require modification by an early intermediate of the cholesterol ]5 Sinensky, M., and Logel, J. Defective macromolecule biosynthesis and cell- biosynthesis pathway, is (are) prevented by the addition of cycle progression in a mammalian cell starved for mevalonate. Proc. Nati. Lovastatin. The activity of these proteins maintains the prolif- Acad. Sci. USA. 82: 3257-3261, 1985. 16. Todaro, G. J., and Green, H. Quantitative studies of the growth of mouse erative ability of cells in culture. These protein are re-activated embryo cells in culture and their development into established lines. J. Cell by the addition of mevalonate and induce the cells to traverse Biol. / 7: 299-313, 1963. 17. Holley, R. W., Baldwin. J. H., Kiernan, J. A., and Messmer, T. O. Control the cell cycle in a synchronous fashion. of growth of benzo(a)pyrene-transformed 3T3 cells. Proc. Nati. Acad. Sci. In summary, we have discussed a unique method to synchro USA, 7.ÃŒ:3229-3232,1976. 18. Sager, R., and Kovac, P. E. Genetic analysis of tumorigenesis: I. Expression nize normal and tumor epithelial as well as fibroblast cells of of tumor-forming ability in hamster hybrid cell lines. Somat. Cell. Genet., 4: several different species in the d phase of the cell cycle. Large 375-392, 1978. quantities of synchronous cell populations are obtained that 19 Smith. B. L., and Sager. R. Multistep origin of tumor-forming ability in Chinese hamster embryo fibroblast cells. Cancer Res., 42: 389-396, 1982. can be used to study cell cycle events that regulate proliferation. 2Q Band, V., and Sager, R. Distinctive traits of normal and tumor-derived Furthermore, cells synchronized by Lovastatin can be followed human mammary epithelial cells expressed in a medium that supports long- for multiple cycles, allowing the study of events that occur in term growth of both cell types. Proc. Nati. Acad. Sci. USA, 86: 1249-1253, mid- to late cyclesâ€”which up to now has not been possible. 1989. 21. Band, V., Zajchowski, D.. Swisshelm, K., Trask, D., Kulesa, V., Cohen, C., Connolly, J., and Sager, R. Tumor progression in four mammary epithelial cell lines derived from the same patient. Cancer Res., 50: 7351-7357, 1990. ACKNOWLEDGMENTS 22 Snedcor, G. W. Curvilinear regression. In: Statistical Methods, pp. 42-52. Ames, IA: Iowa State College Press, 1956. We thank A. W. Alberts of Merck Sharp and Dohme Research 23. Harrock. D. L. The H number concept. Beckman Technical Bulletin, 1095- Laboratories for his generous supply of Lovastatin, Drs. Jean Gudas NUC-77-IT, Irvine, CA, 1977. and Jerry R. Faust for helpful discussions, Drs. Lesley Michalowsky 24 Johnson, G. D.. and Nogueira Araujo. G. M. d. C. A simple method of reducing the fading of immunofluorescence during microscopy. J. Immunol. and Anne Crozat for their critical analyses of this manuscript, and Methods, 43: 349-350. 1981. Pamela Markell and Victoria Kulesa for their excellent technical 25 Chirgwin, J. M., Przbyla, A. E., MacDonald. R. J.. and Rutter. W. J. Isolation assistance. of biologically active ribonucleic acid from sources enriched in ribonucleases. Biochemistry. 18: 5294-5299. 1979. Note Added in Proof 26. Gudas. J. M.. Knight, B. G., and Pardee. A. B. Nuclear posttranscriptional processing of thymidine kinase mRNA at the onset of DNA synthesis. Proc. Since this paper was accepted, confirming information regarding cell cycle Nati. Acad. Sci. USA, 85: 4705-4709, 1988. specific effects of Lovastatin has been reported: Jakobisiak, M.. Bruno, S., 27, Feinberg, A. P., and Vogelstein, B. A. A technique for radiolabeling DNA Skierski, J. S., and Darzynkiewicz, Z. Proc. Nati. Acad. Sci. USA, 88: 3628- restriction endonuclease fragments to high specific activity. Anal. Biochem., 3632, 1991. m-6-13, 1983. 28. Battey, J., Moulding. C., Taub, R., Murphy. W.. Stewart, T., Potter, H., Lenoir. G., and Leder, P. The human c-myc oncogene: structural conse REFERENCES quences of translocation into the IgH locus in Burkitt lymphoma. Cell, 34: 779-787, 1983. 1. Pardee, A. B. GÃ¬events and regulations of cell proliferation. Science (Wash 29. Plumb. M.. Stein. J.. and Stein. G. Coordinate regulation of multiple histone ington DC). 246: 603-608, 1989. mRNAs during the cell cycle in HeLa cells. Nucleic Acids Res., //: 2391- 2. Pardee, A. B. A restriction point for control of normal animal cell prolifera 2410. 1983. tion. Proc. Nati. Acad. Sci. USA. 71: 1286-1290, 1974. 30 Masiakowski, P., Breathnach, R., Bloch, J., Gannon, F., Krust, A., and 3. Ashihara, T., and Baserga, R. Cell synchronization. Methods Enzymol.. 58: ChambÃ³n, P. Cloning of cDNA sequences of hormone-regulated genes from 248-262, 1979. the MCF-7 human breast cancer cell line. Nucleic Acids Res., 10: 7895- 4. (,rdina. D. J., Meistrich, M. L., Meyn, R. E., Johnson, T. S., and White, R. 7903, 1982. A. Cell synchrony techniques. A comparison of methods. In:}. W. Gray and 31 Murphy, J. S., D'Alisa. R., Gershey, E. L., and Landsberger, F. R. Kinetics Z. Darzynkiewicz (eds.). Techniques in Cell Cycle Analysis, pp. 367-402. of desynchronization and distribution of generation times in synchronized Clifton, NJ: Humana Press, 1987. cell populations. Proc. Nati. Acad. Sci. USA. 75:4404-4407, 1978. 5. Zetterberg, A., and Skold. O. The effect of serum starvation or DNA, RNA 32 Engelberg. J. A method of measuring the degree of synchronization of cell 3608

