Changes in Calcitonin Gene RNA Processing During Growth of a Human Medullary Thyroid Carcinoma Cell Line1

(CANCER Ri:Si:\ROI 49. 6949-6952. December 15. 1989] Changes in Calcitonin Gene RNA Processing during Growth of a Human Medullary Thyroid Carcinoma Cell Line1

Barry D. Nelkin,2 Kurt Y. Chen, AndrÃ©edeBustros, Bernard A. Roos, and Stephen B. Baylin The Oncology Center, Johns Hopkins I'niversity School of Medii -inc. Baltimore, Maryland 21205 /B. I). ,V.. K. Y. C., A. de B., S. H. H./; the (ieriatric Research. Education, ami Clinical Center (dRECC). I 'eterans Administration Medical Center, Tacoma. H'ashington 98493 /B. A. R./: and the I 'nirersily of Hushinnton, Seattle, Washington VXIV? [R. A. R.I

ABSTRACT sive product (2). To date, the factors influencing the posttran scriptional choice of CT or CGRP mRNA remain unknown. The ratios of calcitonin (CT) to calcitonin gene-related peptide (CGRP) We have studied CT gene regulation in the TT cell line, mRNA, both generated by alternative RNA processing from the same which was derived from a rapidly growing human MTC (7) and primary RNA transcript, arc shown by Northern blotting of cytoplasmic RNA to vary as a function of growth in a human medullary thyroid produces high levels of both CT and CGRP. We have previously carcinoma cell line ('IT). I'pon initial seeding, CT mRNA levels are shown that, in this cell line, levels of the CT peptide are relatively high, and CGRP mRNA levels are relatively low. During the modulated dynamically during various stages of growth in early logarithmic growth phase, CGRP mRNA levels rise severalfold, culture, being lowest during times of rapid cell proliferation while CT mRNA levels change only slightly. As the cells approach and highest in cells at saturation density (8). In vivo, it has also confluence, both CT and CGRP mRNA levels rise. Subsequently, CGRP been found that CT is decreased in MTC only from patients mRNA levels fall substantially in postconfluent cells, while CT mRNA with rapidly growing, aggressive tumors. In order to determine levels remain high. By actinomycin I) blocking of nascent transcription, the dynamics of CT gene expression which may regulate this we have shown that these growth-related, reversible changes in the ratio modulation of CT production, we have examined, using several of CT to CGRP mRNA arc not due to changes in mRNA stability. Our data rather suggest that 'IT cells reversibly alter alternative RNA- cDNA probes (4), CT gene-derived RNA production during processing patterns dependent upon growth conditions in vitro, such that growth of MTC cells in culture. We now find that the basis for (I mRNA is lowest and CGRP mRNA is highest during rapid growth. this difference is largely at the RNA level and that CT gene The mechanisms underlying this RNA-processing alteration may play a expression bears a complex and dynamically regulated relation role in certain patients with aggressive forms of medullary thyroid ship to cell growth. This regulation includes reversible changes carcinoma, in whom a decrease or loss of CT levels heralds a poor in posttranscriptional processing which lead to changing ratios prognosis. of the relative levels of CGRP to CT mRNAs and in part involves a reversible switch in choice of RNA processing to CT INTRODUCTION or CGRP mRNA. MTC1 is an important experimental system, biologically and clinically, for studies of both peptide hormone gene expression MATERIALS AND METHODS and tumor differentiation and progression. Both the tumor and Cell Culture. TT cells were seeded on day 0 at 1 x IO6cells/25 cm2 its normal counterpart, the thyroid C-cell, express the calcitonin flask, taking care to achieve a single cell suspension, and grown in gene. RPMI 1640 with 16% heat-inactivated fetal bovine serum (Gibco), 2 In both rat and human, the CT gene is composed of six exons HIMglutamine, 100 units/ml penicillin, and 100 i

