- Email: [email protected]
Analysis of Ras Transformation of Human Thyroid Epithelial Cells
648
regulators and effectors of small GTPases: Ras family
[51]
[51] Analysis of Ras Transformation of Human Thyroid Epithelial Cells By ZARUHI POGHOSYAN and DAVID WYNFORD‐THOMAS Abstract
Activation of Ras oncogene by point mutations is an early frequent event in thyroid tumorigenesis. In this chapter, we describe the use of human primary thyroid follicular epithelial cells expressing oncogenic mutant Ras by means of retroviral transduction as a biological model of human cancer initiation that provides powerful insights into thyroid tumorigenesis. We describe protocols for manipulating primary epithelial cells and describe the use of this model to dissect the signaling pathways required for Ras‐induced proliferation in these cells. We also highlight the importance of studying Ras signaling in an appropriate cell context, summarizing some of the key differences identified between more widespread experimental models based on fibroblasts or rodent cell lines and primary epithelial cells.
Introduction
The follicular epithelial cells of human thyroid give rise to a range of pathologically well‐defined tumor phenotypes and stages, and are proving to be one of the most informative models for studying the molecular basis of multistage human tumorigenesis in ‘‘conditional renewal’’–type epithelium. Early thyroid tumor development is closely correlated with mutation of a number of alternative genes, such as ras, ret, trk, gsp, and the TSH receptor, each associated with different tumour phenotypes, presenting a good example of genotype/phenotype association. Genetic analyses of clinical samples by our (Lemoine et al., 1988, 1989) and many other laboratories revealed that independent of clinical stage, thyroid tumors of follicular type showed a high frequency of mutation of the Ras family of oncogenes, whereas those of papillary type show a predominance of either Ret/PTC1 rearrangements or BRAF mutations. Thyroid offers a distinct advantage for modeling tumor development in tissue culture. First, near‐pure cultures of normal thyroid epithelium can be relatively easily obtained from surgical material, which, even in a simple monolayer culture, remain viable with the same simple ‘‘conditional‐ renewal’’ cell kinetics as seen in the intact gland. This contrasts with the METHODS IN ENZYMOLOGY, VOL. 407 Copyright 2006, Elsevier Inc. All rights reserved.
0076-6879/06 $35.00 DOI: 10.1016/S0076-6879(05)07051-5
[51]
Ras in human primary thyroid epithelial cells
649
technical complexity and uncertainty of interpretation involved in modeling the hierarchy of differentiation states seen in a renewing epithelium such as colon. Second, tumor initiation in the thyroid results from inappropriate induction of proliferation in what is otherwise a virtually quiescent cell population, again potentially simple to model in vitro in contrast to renewing tissues in which loss of differentiation may be the initiating event. A major principle of our work is that models of ras‐induced tumorigenesis must be based on a cell type in whose tumors Ras mutation occurs commonly and at an early stage and in which it can be shown to stimulate proliferation in vitro. By use of retroviral vectors and/or microinjection to overcome the difficulty of gene transfer into normal epithelial cells, we have exploited this system to reconstruct the initial steps in thyroid tumorigenesis. We believe that we have developed what is still a unique in vitro model of tumor initiation in a human epithelial cell that demonstrates the determining influence of the nature of the initiating oncogene on tumor phenotype. Overview of Biological Results Obtained with the In Vitro Thyroid Model and Effect of Mutant RAS on Thyroid‐Specific Differentiation
Most other in vitro studies of neoplastic transformation in the thyroid have used immortal but untransformed rodent epithelial cell lines. In these models, expression of mutant Ras leads to loss of tissue‐specific differentiation (Francis‐Lang et al., 1992; Miller et al., 1998) in contrast to the evidence from clinical analyses, implicating Ras mutation as an early event in human thyroid tumor development at a stage before loss of differentiation (Wynford‐Thomas, 1993, 1997). Human primary thyroid epithelial cells are capable of only two to three population doublings (PD) before entering a state of normally irreversible quiescence. The stable expression of mutant RasV12 induces a dramatic proliferative response, resulting in generation of colonies, the final size of which can be up to 107 cells. These colonies show a normal epithelial phenotype and expanding pattern of growth (Bond et al., 1994). However, this proliferation spontaneously ceases after 20–25 PDs, terminating in a viable state of growth arrest, resembling replicative senescence. Because of the starting low proliferative capacity of primary human thyrocytes, the number of colonies obtained after retroviral Ras expression is very low, 50–100 colonies per 105 cells (Fig. 1). To resolve this controversy, we examined the short‐ and long‐term responses of normal human thyroid epithelial cells to RasV12, introduced by microinjection (see ‘‘Methods, 3—Microinjection’’) and retroviral
650
regulators and effectors of small GTPases: Ras family
[51]
FIG. 1. Retroviral transduction of mutant Ras induces proliferation of human primary thyroid epithelial cells. Schematic diagram of a typical outcome of retroviral transduction. As a result of very low proliferative capacity of primary human thyrocytes, generally not more than 50–100 colonies are obtained per 105 cells. Representative photomicrographs of normal thyroid epithelial cells and colonies induced by mutant Ras at an early rapidly proliferating stage and at a late stage at the end of their proliferative life span.
transduction (see ‘‘Methods, 2—Retroviral Transduction Protocol’’), respectively (Gire and Wynford‐Thomas, 2000). In both cases, expression of RasV12 at a level sufficient to induce rapid proliferation did not lead to loss of differentiation, as shown by expression of cytokeratin 18, E‐cadherin, thyroglobulin, TTF‐1, and Pax‐8 proteins. Indeed, Ras was able to prevent, and even to reverse, the loss of thyroglobulin expression that occurs normally in TSH‐deficient culture medium. These responses were partially mimicked by activation of Raf, a major Ras effector, indicating involvement of the MAPK signal pathway. This striking contrast between the effect of RasV12 on differentiation in primary human, compared with immortalized rodent, epithelial cultures is most likely explained by the influence of additional cooperating abnormalities in the latter and again highlights the need for caution in extrapolating from cell line data.