populations. Exp. Cell Res., 23: 218-227, 1961. 44. Matherly, L. H., Schuetz, J. D., Westin, E., and Goldman, I. D. A method 33. Gerdes, J., Schwab. U., Lemke, H., and Stein, H. Production of a mouse for the synchronization of cultured cells with aphidicolin: application to the monoclonal antibody reactive with a human nuclear antigen associated with large-scale synchronization of LI210 cells and the study of the cell cycle cell proliferation. Int. J. Cancer, 31: 13-20. 1983. regulation of thymidylate synthase and dihydrofolate reducÃase. Anal. 34. Gerdes, J., Lemke, H., Baisch, H., Wacker, H-H.. Schwab, U., and Stein, H. Biochem., 182: 338-345, 1989. Cell cycle analysis of a cell proliferation-associated human nuclear antigen 45. Goldstein, J. L.. and Brown, M. S. Regulation of the mevalonate pathway. defined by the monoclonal antibody Ki-67. J. Immunol., 133: 1710-1715. Nature (Lond.). 343: 425-430, 1990. 1984. 46. Kaneko, I., Hazama-Shimada. Y. and Endo, A. Inhibitory effects on lipid 35. Terasima, T., and Tolmach, L. J. Growth and nucleic acid synthesis in metabolism of ML-236B. a potent inhibitor of 3-hydroxy-3-methylglutaryl synchronously dividing populations of HeLa cells. Exp. Cell Res., 30: 344- coenzyme A reducÃase.Eur. J. Biochem., 87: 313-321, 1978. 362. 1963. 47. Faust, J. R., Brown, M. S.. and Goldstein, J. L. Synthesis of isopentenyl 36. Mitchell, B. F., and Tupper, J. T. Synchronization of mouse 3T3 and SV40 tRNA from mevalonate in cultured human fibroblasts. J. Biol. Chem., 255: 3T3 cells by way of centrifugal elutriation. Exp. Cell Res., 106: 351-355, 6546-6548, 1980. 1977. 48. Maltese, W. A. Posttranslational modification of proteins by isoprenoids in 37. Dean, P. N., Gray, J. W., and Dolbeare, P. A. The analysis and interpretation mammalian cells. FASEB J.. 4: 3319-3328, 1990. of DNA distributions measured by flow cytometry. Cytometry, 3: 188-195, 49. Quesney-Huneeus, V., Wiley, M. H., and Siperstein, M. D. Essential role 1982. for mevalonate synthesis in DNA replication. Proc. Nati. Acad. Sci. USA, 38. Tobey, T. A., and Ley, K. D. Regulation of initiation of DNA synthesis in 76:5056-5060, 1979. Chinese hamster cells. 1. Production of a stable, reversible G.-arrested pop 50. Perkins, J. L., Ledin, J. F., and Stubbs, J. D. Linkage of the isoprenoid ulations in suspension culture. J. Cell Biol. 46: 151-157, 1970. biosynthesis pathway with induction of DNA synthesis in mouse lympho 39. Studzinski, G. P., and Lambert, W. C. Thymidine as a synchronizing agent cytes. Effects of compactin on mitogen induced lymphocytes in serum-free I. Nucleic acid and protein formation in with excess thymidine. J. Cell. medium. Biochim. Biophys. Acta 711: 83-89, 1982. Physiol.. 73: 109-115, 1969. 51. Witte, L. D., Fairbanks, K. P.. Barbu, V., and Goodman, D. S. Studies on 40. Sen, S., Erba, E., and D'Incaici, M. Synchronisation of cancer cell lines of cell proliferation and mevalonic acid metabolism in cultured human fibro human origin using methotrexate. Cytometry. //: 595-602, 1990. blasts. Ann. NY Acad. Sci.. 454: 261-269, 1985. 41. Yunis, J. J. High resolution of human chromosomes. Science (Washington 52. Casey, P. J., Solski, P. A., Der, C. J., and Buss, J. E. p21ras is modified by DC), 191: 1268-1270. 1976. a fanesyl isoprenoid. Proc. Nati. Acad. Sci. USA, 86: 8323-8327, 1989. 42. Mironescu, S.. and Ellem, K. A. O. Secondary activities of diverse inhibitors 53. Hancock, J. F.. Magee, A. I., Childs, J. E., and Marshall, C. J. All ras potentiate the response of hamster embryo cultures to a mitotic stimulus. J. proteins are polyisoprenylated but some are palmitoylated. Cell, 57: 1167- Cell. Physiol., 90: 281-294, 1977. 1177, 1989. 43. Tobey, R. A., Valdez, J. G., and Crissman, H. A. Synchronization of human 54. DeClue, J. E., Vass, W. C., Papageorge, A. G., Lowy, D. R., and Willumsen, diploid fibroblasts at multiple stages of the cell cycle. Exp. Cell Res., 179: B. M. Inhibition of cell growth by Lovastatin is independent of ras function. 400-416, 1988. Cancer Res., 51: 712-717, 1991.

3609

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1991 American Association for Cancer Research. Synchronization of Tumor and Normal Cells from G1 to Multiple Cell Cycles by Lovastatin

Khandan Keyomarsi, Larue Sandoval, Vimla Band, et al.

Cancer Res 1991;51:3602-3609.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/51/13/3602

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/51/13/3602. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.