Receivcd 2/12/88: revised 6/5/89; accepted 8/24/89. Johns Hopkins University School of Medicine). The actual fraction of The costs of publication of this article were defrayed in part by the payment the total CT gene derived mRNA expressed as CT or CGRP mRNA of page charges. This article must therefore be hereby marked advertisement in was estimated by |-x-32P|ATP end labeling or reverse transcription of accordance with 18 tJ.S.C. Section 1734 solely to indicate this fact. 1This work was supported in part by grants from the American Cancer Society RNA extracted from TT cells at various time points and hybridization to CT- or CGRP-specific plasmid DNA sequences immobili/cd on (PDT-207 and NP-533): the American Cancer Society. Maryland Division. Inc.: the NIH (AM .16116); and the Veterans Administration: and by gifts from the nitrocellulose (12) (data not shown). W. W. Smith Foundation and the Hodson Trust. CT and CGRP mRNA Half-Life Measurements. TT cells were seeded 2To whom requests for reprints should be addressed, at the Oncology Center as described above. On days 3, 7, or 11, actinomycin D (8 /Â¿g/ml)was Research Laboratories. Johns Hopkins University School of Medicine. 424 N. added to block transcription, and cells were harvested after various Bond Street. Baltimore, MD 21231. 1The abbreviations used are: MTC. medullary thyroid carcinoma; CT. calci further incubation times. This method was chosen instead of pulse- tonin: CGRP. calcitonin gene-related peptide; cDNA, complementary DNA. chase techniques (13) since the high concentration of pyrimidines 6949 Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1989 American Association for Cancer Research. CT RNA PROCESSING CHANGES IN CELL GROWTH

CGRP CGRP Common_nLJiCommon2n~Lr~ CT Coding Codmq Untronsloted A Day iRNACTrâ€”LProcessing 79

CT VGRP mRNA

Common Region Specil'cCD Probe )CGRP Specific Probe ]CT Spec.lie CGRP Probe

Fig. I. Overview of human calcitonin gene structure and expression. Alter native RNA processing produces two mRNAs. one encoding CT [1.0 kilobase (AA)|and one encoding CGRP [1.1 kilobases). Specific common region (exons 2 and 3). CT (exon 4), and CGRP (exons 5 and 6) hybridization probes shown were derived from cloned cDNAs (4). Data from Refs. 1 and 4-6.

required for pulse-chase are toxic to TT cells. RNA was extracted as described above, and 2 Â¿tgwereapplied to nitrocellulose using a slot blot manifold (Schleicher and Schuell, Keene, NH). Slot blots were hybridized with the CT- and CGRP-specific probes described above.