[51]
Ras in human primary thyroid epithelial cells
651
Analysis of Ras Effector Pathways Mediating Mutant Ras‐Induced Proliferation of Primary Human Thyroid Epithelial Cells: A Model of Tumor Initiation
Extensive studies of Ras signaling in the past few years identified that at least three effectors are involved in RAS‐mediated cell proliferation: Raf, PI3K, Ral GDS (Marshall, 1996; Wolthuis and Bos, 1999). However, the particular necessary combination of downstream pathways can be cell type‐ and species‐specific. By use of human primary cultures of normal thyroid epithelial cells as a relevant model, we identified the required effector pathways driving RasV12‐induced proliferation in this cell type. Some of the key findings are summarized in the following. RAF‐MAPK Pathway By use of a combination of transient (scrape‐loading or microinjection, see ‘‘Methods, 3’’) or sustained (retrovirally mediated, see ‘‘Methods, 2’’) expression approaches, we first demonstrated (Gire et al., 1999) that mutant RAS induces a rapid activation of MAPK (predominantly ERK2) as assessed by phosphorylation of the enzyme and by in vitro kinase assay, together with nuclear translocation demonstrated by immunofluorescence. Importantly, this was sustained throughout the period of Ras‐induced proliferation, in contrast to the more transient activation in response to growth factor stimulation of thyroid cells. Such differences in kinetics have been found to be crucial for determining the resulting phenotype and, in this case, may contribute to the ability of Ras to induce sustained proliferation for at least 20–25 population doublings (PD), whereas the response to growth factors is limited to just 1 or 2 PD. We next determined whether this activation of MAPK is necessary for the proliferative response using two approaches for inhibiting the pathway at the level of MAPKK. Coexpression of a dominant‐negative A217‐MAPKK inhibited Ras‐induced proliferation as shown by a 60% reduction in colony yield. (The incompleteness of this effect is explicable by the partial inhibition of MAPK activity observed in a thyroid cell line test system (HT‐ori‐3) and could reflect failure to reach a level of expression sufficient to totally block wild‐type MAPKK interaction with Raf). A more complete inhibition of MAPK activity was achieved using the pharmacological inhibitor of MAPKK, PD98059, resulting in virtually complete inhibition of Ras‐induced epithelial colony formation and inhibition of proliferation in preformed colonies. This effectiveness of PD98059 was also important in excluding effects on other pathways, because it seems to be specific for MAPKK at the concentrations used, unlike the
652
regulators and effectors of small GTPases: Ras family
[51]
dominant‐negative mutant that acts by binding to Raf and could potentially block other Raf‐activated pathways. To determine whether activation of MAPK was sufficient as well as necessary for response to Ras, we used a constitutively active mutant of its immediate upstream partner MAPKK (Glu‐217/Glu‐221) to minimize the chance of stimulating additional pathways. In contrast to a previous study (Cowley et al., 1994), no evidence of proliferation in response to this vector was observed. Although it is known that the biochemical activity of the Glu‐217/Glu‐221 mutant is much lower than the physiologically activated enzyme, we showed, using a thyroid cell line (HT‐ori‐3), the level of expression from the retrovirus vector was sufficient to give activation of MAPK of comparable magnitude to that produced by RasV12. Although this cannot be checked directly in primary cells, it suggests that failure to stimulate proliferation with this vector is not a ‘‘false‐negative’’ result and that activation of MAPK is, therefore, insufficient by itself. We conclude that activation of the MAPK pathway is necessary, but not sufficient, for the proliferogenic action of RasV12 on primary human thyroid cells (Gire et al., 1999). This contrasts with results from the model closest to our own—the rat thyroid cell line WRT—in which MAPK activation seems to be dispensable for Ras‐induced mitogenesis which (Miller et al., 1997, 1998). This emphasizes the risk of extrapolation from rodent cell lines to normal human cells and is particularly important here, because the design of therapeutic strategies targeting Ras will be significantly influenced by the degree of redundancy in RAS signaling pathways. PI3K Pathway With similar approaches, we next examined the role of another major Ras effector—the PI3K pathway (Gire et al., 2000). Following the same strategy as with MAPK, we first showed by use of scrape‐loading that mutant RasV12 activated the PI3K pathway in normal human thyrocytes as revealed by phosphorylation of its downstream target, PKB/Akt. Next, by use of both an effector mutant of RasV12 (C40) specific for PI3K and a constitutively active subunit of PI3K itself (p110*), we showed that activation of this pathway alone was not sufficient to induce proliferation in normal thyrocytes. Coinfection with retroviral vectors expressing a constitutively active MAPKK and p110* did, however, generate a few small colonies, demonstrating a weak synergy between the two pathways but insufficient to fully mimic the effect of Ras. Finally, inhibition of PI3K enzyme activity by LY294002 blocked Ras‐induced colony formation and induced apoptosis in preformed colonies. These data show that activation of PI3K is, like MAPK, necessary (but not sufficient) for
[51]
Ras in human primary thyroid epithelial cells
653
Ras‐induced clonal expansion in thyrocytes, in this case caused, at least in part, by prevention of Ras‐induced apoptosis. The results also suggest that pharmacological inhibitors of PI3K may be potential therapeutic agents for tumors driven by Ras mutation. RalGEF Pathway As discussed previously, Raf and PI3K activation together did not reproduce the full proliferative response of RasV12. Another major downstream Ras effector is RalGEF (Marshall, 1996; Wolthuis and Bos, 1999); we sought, therefore, to identify the involvement of RalGEF in Ras‐ induced proliferation in thyroid cells. RalBD pull‐down assay carried out with Ras‐infected thyrocytes showed a fivefold increase in Ral‐GTP binding than with normal thyrocytes (Bounacer et al., 2004). However, RasV12/ G37 mutant was not able to initiate proliferation in normal thyroid cells. Coinfection of primary thyrocytes with RasV12 and the dominant‐negative Ral construct (RalN28) yielded 2.35 times less colonies than RasV12 and an empty vector. The growing doubly infected colonies had altered morphology with marked vacuolation and poorly defined colony edges. Because the efficiency of multiple retroviral infection is very low in primary thyrocytes, we used scrape‐loading to assess the combined effects of all three Ras effector mutants. BrdU LI was analyzed after 48 h, showing that scrape‐loading of thyrocytes with the three effector mutants together leads to a BrdU LI more than 80% of that seen with RasV12 (Fig. 2). Taken together, our data demonstrate the necessity of all three Ras effector pathways for RAS‐induced proliferation and colony formation of primary thyroid epithelial cells.
Differential Response of Human Fibroblasts and Thyrocytes to Mutant Ras Oncoprotein
Many types of human cancers show mutations in the ras family of oncogenes in vivo (Bos, 1989). However, most in vitro studies using cell cultures demonstrate the largely growth inhibitory effect of mutant ras, explained by the induction of cell cycle inhibitors p16 and/or p21 (Serrano et al., 1997; Wei et al., 2001). Most of these in vitro studies use either fibroblasts (Hahn et al., 1999), breast epithelial cells (Elenbaas et al., 2001), or astrocytes (Rich et al., 2001), naturally accruing tumors that seldom have Ras oncogene mutations (Bos, 1989; Bredel and Pollack, 1999). Follicular tumors of the thyroid, however, demonstrate the high frequency of Ras mutations (50%) in both benign and malignant stages (Lemoine et al., 1989;
654
regulators and effectors of small GTPases: Ras family
[51]
FIG. 2. Scrape‐loading of the three Ras effector mutants mimics the proliferative effect of mutant Ras expression; 2‐day thyrocyte cultures were scrape‐loaded with 1 g/ml purified recombinant RasV12 or combinations of effector mutants and reseeded onto poly‐D‐lysine– coated dishes. The BrdU labeling index was measured 48 h after scrape‐loading. The data shown are the means of two independent experiments (Adapted from Bounacer et al., 2004).
Suarez et al., 1990), which is matched with the in vitro cell culture data, where mutant Ras induces long‐term proliferation in thyroid follicular epithelial cells (as discussed in previous sections). The paradoxical issue of the growth‐inhibitory effect of mutant Ras in fibroblasts (Serrano et al., 1997; Wei et al., 2001) and proliferogenic response in primary thyrocytes could potentially be explained by varying levels of Ras expression because of different expression systems used in different laboratories. To address that, we exploited the microinjection approach to deliver controlled levels of recombinant Ras protein into fibroblasts and thyrocytes (Skinner et al., 2004). Monolayers of normal human fibroblasts (HCA2) or thyroid epithelial cells were microinjected with varying concentrations of recombinant mutant RasV12 protein (together with rabbit IgG as a ‘‘marker’’). Proliferative response (nuclear DNA synthesis) was assessed 48 h later by labeling cultures with BrdU.
