U0126, a recently introduced mitogen-activated protein kinase kinase (MAPK)/extracellular signal-regulated kinase kinase inhibitor reversed morphology and inhibited anchorage-independent growth of Ki-ras-transformed rat fibroblasts. Immunoblot analyses with phosphospecific antibodies indicated that in addition to MAPK, U0126 suppressed activation of p70S6K, but not Akt,at concentrations at which it normalized the transformed phenotypes. Another MAPK/extracellular signal-regulated kinase kinase inhibitor,PD98059, showed only marginal effects on p70S6Kphosphorylation and did not effectively block Ki-ras-induced transformation. However, simultaneous inhibition of the MAPK pathway and the p70S6K pathway by PD98059 in conjunction with the p70S6K inhibitor rapamycin essentially restored the normal phenotype. U0126 or the combination of PD98059 and rapamycin flattened morphology of v-src-transformed cells, but did not reverse anchorage independence, although activation of both MAPK and p70S6Kwas blocked. The results suggest that normalization of Ki-ras-induced transformed phenotypes by U0126 is a consequence of concurrent inhibition of the MAPK and p70S6Kpathways. Intervention of other pathway(s) appears to be required to completely antagonize transformation by v-src. Simultaneous blockade of more than one signal transduction pathway by combining selective inhibitors might be effective in suppressing uncontrolled tumorigenic growth.
Much of the current effort in anticancer drug development focuses on signal transduction pathways. The strategy of precisely targeting abnormalities that propel uncontrolled growth is more rational and disease oriented than the traditional cell-killing approach. A potential difficulty in development of target therapies is assessment of their efficacy. Signal transduction inhibitors are inherently cytostatic, and simple cytotoxicity assays may not be relevant. Anchorage-independent growth is the hallmark of uncontrolled tumorigenic proliferation (1, 2). In theory, specific blockade of aberrant signal transduction would normalize the transformed phenotypes and thus render tumor cells anchorage dependent. The cells are expected to lose the ability to proliferate without firm attachment but still to be competent for growth on solid support. We developed a simple, objective, and quantitative method to measure anchorage-independent growth using microtiter plates coated with the antiadhesive polymer polyHEMA3(3, 4). We anticipate the polyHEMA-coated plates will provide a system for evaluating inhibitors of oncogenic signal transduction.
The MAPK (extracellular signal-regulated kinase 1/2) pathway is considered to be one of the promising targets for signal transduction-based cancer chemotherapy. We examined the effects of U0126, a recently introduced MEK inhibitor (5), on growth properties of Ki-ras- and v-src-transformed rat fibroblasts. U0126 selectively repressed anchorage-independent growth of Ki-ras transformed cells. Subsequent analyses revealed that U0126 blocked not only the MAPK pathway but also the p70S6K pathway. Experiments with PD98059(6) and rapamycin (7) suggested that separate intervention of the MAPK pathway or the p70S6Kpathway has little effect on Ki-ras-induced transformation,but simultaneous blockade of the two pathways restores the normal phenotype. The results imply significant benefits of combining inhibitors of signal transduction pathways. Synergy arising from combinations of selective pharmacological agents might extend the range of signal transduction-based cancer chemotherapy.
Materials and Methods
U0126 was from Promega (Madison, WI); PD98059 and rapamycin were from Calbiochem (San Diego, CA); and polyHEMA was from Sigma Chemical Co.(St. Louis, MO). Antibody against phosphorylated MAPK was from Promega,and other phosphospecific antibodies were from New England Biolabs(Beverly, MA). Antibodies against MAPK, p70S6K,and Akt were from Zymed Laboratories (South San Francisco, CA), Santa Cruz Biotechnology (Santa Cruz, CA), and New England Biolabs,respectively.
Cell Culture and Measurement of Anchorage-independent Growth.