RESULTS We have examined, by Northern blotting using CT- and CGRP-specific mRNA probes, the dynamic variation of the relative levels of these mRNAs in TT cells during growth. Fig. 2, A and B, shows that, during TT cell growth, CT-specific mRNA levels generally paralleled those of total CT gene- derived mRNA (Fig. 2Ã„);these relative mRNA levels remained constant or decreased for several days after cell seeding and then gradually increased as the cells progressed toward conflu ence (after day 6). These changes in CT mRNA levels were consistent with our previous observation that CT peptide levels fall during the initial growth phase of the TT cells and increase to initial levels at confluence (8). Thus, CT peptide production in these cells during growth appears to be largely regulated at the mRNA level. When the hybridization was repeated using a CGRP-specific probe (exons 5 and 6), a different pattern emerged. It became apparent that the relative levels of CGRP mRNA varied during growth, in a pattern unlike that of CT-specific or overall CT gene-derived mRNA. Thus, for the first 2-3 days after the cells Fig. 2. Measurement by hybridization of relative levels of CT and CGRP were plated, relative levels of CGRP mRNA were quite low, mRNAs in TT cells during growth in culture. A, Northern blot hybridization of relative to CT-specific or total CT-gene derived mRNAs (Fig. total cytoplasmic RNA from TT cells during growth in culture, using probes 2B). Then, during rapid cell growth on days 4-6 after plating, specific for CT (exon 4) and CGRP (exons 5 and 6) mRNA. B. mean of normalized densitomctric analysis of three independent experiments, including the experi when CT mRNA levels were decreased somewhat, levels of ment in A, measuring, by Northern blotting. CT and CGRP mRNA in TT cells CGRP mRNA actually began to increase. That this initial during growth in culture. In this analysis, a common region (exons 2 and 3) probe was also used to assess total CT gene-derived mRNAs. Common region probes increase in CGRP mRNA was not due solely to an increase in hybridize to a single band, since CT and CGRP mRNAs, 1.0-1.1 kilobases. are total CT gene mRNA levels is demonstrated in Fig. 2C, since, not separated on these blots. â€¢¿.CT-specificmRNA; x, CGRP-specific mRNA; from day 2 to day 6, the ratio of relative levels of CGRP mRNA O. common region. C, changing ratio of relative levels of CGRP mRNA to CT to CT mRNA rose 10-fold. During days 6-9, as the cells mRNA in TT cells, derived from data in B. approached confluence, levels of total CT gene-derived mRNA increased 2-fold (common region probe); the ratio of relative primary transcript. The differences could be mediated by differ levels of CGRP to CT mRNA stabilized at the new higher ential transport efficiency of the mRNA species from the nu levels as CT and CGRP mRNA increased in parallel (Fig. 2, A cleus, differential mRNA stability, or events involving alterna and B). When the cells reached superconfluence, the ratio of tive RNA processing. We have investigated these three possi relative levels of CGRP mRNA to CT mRNA rapidly fell, bilities. First, the differences are probably not due to differential despite the fact that the steady state level of total CT gene transport efficiency, since, by Northern blotting, nuclear RNAs mRNA (common region probe) and CT-specific mRNA re have the same CT/CGRP RNA ratio as do the cytoplasmic mained fairly constant (day 11). RNAs on days 3, 7, and 11 (data not shown). Second, the These differences in the relative levels of CT and CGRP differences do not appear to be due to divergent changes in the mRNAs at different times of growth in the TT cells must be stability of the CT and CGRP mRNAs, since no differences in posttranscriptional, since both messages derive from a common mRNA half-life (f 1/2)between these two mRNA species could 6950 Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1989 American Association for Cancer Research. CT RNA PROCESSING CHANGES IN CELL GROWTH be detected by Northern blotting. Densitometric plots of the Day hybridization signals are shown in Fig. 3. This measurement 7 showed that CT and CGRP mRNAs were relatively stable, and A differences in /,/; sufficient to explain the changes in CT/CGRP mRNA ratios were not observed. CT Cote and Gagel (14) have reported that dexamethasone treat ment of TT cells changed their CCRF/CT mRNA ratio, but, unlike our results, they showed only minor growth-dependent modulation of this ratio in the absence of dexamethasone. We cannot account for the discordance of our results with theirs, but we speculate that subtle differences in culture conditions may be involved. In this study, we have used 16% fetal bovine Â»â€¢â€¢* CGRP serum, while Cote and Gagel used 10%, and we changed the medium only on day 6, while Cote and Gagel changed the medium every second day. However, we have repeated our study 3.2 using their conditions, and we observed changes in CGRP/CT o mRNA ratios similar to those seen when TT cells were grown rr 2.8 under our original culture conditions (Fig. 4). Thus, the differ I- 2.4 o ences between our results and those of Cote and Gagel remain 2.0 Q_ to be elucidated. rr o 1.6 o 1.2 DISCUSSION 0.8 We have shown that both CT and CGRP mRNA levels and, most importantly, the ratio of relative levels of CGRP/CT rr mRNA change dramatically during routine growth of TT cells. B 13 The stability of these mRNAs is unchanged. Taken together, Day the above data strongly indicate that a change in posttranscrip- Fig. 4. Effect of alteration of culture condition on the ratio of relative levels tional RNA processing explains the change in CGRP/CT of CT and CGRP mRNA. TT cells were seeded as described in "Materials and mRNA ratios. Methods," in RPMI 1640 containing 10% fetal bovine serum, and the medium Among several factors influencing CT gene expression in TT was changed every 2 days. These conditions simulate the growth conditions in Ref. 14. A, Northern blot hybridization of total cytoplasmic RNA from TT cells cells, the factor accounting for the dramatic increase we see in during growth in culture, using the probes described in Fig. 2A. These hybridi zations were done sequentially on the same blot. B, ratio of relative levels of CGRP mRNA to CT mRNA in TT cells under these growth conditions, derived from data from the average of duplicate Northern blots of the experiment in A, normalized for fi-actin hybridization. The changes in relative CGRP/CT mRNA ratio seen in TT cells grown in 16% serum thus are also seen in TT cells grown in 10% scrum with frequent media changes.