[51]
Ras in human primary thyroid epithelial cells
655
Microinjection of recombinant mutant Ras proteins induced a concentration‐dependent stimulation of DNA synthesis in thyrocytes, matched by a reciprocal inhibition in fibroblasts. Induction of p21 reached similar high levels in both cell types, but p16 was rapidly induced only in fibroblast (Skinner et al., 2004). We initially speculated that this accounted for the fibroblast specificity of growth inhibition, but we subsequently showed that Ras inhibits proliferation to the same extent and over the same dose range in p16‐deficient fibroblasts (‘‘Leiden’’ strain) (Skinner et al., 2004). Our current hypothesis, therefore, is that Ras signaling can overcome the growth inhibitory action of p21 in thyrocytes but not in fibroblasts. In summary, therefore, by use of an approach that allows direct control of protein levels, we have shown that RasV12 induces a concentration‐ dependent stimulation of proliferation in thyroid epithelial cells that is matched by a reciprocal inhibition of proliferation in fibroblasts. Importantly, these opposing effects are unidirectional (with no evidence of a ‘‘bell‐shaped’’ curve) and are observed over very similar concentration ranges. Methods 1: Disaggregation of Thyroid Tissue to Produce Primary Monolayer Cultures of Follicular Epithelium
1. Rinse freshly collected histologically normal human thyroid tissue twice in HBSS to remove blood and contaminants. Transfer the tissue to a Petri dish containing a few milliliters of ice‐cold HBSS. 2. Trim off any connective tissue and discard. 3. Mince the thyroid tissue with sterile ‘‘crossed’’ scalpel blades as finely as possible, ensuring that the tissue does not dry out. 4. Transfer the fragments into a 25‐ml universal container (for up to 2 g of tissue) and wash with ice‐cold serum‐free RPMI to remove as much blood as possible (two to three times), allowing tissue fragments to sediment under gravity and carefully aspirating the supernatant. If necessary, at this stage, the process can be suspended overnight. Fill the container with RPMI, seal, and keep on ice until restart. 5. Wash the fragments in a minimal volume of enzyme mixture (50 mg of collagenase (Boehringer No. 1088793) and 60 mg of Dispase (Boehringer No. 165859) in 60 ml serum‐free RPMI, prepared fresh, filter‐ sterilized, and warmed up to 37 ). 6. Resuspend in 10 ml of prewarmed enzyme mixture and place in 37 waterbath. Remove tube every 15 min and agitate gently. 7. After 1 h, harvest the first fraction. Remove the tube from the water bath, wipe with 70% ethanol, and agitate for 20 sec.
656
regulators and effectors of small GTPases: Ras family
[51]
8. Allow undigested tissue fragments to sediment under gravity. Carefully remove and save the supernatant (containing single cells and follicles) using a plastic pipette into a 15‐ml centrifuge tube. The strands of connective tissue occasionally contaminating the supernatant can be removed by stirring it with a glass Pasteur pipette. 9. Add 10 volumes of enzyme mixture to the remaining tissue fragments and continue the incubation. 10. Harvest at 30‐min intervals until disaggregation is complete. 11. While the digestion of the next fraction is proceeding, wash the latest fraction to remove enzyme by centrifuging at 200g (1000 rpm in the bench top centrifuge) for 2 min. Discard the supernatant. 12. Resuspend in 1.5 ml RPMI þ 05% newborn calf serum (NBCS, reduces clumping). Resuspension should be achieved by ‘‘flicking’’ the tube and not by pipetting, which will result in greater cell loss. 13. Take a small sample (10 l) for examination by phase‐contrast microscopy to assess the progress of follicle disaggregation. The content of follicles should reach a maximum from fraction 3 onwards. Digestion is normally complete in 3–4 h. 14. Allow the suspended mixtures of single cells and follicles to sediment on ice for at least 45 min. 15. Carefully remove most of the supernatant containing single cells and erythrocytes and discard. 16. Progressively pool the remaining pallets as successive fractions are processed. Make up to 5 ml with RPMI (containing 0.5% NBCS), rinse the tubes with a further 5 ml, and add to the first. Centrifuge the cell suspension 200g (1000 rpm bench‐top) for 3 min. Discard the supernatant. 17. At this stage, cells can be frozen in 1:1 mixture of RPMI/10% NBCS and freezing mixture (DMSO with NBCS at 1:4).
Methods 2: The Stable Expression of Mutant Ras in Normal Thyroid Epithelial Cells Using Retroviral Vectors
Production of the Amphotropic Vector To generate an amphotropic vector expressing mutant human ras, the amphotropic packaging line psi‐CRIP was transduced with ecotropic virus from the producer line psi‐2‐DOEJ (Wolthuis and Bos, 1999). This codes for RasV12 driven by the Moloney murine leukemia virus–long terminal repeat of the ‘‘defective’’ retroviral vector DOL. Producer clones were
[51]
Ras in human primary thyroid epithelial cells
657
assessed for viral titer by the ability to transduce G418‐resistance using the human epithelial A431 cells as a target. The highest titer producer (5 105 colony forming units/ml) was used for subsequent work. Retroviral Transduction Protocol 1. The producer cells should be more than 95% confluent for best results. Maintain the producer cells in selective medium and preferably in a flask for safety reasons. 2. 12–18 h before infection, remove the medium and wash the cells in prewarmed HBSS or medium, before adding prewarmed harvest medium without the selective agent. 3. Plate primary thyroid epithelial cultures at 5 105 cells per 60‐mm dish and allow at least 48 h to attach before the infection with the retrovirus vector psi‐CRIP‐DOEJ. 4. One hour before the infection, refeed the cells with fresh medium containing 8 g/ml polybrene (Sigma). 5. Harvest the medium and centrifuge for 5 min at 1000 rpm to sediment cell debris. 6. Filter the viral supernatant through the 0.45‐m membrane filter and add polybrene to a final concentration 8 g/ml. 7. Remove the medium from the target cells and replace with 2 ml of retrovirus‐containing medium per 60‐ml dish. 8. Return to the 37 incubator for 2–3 h. 9. Then add 3 ml (for a 60‐ml dish) of the medium used to culture the target cells. 10. Refeed the cells with nonselective medium 18–24 h after infection. 11. 48 hours after infection, pass the cells and maintain in medium with or without G418 (400 g/ml). Typically, each 60‐mm dish can be split into three. Methods for Transient Expression of Mutant Ras in Thyroid Epithelial Cells Scrape‐Loading 1. Plate primary cultures of thyroid epithelium at 5 105 cells per 60‐mm dish. 2. Replace the medium 3 days later with 150 ml buffer (10 mM Tris, pH 7, 114 mM KCl, 15 mM NaCl, 5.5 mM MgCl2) containing IgG (control) or recombinant Ras protein, produced as described in Trahey et al. (1987).
658
regulators and effectors of small GTPases: Ras family
[51]
3. Detach cells by gentle scraping with a ‘‘rubber policeman’’ (Leevers and Marshall, 1992), and after 1 min reseed onto poly‐D‐lysine (Sigma) dishes in complete medium (or maintained in suspension for analysis of early time points). Microinjection 1. Plate thyrocytes in 60‐mm dishes 24–48 h before microinjection. 2. Mix recombinant proteins or affinity‐purified rat immunoglobulin (IgG) (1 mg/ml) with nonimmune rabbit IgG (5 mg/ml) in 10 mM Tris‐HCl, pH 7.5, 114 mM KCl, 15 mM NaCl, and 5 mM MgCl2. 3. Inject approximately 20 femto mol of protein solution into the cytoplasm of cells within a marked area of the culture dish, using an Eppendorf system (micromanipulator 5171, transjector 5142; Carl Zeiss, Oberkochen, Germany) mounted on a Zeiss microscope.
Summary
In summary, we showed that the main candidate initiating event—Ras mutation—generates a phase of clonal expansion in vitro with retention of thyroid‐specific gene expression, a phenotype highly consistent with that of the first stage of tumorigenesis in vivo follicular adenoma. We have exploited this model to dissect the signaling pathways required for Ras‐ induced proliferation in these cells, revealing important differences from other more ‘‘convenient’’ experimental models based on fibroblasts or rodent cell lines, thus highlighting the importance of studying Ras signaling in the appropriate cell context.