the ratio of relative levels of CGRP mRNA to CT mRNA, from day 2 to day 6, and the fall in this ratio from days 9 to 11, is a dynamic modulation of initial RNA processing choice. This processing modulation is superimposed on events from days 6 to 9 of growth, which result in somewhat higher levels of both CT and CGRP mRNAs. The TT cells thus provide a dynamic system for studying the control of posttranscriptional RNA processing events. Tissue- or differentiation-specific control of alternative RNA process ing mediated at the polyadenylation and/or splicing level has been reported for numerous genes (for review, see Ref. 15). We believe that the reversibility of the CT gene RNA processing choice under routine in vitro growth conditions and its depend ence upon the growth patterns of the TT cells will make this cell system a fruitful one in which to study these functions. An alternative possibility is that the TT cell line contains two populations of cells, one expressing relatively more CT mRNA and one expressing relatively more CGRP mRNA. Regulation

Fig. 3. Estimation of half-life of CT and CGRP mRNAs. TT cell transcription of CT gene transcription in these two putative cell populations was blocked on indicated days by exposure to actinomycin D (8 /ig/ml). RNA could then result in alterations in the CGRP/CT mRNA ratio was harvested after various incubation times, slot-blotted, and hybridized to CT such as we have shown. We consider this possibility unlikely, or CGRP-specific probes. Densitometric intensity of hybridization signals are expressed as A///V0. the fraction of the hybridization signal remaining at a given however. The reversible changes we have seen have remained time, relative to the hybridization signal at the time of addition of actinomycin. for over 100 passages in culture; this would demand an exactly Â».day 3: â€¢¿.day7; A. day 11. A. CT-specific mRNA. l,n on days 3. 7. and 11 were 25, 22, and 52 h. respectively. B. CGRP-specific mRNA. /l/2 on days 3, 7, equal growth rate for each putative cell type to maintain both and 11 were 26, 22, and 29 h. respectively. populations in the TT cell line. 6951

One important challenge will be to identify the physiological 2. Rosenfeld. M. G.. Lin. C.R.. Amara, S. G.. Stolarsky, L., Roos. B.A.. Ong. E. S., and Evans. R. M. Calcitonin mRNA polymorphism: peptidc switching stimuli for the type of modulation of RNA processing we now associated with alternative RNA splicing events. Proc. Nati. Acad. Sci. USA, report. Among the possibilities are regulation by soluble se 79: 1717-1721. 1982. 3. Rosenfeld, M. G.. Mermod, J-J., Amara, S. G., Swanson, L. W., Sawchenko, creted factors, cell shape, and direct cell-cell interactions. Pre P. E., Rivier. J., Vale. W. \\'., and Evans. R. M. Production of a novel sumably, such signals direct the cells to synthesize trans-acting neuropeptide encoded by the calcitonin gene via tissue-specific RNA proc factors, such as RNA splicing or polyadenylation cofactors, essing. Nature (Lond.), 304: 129-134, 1983. 4. Nelkin, B. D.. Rosenfeld. K. I., de Bustros, Aâ€žLeong, S. S.. Roos, B. A., which mediate, as a function of growth, the choice to process and Baylin. S. B. Structure and expression of a gene encoding human the primary RNA transcript to cytoplasmic CT or CGRP calcitonin and calcitonin gene related peptide. Biochem. Biophys. Res. Com mRNA. Similar factors may control CT gene processing in vivo, mun., 123: 648-653. 1984. 5. Edbrooke, M. R.. Parker. D.. McVey, J. H., Riley. J. H.. Sorenson, G. D., since CT and CGRP mRNA are expressed in a tissue-specific Pettengill. O. S.. and Craig. R. K. Expression of the human calcitonin/ manner (3). In the TT cells, we envision that from day 2 to day CGRP gene in lung and thyroid carcinoma. EMBO J.. 4: 715-724, 1985. 6. Jonas. V., Lin. C. R.. Kawashima. E., Semon, D.. Swanson. L. W., Mermod. 6 these factors change to increase the fraction of CT gene J-J.. Evans, R. M., and Rosenfeld, M. G. Alternative RNA processing events transcript processed to cytoplasmic CGRP mRNA; then, from in human calcitonin/calcitonin gene-related peptide gene expression. Proc. day 6 to day 9, the factors presumably remain constant to Nati. Acad. Sci. USA, 81: 1994-1998. 1985. 7. Leong. S. S., Horoszewicz, J. S., Shimaoka. K., Friedman, M., Kawinski. E.. stabilize the fraction of CGRP mRNA at its higher level and Song, M. J.. Zeigel. R.. Chu. T. M.. Baylin, S. B., and Mirand. E. A. A new subsequently, after day 9, change again in favor of CT mRNA. cell line for study of human medullary thyroid carcinoma. In: M. Andreoli. The nature of such putative trans-acting factors is not known. F. Monaco, and J. Robbins (eds.). Advances in Thyroid Neoplasia 1981. pp. 95-108. Rome: Field Educational Italia, 1985. Small nuclear ribonucleoproteins have been suggested to be 8. Berger, C. L., de Bustros. A., Roos, B. A.. Leong. S. S., Mendelsohn. G., involved in RNA splicing recognition (16-21), polyadenylation Gesell. M. S., and Baylin. S. B. Human medullary thyroid carcinoma in culture provides a model relating growth dynamics, endocrine cell differen (22), and other RNA processing steps (23). Microheterogeneity tiation, and tumor progression. J. Clin. Endocrino!. Metab.. 59: 338-343. of different members of small nuclear RNA gene families, such 1984. as has been observed in Xenopus laevis U l genes (24), or other 9. Berger, S. L., Wallace. D. M.. Siegal, G. P., Hitchcock. M. J. M.. Birken- meier, and Reber, S. B. Preparation of interferon messenger RNAs with the factors could possibly mediate a change in recognition of RNA- use of ribonucleoside-vanadyl complexes. Methods Enzymol., 79: 59-68. processing sequences in the primary transcript. However, by 1981. urea-polyacrylamide gel electrophoresis, we have detected no 10. Maniatis, T., Fritsch, E. F.. and Sambrook. J. Molecular Cloning, a Labo ratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, gross differences in small nuclear RNAs from TT cells har 1982. vested after various times in culture (data not shown). Leff et 11. Steenbergh, P. H., Hoppener. J. W.. Zandberg, J., Lips, C. J., and Jansz, H. al. (25) have reported an extensive study of the effects of site- S. A second human calcitonin/CGRP gene. FEBS Lett., 183:403-407. 1985. 12. Derman. E.. Krauter, K.. Walling, L.. Weinberger. C., Ray. M., and Darnell. directed mutagenesis of the CT gene on RNA processing. They J. E., Jr. Transcriptional control in the production of liver-specific mRNAs. concluded that neither the splice acceptor sites nor the poly Cell, 23: 731-739. 1981. 13. Rodgers, J. R., Johnson, M. L., and Rosen, J. M. Measurement of mRNA adenylation sites alone were sufficient to account for cell spec concentration and mRNA half-life as a function of hormonal treatment. ificity of processing, and they suggested that a factor acts in Methods Enzymol.. 709: 572-592, 1985. 