References Bond, J. A., Wyllie, F. S., Rowson, J., Radulescu, A., and Wynford‐Thomas, D. (1994). In vitro reconstruction of tumour initiation in a human epithelium. Oncogene 9, 281–290. Bos, J. L. (1989). Ras oncogenes in human cancer: A review. Cancer Res. 49, 4682–4689. Bounacer, A., McGregor, A., Skinner, J., Bond, J., Poghosyan, Z., and Wynford‐Thomas, D. (2004). Mutant ras‐induced proliferation of human thyroid epithelial cells requires three effector pathways. Oncogene 23, 7839–7845. Bredel, M., and Pollack, I. F. (1999). The p21‐Ras signal transduction pathway and growth regulation in human high‐grade gliomas. Brain Res. Brain Res. Rev. 29, 232–249. Cowley, S., Paterson, H., Kemp, P., and Marshall, C. J. (1994). Activation of MAP kinase kinase is necessary and sufficient for PC12 differentiation and for transformation of NIH 3T3 cells. Cell 77, 841–852. Elenbaas, B., Spirio, L., Koerner, F., Fleming, M. D., Zimonjic, D. B., Donaher, J. L., Popescu, N. C., Hahn, W. C., and Weinberg, R. A. (2001). Human breast cancer cells
[51]
Ras in human primary thyroid epithelial cells
659
generated by oncogenic transformation of primary mammary epithelial cells. Genes Dev. 15, 50–65. Francis‐Lang, H., Zannini, M., De Felice, M., Berlingieri, M. T., Fusco, A., and Di Lauro, R. (1992). Multiple mechanisms of interference between transformation and differentiation in thyroid cells. Mol. Cell. Biol. 12, 5793–5800. Gire, V., Marshall, C., and Wynford‐Thomas, D. (2000). PI‐3‐kinase is an essential antiapoptotic effector in the proliferative response of primary human epithelial cells to mutant RAS. Oncogene 19, 2269–2276. Gire, V., Marshall, C. J., and Wynford‐Thomas, D. (1999). Activation of mitogen‐activated protein kinase is necessary but not sufficient for proliferation of human thyroid epithelial cells induced by mutant Ras. Oncogene 18, 4819–4832. Gire, V., and Wynford‐Thomas, D. (2000). RAS oncogene activation induces proliferation in normal human thyroid epithelial cells without loss of differentiation. Oncogene 19, 737–744. Hahn, W. C., Counter, C. M., Lundberg, A. S., Beijersbergen, R. L., Brooks, M. W., and Weinberg, R. A. (1999). Creation of human tumour cells with defined genetic elements. Nature 400, 464–468. Leevers, S. J., and Marshall, C. J. (1992). Activation of extracellular signal‐regulated kinase, ERK2, by p21ras oncoprotein. EMBO J. 11, 569–574. Lemoine, N. R., Mayall, E. S., Wyllie, F. S., Farr, C. J., Hughes, D., Padua, R. A., Thurston, V., Williams, E. D., and Wynford‐Thomas, D. (1988). Activated ras oncogenes in human thyroid cancers. Cancer Res. 48, 4459–4463. Lemoine, N. R., Mayall, E. S., Wyllie, F. S., Williams, E. D., Goyns, M., Stringer, B., and Wynford‐Thomas, D. (1989). High frequency of ras oncogene activation in all stages of human thyroid tumorigenesis. Oncogene 4, 159–164. Marshall, C. J. (1996). Ras effectors. Curr. Opin. Cell. Biol. 8, 197–204. Miller, M. J., Prigent, S., Kupperman, E., Rioux, L., Park, S. H., Feramisco, J. R., White, M. A., Rutkowski, J. L., and Meinkoth, J. L. (1997). RalGDS functions in Ras‐and cAMP‐ mediated growth stimulation. J. Biol. Chem. 272, 5600–5605. Miller, M. J., Rioux, L., Prendergast, G. V., Cannon, S., White, M. A., and Meinkoth, J. L. (1998). Differential effects of protein kinase A on Ras effector pathways. Mol. Cell. Biol. 18, 3718–3726. Rich, J. N., Guo, C., McLendon, R. E., Bigner, D. D., Wang, X. F., and Counter, C. M. (2001). A genetically tractable model of human glioma formation. Cancer Res. 61, 3556–3560. Serrano, M., Lin, A. W., McCurrach, M. E., Beach, D., and Lowe, S. W. (1997). Oncogenic ras provokes premature cell senescence associated with accumulation of p53 and p16INK4a. Cell 88, 593–602. Skinner, J., Bounacer, A., Bond, J. A., Haughton, M. F., deMicco, C., and Wynford‐Thomas, D. (2004). Opposing effects of mutant ras oncoprotein on human fibroblast and epithelial cell proliferation: Implications for models of human tumorigenesis. Oncogene 23, 5994–5999. Suarez, H. G., du Villard, J. A., Severino, M., Caillou, B., Schlumberger, M., Tubiana, M., Parmentier, C., and Monier, R. (1990). Presence of mutations in all three ras genes in human thyroid tumors. Oncogene 5, 565–570. Trahey, M., Milley, R. J., Cole, G. E., Innis, M., Paterson, H., Marshall, C. J., Hall, A., and McCormick, F. (1987). Biochemical and biological properties of the human N‐ras p21 protein. Mol. Cell. Biol. 7, 541–544. Wei, W., Hemmer, R. M., and Sedivy, J. M. (2001). Role of p14(ARF) in replicative and induced senescence of human fibroblasts. Mol. Cell. Biol. 21, 6748–6757.
660
regulators and effectors of small GTPases: Ras family
[52]
Wolthuis, R. M., and Bos, J. L. (1999). Ras caught in another affair: The exchange factors for Ral. Curr. Opin. Genet. Dev. 9, 112–117. Wynford‐Thomas, D. (1993). Molecular basis of epithelial tumorigenesis: The thyroid model. Crit. Rev. Oncog. 4, 1–23. Wynford‐Thomas, D. (1997). Origin and progression of thyroid epithelial tumours: Cellular and molecular mechanisms. Horm. Res. 47, 145–157.
[52] Use of Ras‐Transformed Human Ovarian Surface Epithelial Cells as a Model for Studying Ovarian Cancer By DANIEL G. ROSEN,* GONG YANG,* ROBERT C. BAST, JR., and JINSONG LIU Abstract
The Ras gene family has been implicated in the development of many human epithelial cancers. Mutations in K‐ras or its downstream mediator BRAF have been detected in about two thirds of low‐grade serous carcinomas and borderline serous tumors; mutations in K‐ras are also often present in benign and invasive mucinous ovarian cancers. Although the oncogenic allele H‐rasV12 is present in only approximately 6% of ovarian cancers, physiologically activated H‐ras protein is commonly detected in human ovarian cancer, presumably because of an increase in upstream signals from tyrosine kinase growth factor receptors such as Her‐2/neu, despite the lack of a Ras mutation. The mechanisms by which ras oncogenes transform human epithelial cells are not clear. The methods described here are what we use to culture human ovarian surface epithelial cells, to immortalize those cells, and to transform the immortalized cells with oncogenic H‐ras or K‐ras. These Ras‐transformed human ovarian surface epithelial cells form tumors in nude mice and recapitulate many features of human ovarian cancer, thus providing an excellent model system for studying the initiation and progression of human ovarian cancer. Introduction
Ovarian cancer is the most lethal form of cancer among women in the United States, accounting for more than 25,000 new cases and approximately 16,000 deaths in 2004 (Jemal et al., 2004). Ras genes encode highly * These two authors contributed equally to this work. METHODS IN ENZYMOLOGY, VOL. 407 Copyright 2006, Elsevier Inc. All rights reserved.
0076-6879/06 $35.00 DOI: 10.1016/S0076-6879(05)07052-7
regulators and effectors of small GTPases: Ras family
[51]
[51] Analysis of Ras Transformation of Human Thyroid Epithelial Cells By ZARUHI POGHOSYAN and DAVID WYNFORD‐THOMAS Abstract
Activation of Ras oncogene by point mutations is an early frequent event in thyroid tumorigenesis. In this chapter, we describe the use of human primary thyroid follicular epithelial cells expressing oncogenic mutant Ras by means of retroviral transduction as a biological model of human cancer initiation that provides powerful insights into thyroid tumorigenesis. We describe protocols for manipulating primary epithelial cells and describe the use of this model to dissect the signaling pathways required for Ras‐induced proliferation in these cells. We also highlight the importance of studying Ras signaling in an appropriate cell context, summarizing some of the key differences identified between more widespread experimental models based on fibroblasts or rodent cell lines and primary epithelial cells.
Introduction
The follicular epithelial cells of human thyroid give rise to a range of pathologically well‐defined tumor phenotypes and stages, and are proving to be one of the most informative models for studying the molecular basis of multistage human tumorigenesis in ‘‘conditional renewal’’–type epithelium. Early thyroid tumor development is closely correlated with mutation of a number of alternative genes, such as ras, ret, trk, gsp, and the TSH receptor, each associated with different tumour phenotypes, presenting a good example of genotype/phenotype association. Genetic analyses of clinical samples by our (Lemoine et al., 1988, 1989) and many other laboratories revealed that independent of clinical stage, thyroid tumors of follicular type showed a high frequency of mutation of the Ras family of oncogenes, whereas those of papillary type show a predominance of either Ret/PTC1 rearrangements or BRAF mutations. Thyroid offers a distinct advantage for modeling tumor development in tissue culture. First, near‐pure cultures of normal thyroid epithelium can be relatively easily obtained from surgical material, which, even in a simple monolayer culture, remain viable with the same simple ‘‘conditional‐ renewal’’ cell kinetics as seen in the intact gland. This contrasts with the METHODS IN ENZYMOLOGY, VOL. 407 Copyright 2006, Elsevier Inc. All rights reserved.
0076-6879/06 $35.00 DOI: 10.1016/S0076-6879(05)07051-5
[51]
Ras in human primary thyroid epithelial cells
649
technical complexity and uncertainty of interpretation involved in modeling the hierarchy of differentiation states seen in a renewing epithelium such as colon. Second, tumor initiation in the thyroid results from inappropriate induction of proliferation in what is otherwise a virtually quiescent cell population, again potentially simple to model in vitro in contrast to renewing tissues in which loss of differentiation may be the initiating event. A major principle of our work is that models of ras‐induced tumorigenesis must be based on a cell type in whose tumors Ras mutation occurs commonly and at an early stage and in which it can be shown to stimulate proliferation in vitro. By use of retroviral vectors and/or microinjection to overcome the difficulty of gene transfer into normal epithelial cells, we have exploited this system to reconstruct the initial steps in thyroid tumorigenesis. We believe that we have developed what is still a unique in vitro model of tumor initiation in a human epithelial cell that demonstrates the determining influence of the nature of the initiating oncogene on tumor phenotype. Overview of Biological Results Obtained with the In Vitro Thyroid Model and Effect of Mutant RAS on Thyroid‐Specific Differentiation
Most other in vitro studies of neoplastic transformation in the thyroid have used immortal but untransformed rodent epithelial cell lines. In these models, expression of mutant Ras leads to loss of tissue‐specific differentiation (Francis‐Lang et al., 1992; Miller et al., 1998) in contrast to the evidence from clinical analyses, implicating Ras mutation as an early event in human thyroid tumor development at a stage before loss of differentiation (Wynford‐Thomas, 1993, 1997). Human primary thyroid epithelial cells are capable of only two to three population doublings (PD) before entering a state of normally irreversible quiescence. The stable expression of mutant RasV12 induces a dramatic proliferative response, resulting in generation of colonies, the final size of which can be up to 107 cells. These colonies show a normal epithelial phenotype and expanding pattern of growth (Bond et al., 1994). However, this proliferation spontaneously ceases after 20–25 PDs, terminating in a viable state of growth arrest, resembling replicative senescence. Because of the starting low proliferative capacity of primary human thyrocytes, the number of colonies obtained after retroviral Ras expression is very low, 50–100 colonies per 105 cells (Fig. 1). To resolve this controversy, we examined the short‐ and long‐term responses of normal human thyroid epithelial cells to RasV12, introduced by microinjection (see ‘‘Methods, 3—Microinjection’’) and retroviral
650
regulators and effectors of small GTPases: Ras family
[51]
FIG. 1. Retroviral transduction of mutant Ras induces proliferation of human primary thyroid epithelial cells. Schematic diagram of a typical outcome of retroviral transduction. As a result of very low proliferative capacity of primary human thyrocytes, generally not more than 50–100 colonies are obtained per 105 cells. Representative photomicrographs of normal thyroid epithelial cells and colonies induced by mutant Ras at an early rapidly proliferating stage and at a late stage at the end of their proliferative life span.
transduction (see ‘‘Methods, 2—Retroviral Transduction Protocol’’), respectively (Gire and Wynford‐Thomas, 2000). In both cases, expression of RasV12 at a level sufficient to induce rapid proliferation did not lead to loss of differentiation, as shown by expression of cytokeratin 18, E‐cadherin, thyroglobulin, TTF‐1, and Pax‐8 proteins. Indeed, Ras was able to prevent, and even to reverse, the loss of thyroglobulin expression that occurs normally in TSH‐deficient culture medium. These responses were partially mimicked by activation of Raf, a major Ras effector, indicating involvement of the MAPK signal pathway. This striking contrast between the effect of RasV12 on differentiation in primary human, compared with immortalized rodent, epithelial cultures is most likely explained by the influence of additional cooperating abnormalities in the latter and again highlights the need for caution in extrapolating from cell line data.