14. Cote, C. J.. and Gagel. R. F. Dexamethasone differentially affects the levels trans upon the RNA secondary structure to direct specific of calcitonin and calcitonin gene-related peptide mRNAs expressed in a processing. human medullary thyroid carcinoma cell line. J. Biol. Chem., 261: 15524- Finally, the physiological effect of the variations of CT and 15528. 1986. 15. Leff, S. E., Rosenfeld. M. G., and Evans, R. M. Complex transcriptional CGRP levels in the TT cells must be considered. The type of units: diversity in gene expression by alternative RNA processing. Annu. RNA processing changes we report may represent an important Rev. Biochem., 55: 1091-1117. 1986. type of regulation pathway for cell growth. One may speculate 16. Lerner, M. R., Boyle. J. A.. Mount. S. M., Wolin, S. L., and Steitz. J. A. Are snRNAs involved in splicing? Nature (Lond.), 283: 220-224. 1980. that this specific regulation of CT and CGRP production might 17. Rogers, J.. and Wall, R. A mechanism for RNA splicing. Proc. Nati. Acad. be involved in autocrine control of growth of normal and Sci. USA, 77: 1877-1879, 1980. neoplastic C-cells. This possibility is especially intriguing, con 18. Padgett, R. A., Mount. S. M.. Steitz, J. A., and Sharp, P. A. Splicing of messenger RNA precursors is inhibited by antiscra to small nuclear ribonu- sidering the dynamics of CT production of MTC tumors in cleoprotein. Cell, 35: 101-107. 1983. 19. Keller. E. B., and Noon, W. A. Intron splicing: a conserved internal signal vivo. Human MTC is usually a relatively indolent, well differ in introns of animal pre-RNA's. Proc. Nati. Acad. Sci. USA, 81: 7417-7421, entiated endocrine tumor, which, like its nonneoplastic parent 1984. C-cell, produces high levels of CT. Occasionally, however, MTC 20. Berget, S. M., and Robberson, B. L. Ul, U2, and U4/U6 small nuclear can be an aggressive lethal disease with widespread, steadily ribonucleoproteins are required for in t'itro splicing but not polyadenylation. Cell. 46: 691-696, 1986. growing mÃ©tastases.In these latter tumors, there is heteroge 21. Black, D. L., and Steitz, J. A. Pre-mRNA splicing in vitro requires intact neous CT production, with large cell populations within the U4/U6 small nuclear ribonucleoprotein. Cell. 46: 697-704, 1986. tumor producing little or no immunohistochemically detectable 22. Hashimoto. C., and Steitz, J. A. A small nuclear ribonucleoprotein associates with the AAUAAA polyadenylation signal in vitro. Cell. 45: 581-591. 1986. CT (26, 27). In light of our present results, one may speculate 23. Strub, K.. Galli,G.. Busslinger. M., and Birnsticl. M. L. ThecDNA sequences that alterations in regulation of the RNA-processing system, of (he sea urchin U7 small nuclear RNA suggest specific contacts between histone mRNA precursor and 117 RNA during RNA processing. EMBO J.. which accompany growth changes, could play a role in the loss 5:2801-2807. 1984. of CT from rapidly growing tumors. Furthermore, in this and 24. Forbes, D. J., Kirschner, M. W., Caput, D., Dahlberg, J. E., and Lund, E. other cell systems, such regulation of factors involved in RNA Differential expression of multiple UI small nuclear RNAs in oocytes and processing may simultaneously effect pleiotropic changes on embryos of Xenopus laevis. Cell, 38: 681-689, 1984. 25. Leff, S. E., Evans, R. M., and Rosenfeld, M. G. Splice commitment dictates numerous alternatively processed genes, resulting in profound neuron-specific alternative RNA processing in calcitonin/CGRP gene expres phenotypic effects on the cell. sion. Cell, 48: 517-524, 1987. 26. Baylin, S. B.. and Mendelsohn, G. Medullary thyroid carcinoma: a model for the study of tumor progression and cell heterogeneity. In: A. H. Owens, REFERENCES Jr., D. S. Coffey, and S. B. Baylin (eds.). Tumor Cell Heterogeneity: Origins and Implications, pp. 9-21. New York: Academic Press, 1982. 1. Amara. S. G.. Jonas, V.. Rosenfeld. M. G.. Ong, E. S., and Evans. R. M. 27. Lippman, S. M., Mendelsohn, G. M., Trump. D. L.. Wells, S. A., Jr.. and Alternative RNA processing in calcitonin gÃ¨neexpression generates mRNAs Baylin. S. B. The prognostic and biologic significance of cellular heteroge encoding different polypeptide products. Nature (Lond.). 298: 240-244. neity in medullary thyroid carcinoma: a study of calcitonin, L-dopa decar- 1982. boxylase, and histaminase. J. Clin. Endocrinol. Metab.. 54: 233-240. 1982.

6952 Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1989 American Association for Cancer Research. Changes in Calcitonin Gene RNA Processing during Growth of a Human Medullary Thyroid Carcinoma Cell Line

Barry D. Nelkin, Kurt Y. Chen, Andrée de Bustros, et al.

Cancer Res 1989;49:6949-6952.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/49/24_Part_1/6949

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/49/24_Part_1/6949. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.