[51]
Ras in human primary thyroid epithelial cells
651
Analysis of Ras Effector Pathways Mediating Mutant Ras‐Induced Proliferation of Primary Human Thyroid Epithelial Cells: A Model of Tumor Initiation
Extensive studies of Ras signaling in the past few years identified that at least three effectors are involved in RAS‐mediated cell proliferation: Raf, PI3K, Ral GDS (Marshall, 1996; Wolthuis and Bos, 1999). However, the particular necessary combination of downstream pathways can be cell type‐ and species‐specific. By use of human primary cultures of normal thyroid epithelial cells as a relevant model, we identified the required effector pathways driving RasV12‐induced proliferation in this cell type. Some of the key findings are summarized in the following. RAF‐MAPK Pathway By use of a combination of transient (scrape‐loading or microinjection, see ‘‘Methods, 3’’) or sustained (retrovirally mediated, see ‘‘Methods, 2’’) expression approaches, we first demonstrated (Gire et al., 1999) that mutant RAS induces a rapid activation of MAPK (predominantly ERK2) as assessed by phosphorylation of the enzyme and by in vitro kinase assay, together with nuclear translocation demonstrated by immunofluorescence. Importantly, this was sustained throughout the period of Ras‐induced proliferation, in contrast to the more transient activation in response to growth factor stimulation of thyroid cells. Such differences in kinetics have been found to be crucial for determining the resulting phenotype and, in this case, may contribute to the ability of Ras to induce sustained proliferation for at least 20–25 population doublings (PD), whereas the response to growth factors is limited to just 1 or 2 PD. We next determined whether this activation of MAPK is necessary for the proliferative response using two approaches for inhibiting the pathway at the level of MAPKK. Coexpression of a dominant‐negative A217‐MAPKK inhibited Ras‐induced proliferation as shown by a 60% reduction in colony yield. (The incompleteness of this effect is explicable by the partial inhibition of MAPK activity observed in a thyroid cell line test system (HT‐ori‐3) and could reflect failure to reach a level of expression sufficient to totally block wild‐type MAPKK interaction with Raf). A more complete inhibition of MAPK activity was achieved using the pharmacological inhibitor of MAPKK, PD98059, resulting in virtually complete inhibition of Ras‐induced epithelial colony formation and inhibition of proliferation in preformed colonies. This effectiveness of PD98059 was also important in excluding effects on other pathways, because it seems to be specific for MAPKK at the concentrations used, unlike the
652
regulators and effectors of small GTPases: Ras family
[51]
dominant‐negative mutant that acts by binding to Raf and could potentially block other Raf‐activated pathways. To determine whether activation of MAPK was sufficient as well as necessary for response to Ras, we used a constitutively active mutant of its immediate upstream partner MAPKK (Glu‐217/Glu‐221) to minimize the chance of stimulating additional pathways. In contrast to a previous study (Cowley et al., 1994), no evidence of proliferation in response to this vector was observed. Although it is known that the biochemical activity of the Glu‐217/Glu‐221 mutant is much lower than the physiologically activated enzyme, we showed, using a thyroid cell line (HT‐ori‐3), the level of expression from the retrovirus vector was sufficient to give activation of MAPK of comparable magnitude to that produced by RasV12. Although this cannot be checked directly in primary cells, it suggests that failure to stimulate proliferation with this vector is not a ‘‘false‐negative’’ result and that activation of MAPK is, therefore, insufficient by itself. We conclude that activation of the MAPK pathway is necessary, but not sufficient, for the proliferogenic action of RasV12 on primary human thyroid cells (Gire et al., 1999). This contrasts with results from the model closest to our own—the rat thyroid cell line WRT—in which MAPK activation seems to be dispensable for Ras‐induced mitogenesis which (Miller et al., 1997, 1998). This emphasizes the risk of extrapolation from rodent cell lines to normal human cells and is particularly important here, because the design of therapeutic strategies targeting Ras will be significantly influenced by the degree of redundancy in RAS signaling pathways. PI3K Pathway With similar approaches, we next examined the role of another major Ras effector—the PI3K pathway (Gire et al., 2000). Following the same strategy as with MAPK, we first showed by use of scrape‐loading that mutant RasV12 activated the PI3K pathway in normal human thyrocytes as revealed by phosphorylation of its downstream target, PKB/Akt. Next, by use of both an effector mutant of RasV12 (C40) specific for PI3K and a constitutively active subunit of PI3K itself (p110*), we showed that activation of this pathway alone was not sufficient to induce proliferation in normal thyrocytes. Coinfection with retroviral vectors expressing a constitutively active MAPKK and p110* did, however, generate a few small colonies, demonstrating a weak synergy between the two pathways but insufficient to fully mimic the effect of Ras. Finally, inhibition of PI3K enzyme activity by LY294002 blocked Ras‐induced colony formation and induced apoptosis in preformed colonies. These data show that activation of PI3K is, like MAPK, necessary (but not sufficient) for
[51]
Ras in human primary thyroid epithelial cells
653
Ras‐induced clonal expansion in thyrocytes, in this case caused, at least in part, by prevention of Ras‐induced apoptosis. The results also suggest that pharmacological inhibitors of PI3K may be potential therapeutic agents for tumors driven by Ras mutation. RalGEF Pathway As discussed previously, Raf and PI3K activation together did not reproduce the full proliferative response of RasV12. Another major downstream Ras effector is RalGEF (Marshall, 1996; Wolthuis and Bos, 1999); we sought, therefore, to identify the involvement of RalGEF in Ras‐ induced proliferation in thyroid cells. RalBD pull‐down assay carried out with Ras‐infected thyrocytes showed a fivefold increase in Ral‐GTP binding than with normal thyrocytes (Bounacer et al., 2004). However, RasV12/ G37 mutant was not able to initiate proliferation in normal thyroid cells. Coinfection of primary thyrocytes with RasV12 and the dominant‐negative Ral construct (RalN28) yielded 2.35 times less colonies than RasV12 and an empty vector. The growing doubly infected colonies had altered morphology with marked vacuolation and poorly defined colony edges. Because the efficiency of multiple retroviral infection is very low in primary thyrocytes, we used scrape‐loading to assess the combined effects of all three Ras effector mutants. BrdU LI was analyzed after 48 h, showing that scrape‐loading of thyrocytes with the three effector mutants together leads to a BrdU LI more than 80% of that seen with RasV12 (Fig. 2). Taken together, our data demonstrate the necessity of all three Ras effector pathways for RAS‐induced proliferation and colony formation of primary thyroid epithelial cells.
Differential Response of Human Fibroblasts and Thyrocytes to Mutant Ras Oncoprotein
Many types of human cancers show mutations in the ras family of oncogenes in vivo (Bos, 1989). However, most in vitro studies using cell cultures demonstrate the largely growth inhibitory effect of mutant ras, explained by the induction of cell cycle inhibitors p16 and/or p21 (Serrano et al., 1997; Wei et al., 2001). Most of these in vitro studies use either fibroblasts (Hahn et al., 1999), breast epithelial cells (Elenbaas et al., 2001), or astrocytes (Rich et al., 2001), naturally accruing tumors that seldom have Ras oncogene mutations (Bos, 1989; Bredel and Pollack, 1999). Follicular tumors of the thyroid, however, demonstrate the high frequency of Ras mutations (50%) in both benign and malignant stages (Lemoine et al., 1989;
654
regulators and effectors of small GTPases: Ras family
[51]
FIG. 2. Scrape‐loading of the three Ras effector mutants mimics the proliferative effect of mutant Ras expression; 2‐day thyrocyte cultures were scrape‐loaded with 1 g/ml purified recombinant RasV12 or combinations of effector mutants and reseeded onto poly‐D‐lysine– coated dishes. The BrdU labeling index was measured 48 h after scrape‐loading. The data shown are the means of two independent experiments (Adapted from Bounacer et al., 2004).
Suarez et al., 1990), which is matched with the in vitro cell culture data, where mutant Ras induces long‐term proliferation in thyroid follicular epithelial cells (as discussed in previous sections). The paradoxical issue of the growth‐inhibitory effect of mutant Ras in fibroblasts (Serrano et al., 1997; Wei et al., 2001) and proliferogenic response in primary thyrocytes could potentially be explained by varying levels of Ras expression because of different expression systems used in different laboratories. To address that, we exploited the microinjection approach to deliver controlled levels of recombinant Ras protein into fibroblasts and thyrocytes (Skinner et al., 2004). Monolayers of normal human fibroblasts (HCA2) or thyroid epithelial cells were microinjected with varying concentrations of recombinant mutant RasV12 protein (together with rabbit IgG as a ‘‘marker’’). Proliferative response (nuclear DNA synthesis) was assessed 48 h later by labeling cultures with BrdU.
[51]
Ras in human primary thyroid epithelial cells
655
Microinjection of recombinant mutant Ras proteins induced a concentration‐dependent stimulation of DNA synthesis in thyrocytes, matched by a reciprocal inhibition in fibroblasts. Induction of p21 reached similar high levels in both cell types, but p16 was rapidly induced only in fibroblast (Skinner et al., 2004). We initially speculated that this accounted for the fibroblast specificity of growth inhibition, but we subsequently showed that Ras inhibits proliferation to the same extent and over the same dose range in p16‐deficient fibroblasts (‘‘Leiden’’ strain) (Skinner et al., 2004). Our current hypothesis, therefore, is that Ras signaling can overcome the growth inhibitory action of p21 in thyrocytes but not in fibroblasts. In summary, therefore, by use of an approach that allows direct control of protein levels, we have shown that RasV12 induces a concentration‐ dependent stimulation of proliferation in thyroid epithelial cells that is matched by a reciprocal inhibition of proliferation in fibroblasts. Importantly, these opposing effects are unidirectional (with no evidence of a ‘‘bell‐shaped’’ curve) and are observed over very similar concentration ranges. Methods 1: Disaggregation of Thyroid Tissue to Produce Primary Monolayer Cultures of Follicular Epithelium
1. Rinse freshly collected histologically normal human thyroid tissue twice in HBSS to remove blood and contaminants. Transfer the tissue to a Petri dish containing a few milliliters of ice‐cold HBSS. 2. Trim off any connective tissue and discard. 3. Mince the thyroid tissue with sterile ‘‘crossed’’ scalpel blades as finely as possible, ensuring that the tissue does not dry out. 4. Transfer the fragments into a 25‐ml universal container (for up to 2 g of tissue) and wash with ice‐cold serum‐free RPMI to remove as much blood as possible (two to three times), allowing tissue fragments to sediment under gravity and carefully aspirating the supernatant. If necessary, at this stage, the process can be suspended overnight. Fill the container with RPMI, seal, and keep on ice until restart. 5. Wash the fragments in a minimal volume of enzyme mixture (50 mg of collagenase (Boehringer No. 1088793) and 60 mg of Dispase (Boehringer No. 165859) in 60 ml serum‐free RPMI, prepared fresh, filter‐ sterilized, and warmed up to 37 ). 6. Resuspend in 10 ml of prewarmed enzyme mixture and place in 37 waterbath. Remove tube every 15 min and agitate gently. 7. After 1 h, harvest the first fraction. Remove the tube from the water bath, wipe with 70% ethanol, and agitate for 20 sec.
656
regulators and effectors of small GTPases: Ras family
[51]
8. Allow undigested tissue fragments to sediment under gravity. Carefully remove and save the supernatant (containing single cells and follicles) using a plastic pipette into a 15‐ml centrifuge tube. The strands of connective tissue occasionally contaminating the supernatant can be removed by stirring it with a glass Pasteur pipette. 9. Add 10 volumes of enzyme mixture to the remaining tissue fragments and continue the incubation. 10. Harvest at 30‐min intervals until disaggregation is complete. 11. While the digestion of the next fraction is proceeding, wash the latest fraction to remove enzyme by centrifuging at 200g (1000 rpm in the bench top centrifuge) for 2 min. Discard the supernatant. 12. Resuspend in 1.5 ml RPMI þ 05% newborn calf serum (NBCS, reduces clumping). Resuspension should be achieved by ‘‘flicking’’ the tube and not by pipetting, which will result in greater cell loss. 13. Take a small sample (10 l) for examination by phase‐contrast microscopy to assess the progress of follicle disaggregation. The content of follicles should reach a maximum from fraction 3 onwards. Digestion is normally complete in 3–4 h. 14. Allow the suspended mixtures of single cells and follicles to sediment on ice for at least 45 min. 15. Carefully remove most of the supernatant containing single cells and erythrocytes and discard. 16. Progressively pool the remaining pallets as successive fractions are processed. Make up to 5 ml with RPMI (containing 0.5% NBCS), rinse the tubes with a further 5 ml, and add to the first. Centrifuge the cell suspension 200g (1000 rpm bench‐top) for 3 min. Discard the supernatant. 17. At this stage, cells can be frozen in 1:1 mixture of RPMI/10% NBCS and freezing mixture (DMSO with NBCS at 1:4).
Methods 2: The Stable Expression of Mutant Ras in Normal Thyroid Epithelial Cells Using Retroviral Vectors
Production of the Amphotropic Vector To generate an amphotropic vector expressing mutant human ras, the amphotropic packaging line psi‐CRIP was transduced with ecotropic virus from the producer line psi‐2‐DOEJ (Wolthuis and Bos, 1999). This codes for RasV12 driven by the Moloney murine leukemia virus–long terminal repeat of the ‘‘defective’’ retroviral vector DOL. Producer clones were
[51]
Ras in human primary thyroid epithelial cells
657
assessed for viral titer by the ability to transduce G418‐resistance using the human epithelial A431 cells as a target. The highest titer producer (5 105 colony forming units/ml) was used for subsequent work. Retroviral Transduction Protocol 1. The producer cells should be more than 95% confluent for best results. Maintain the producer cells in selective medium and preferably in a flask for safety reasons. 2. 12–18 h before infection, remove the medium and wash the cells in prewarmed HBSS or medium, before adding prewarmed harvest medium without the selective agent. 3. Plate primary thyroid epithelial cultures at 5 105 cells per 60‐mm dish and allow at least 48 h to attach before the infection with the retrovirus vector psi‐CRIP‐DOEJ. 4. One hour before the infection, refeed the cells with fresh medium containing 8 g/ml polybrene (Sigma). 5. Harvest the medium and centrifuge for 5 min at 1000 rpm to sediment cell debris. 6. Filter the viral supernatant through the 0.45‐m membrane filter and add polybrene to a final concentration 8 g/ml. 7. Remove the medium from the target cells and replace with 2 ml of retrovirus‐containing medium per 60‐ml dish. 8. Return to the 37 incubator for 2–3 h. 9. Then add 3 ml (for a 60‐ml dish) of the medium used to culture the target cells. 10. Refeed the cells with nonselective medium 18–24 h after infection. 11. 48 hours after infection, pass the cells and maintain in medium with or without G418 (400 g/ml). Typically, each 60‐mm dish can be split into three. Methods for Transient Expression of Mutant Ras in Thyroid Epithelial Cells Scrape‐Loading 1. Plate primary cultures of thyroid epithelium at 5 105 cells per 60‐mm dish. 2. Replace the medium 3 days later with 150 ml buffer (10 mM Tris, pH 7, 114 mM KCl, 15 mM NaCl, 5.5 mM MgCl2) containing IgG (control) or recombinant Ras protein, produced as described in Trahey et al. (1987).
658
regulators and effectors of small GTPases: Ras family
[51]
3. Detach cells by gentle scraping with a ‘‘rubber policeman’’ (Leevers and Marshall, 1992), and after 1 min reseed onto poly‐D‐lysine (Sigma) dishes in complete medium (or maintained in suspension for analysis of early time points). Microinjection 1. Plate thyrocytes in 60‐mm dishes 24–48 h before microinjection. 2. Mix recombinant proteins or affinity‐purified rat immunoglobulin (IgG) (1 mg/ml) with nonimmune rabbit IgG (5 mg/ml) in 10 mM Tris‐HCl, pH 7.5, 114 mM KCl, 15 mM NaCl, and 5 mM MgCl2. 3. Inject approximately 20 femto mol of protein solution into the cytoplasm of cells within a marked area of the culture dish, using an Eppendorf system (micromanipulator 5171, transjector 5142; Carl Zeiss, Oberkochen, Germany) mounted on a Zeiss microscope.
Summary
In summary, we showed that the main candidate initiating event—Ras mutation—generates a phase of clonal expansion in vitro with retention of thyroid‐specific gene expression, a phenotype highly consistent with that of the first stage of tumorigenesis in vivo follicular adenoma. We have exploited this model to dissect the signaling pathways required for Ras‐ induced proliferation in these cells, revealing important differences from other more ‘‘convenient’’ experimental models based on fibroblasts or rodent cell lines, thus highlighting the importance of studying Ras signaling in the appropriate cell context.
References Bond, J. A., Wyllie, F. S., Rowson, J., Radulescu, A., and Wynford‐Thomas, D. (1994). In vitro reconstruction of tumour initiation in a human epithelium. Oncogene 9, 281–290. Bos, J. L. (1989). Ras oncogenes in human cancer: A review. Cancer Res. 49, 4682–4689. Bounacer, A., McGregor, A., Skinner, J., Bond, J., Poghosyan, Z., and Wynford‐Thomas, D. (2004). Mutant ras‐induced proliferation of human thyroid epithelial cells requires three effector pathways. Oncogene 23, 7839–7845. Bredel, M., and Pollack, I. F. (1999). The p21‐Ras signal transduction pathway and growth regulation in human high‐grade gliomas. Brain Res. Brain Res. Rev. 29, 232–249. Cowley, S., Paterson, H., Kemp, P., and Marshall, C. J. (1994). Activation of MAP kinase kinase is necessary and sufficient for PC12 differentiation and for transformation of NIH 3T3 cells. Cell 77, 841–852. Elenbaas, B., Spirio, L., Koerner, F., Fleming, M. D., Zimonjic, D. B., Donaher, J. L., Popescu, N. C., Hahn, W. C., and Weinberg, R. A. (2001). Human breast cancer cells
[51]
Ras in human primary thyroid epithelial cells
659
generated by oncogenic transformation of primary mammary epithelial cells. Genes Dev. 15, 50–65. Francis‐Lang, H., Zannini, M., De Felice, M., Berlingieri, M. T., Fusco, A., and Di Lauro, R. (1992). Multiple mechanisms of interference between transformation and differentiation in thyroid cells. Mol. Cell. Biol. 12, 5793–5800. Gire, V., Marshall, C., and Wynford‐Thomas, D. (2000). PI‐3‐kinase is an essential antiapoptotic effector in the proliferative response of primary human epithelial cells to mutant RAS. Oncogene 19, 2269–2276. Gire, V., Marshall, C. J., and Wynford‐Thomas, D. (1999). Activation of mitogen‐activated protein kinase is necessary but not sufficient for proliferation of human thyroid epithelial cells induced by mutant Ras. Oncogene 18, 4819–4832. Gire, V., and Wynford‐Thomas, D. (2000). RAS oncogene activation induces proliferation in normal human thyroid epithelial cells without loss of differentiation. Oncogene 19, 737–744. Hahn, W. C., Counter, C. M., Lundberg, A. S., Beijersbergen, R. L., Brooks, M. W., and Weinberg, R. A. (1999). Creation of human tumour cells with defined genetic elements. Nature 400, 464–468. Leevers, S. J., and Marshall, C. J. (1992). Activation of extracellular signal‐regulated kinase, ERK2, by p21ras oncoprotein. EMBO J. 11, 569–574. Lemoine, N. R., Mayall, E. S., Wyllie, F. S., Farr, C. J., Hughes, D., Padua, R. A., Thurston, V., Williams, E. D., and Wynford‐Thomas, D. (1988). Activated ras oncogenes in human thyroid cancers. Cancer Res. 48, 4459–4463. Lemoine, N. R., Mayall, E. S., Wyllie, F. S., Williams, E. D., Goyns, M., Stringer, B., and Wynford‐Thomas, D. (1989). High frequency of ras oncogene activation in all stages of human thyroid tumorigenesis. Oncogene 4, 159–164. Marshall, C. J. (1996). Ras effectors. Curr. Opin. Cell. Biol. 8, 197–204. Miller, M. J., Prigent, S., Kupperman, E., Rioux, L., Park, S. H., Feramisco, J. R., White, M. A., Rutkowski, J. L., and Meinkoth, J. L. (1997). RalGDS functions in Ras‐and cAMP‐ mediated growth stimulation. J. Biol. Chem. 272, 5600–5605. Miller, M. J., Rioux, L., Prendergast, G. V., Cannon, S., White, M. A., and Meinkoth, J. L. (1998). Differential effects of protein kinase A on Ras effector pathways. Mol. Cell. Biol. 18, 3718–3726. Rich, J. N., Guo, C., McLendon, R. E., Bigner, D. D., Wang, X. F., and Counter, C. M. (2001). A genetically tractable model of human glioma formation. Cancer Res. 61, 3556–3560. Serrano, M., Lin, A. W., McCurrach, M. E., Beach, D., and Lowe, S. W. (1997). Oncogenic ras provokes premature cell senescence associated with accumulation of p53 and p16INK4a. Cell 88, 593–602. Skinner, J., Bounacer, A., Bond, J. A., Haughton, M. F., deMicco, C., and Wynford‐Thomas, D. (2004). Opposing effects of mutant ras oncoprotein on human fibroblast and epithelial cell proliferation: Implications for models of human tumorigenesis. Oncogene 23, 5994–5999. Suarez, H. G., du Villard, J. A., Severino, M., Caillou, B., Schlumberger, M., Tubiana, M., Parmentier, C., and Monier, R. (1990). Presence of mutations in all three ras genes in human thyroid tumors. Oncogene 5, 565–570. Trahey, M., Milley, R. J., Cole, G. E., Innis, M., Paterson, H., Marshall, C. J., Hall, A., and McCormick, F. (1987). Biochemical and biological properties of the human N‐ras p21 protein. Mol. Cell. Biol. 7, 541–544. Wei, W., Hemmer, R. M., and Sedivy, J. M. (2001). Role of p14(ARF) in replicative and induced senescence of human fibroblasts. Mol. Cell. Biol. 21, 6748–6757.
660
regulators and effectors of small GTPases: Ras family
[52]
Wolthuis, R. M., and Bos, J. L. (1999). Ras caught in another affair: The exchange factors for Ral. Curr. Opin. Genet. Dev. 9, 112–117. Wynford‐Thomas, D. (1993). Molecular basis of epithelial tumorigenesis: The thyroid model. Crit. Rev. Oncog. 4, 1–23. Wynford‐Thomas, D. (1997). Origin and progression of thyroid epithelial tumours: Cellular and molecular mechanisms. Horm. Res. 47, 145–157.
[52] Use of Ras‐Transformed Human Ovarian Surface Epithelial Cells as a Model for Studying Ovarian Cancer By DANIEL G. ROSEN,* GONG YANG,* ROBERT C. BAST, JR., and JINSONG LIU Abstract
The Ras gene family has been implicated in the development of many human epithelial cancers. Mutations in K‐ras or its downstream mediator BRAF have been detected in about two thirds of low‐grade serous carcinomas and borderline serous tumors; mutations in K‐ras are also often present in benign and invasive mucinous ovarian cancers. Although the oncogenic allele H‐rasV12 is present in only approximately 6% of ovarian cancers, physiologically activated H‐ras protein is commonly detected in human ovarian cancer, presumably because of an increase in upstream signals from tyrosine kinase growth factor receptors such as Her‐2/neu, despite the lack of a Ras mutation. The mechanisms by which ras oncogenes transform human epithelial cells are not clear. The methods described here are what we use to culture human ovarian surface epithelial cells, to immortalize those cells, and to transform the immortalized cells with oncogenic H‐ras or K‐ras. These Ras‐transformed human ovarian surface epithelial cells form tumors in nude mice and recapitulate many features of human ovarian cancer, thus providing an excellent model system for studying the initiation and progression of human ovarian cancer. Introduction
Ovarian cancer is the most lethal form of cancer among women in the United States, accounting for more than 25,000 new cases and approximately 16,000 deaths in 2004 (Jemal et al., 2004). Ras genes encode highly * These two authors contributed equally to this work. METHODS IN ENZYMOLOGY, VOL. 407 Copyright 2006, Elsevier Inc. All rights reserved.
0076-6879/06 $35.00 DOI: 10.1016/S0076-6879(05)07052-7