Cell lines used in the experiments have been described(3). Anchorage-independent growth was measured on polyHEMA-coated 96-well plates as described (3, 4). Briefly, 50 μl of polyHEMA solution (5 mg/ml in 95% ethanol) were pipetted into wells of 96-well plates and dried for 2 days with lids in place. Cells were inoculated in a volume of 135 μl at a density of 1000 cells per well for Ki-ras/NRK and 5000 per well for others. Inhibitors dissolved in 15 μl of medium were added, and the cells were cultured for 4 days. Fifteen microliters of 3-[4,5-dimethylthiazol-2yl]-2,5-diphenyltetrazolium bromide solution (5 mg/ml in PBS) were added, and the mixture was further incubated for 4 h. The resulting 3-[4,5-dimethylthiazol-2yl]-2,5-diphenyltetrazolium bromide formazan was solubilized by addition of 100 μl of SDS solution (20% in 10 mm HCl), and the absorbance was measured after 24 h at 570 nm and a reference wavelength of 690 nm using a microplate reader.
Cells were seeded in a volume of 1 ml in 24-well plates at a density of 1 × 105 per well and cultured for 24 h. Cells were then treated with inhibitors for 24 h,washed with cold PBS, fixed for 10 min with 10% cold trichloroacetic acid, and lysed with 80 μl of 9 m urea and 2% Triton X-100 and 20 μl of 10% lithium dodecyl sulfate. Lysates were neutralized with 2 m Tris, passed through a 250-μl syringe (Hamilton, Reno, NV) to reduce viscosity and normalized for protein using a Pierce (Rockford, IL) bicinchoninic acid kit. After addition of bromphenol blue and DTT to 0.001% and 50 mm,respectively, proteins were electrophoresed through 10% SDS-PAGE and analyzed by immunoblotting using phosphospecific antibodies. Proteins were visualized with Renaissance Western blot chemiluminescence reagent(DuPont New England Nuclear, Boston, MA).
U0126 Inhibits Anchorage-independent Growth and Normalizes Morphology of Ki-ras-transformed Cells.
Anchorage independence is the prime in vitro parameter of transformation (1, 2). We developed a microplate assay for efficient quantitation of anchorage-independent growth using plates coated with an antiadhesive polymer, polyHEMA (3, 4). This method can substitute for soft agar colony formation in many experimental situations and is more practical when many samples are to be tested.
We assessed the growth-inhibitory effects of MEK inhibitors on Ki-ras/NRK and v-src/NRK cells in polyHEMA-coated and normal tissue culture plastic plates. PD98059 inhibited growth of Ki-ras/NRK cells comparably in polyHEMA-coated and plastic plates (Fig. 1,A, middle panel). In contrast, U0126 clearly suppressed the growth of Ki-ras/NRK cells on the nonadhesive polyHEMA surface at concentrations lower than on plastic (Fig. 1,A, top panel). IC50 was 2μ m on polyHEMA, compared with 8μ m on plastic. Against the v-src/NRK cells, on the other hand, neither PD98059 nor U0126 showed preference for polyHEMA to tissue culture plastic (Fig. 1 B, top and middle panels).
Treatment of Ki-ras/NRK cells with U0126 caused apparent morphological reversion to a flat, normal appearance (Fig. 2,A; 24-h treatment). Although not as clear, U0126 also flattened morphology of v-src/NRK cells (Fig. 2,B;48-h treatment). PD98059 induced a slight morphological change in Ki-ras/NRK cells at high concentrations (Fig. 2,A), but the morphology of v-src/NRK cells did not change (Fig. 2 B).
U0126 Inhibits Both MAPK and p70S6K Pathways.
The above results indicate that the two MEK inhibitors U0126 and PD98059 differ in ability to block transformation by Ki-ras. Because the selectivity of PD98059 has been well tested(8), we speculated that U0126 inhibits other pathway(s) in addition to the MAPK pathway.
To dissect the issue, we monitored the effects of the two compounds on activation of three serine/threonine kinases that are downstream targets of Ras (9). Cell lysates from Ki-ras/NRK cells treated for 24 h were examined by immunoblotting with phosphospecific antibodies that recognize the activated form of MAPK, p70S6K or Akt. As shown in Fig. 3 A, both U0126 and PD98059 inhibited activation of MAPK, U0126 being more potent as reported (5). Neither compound displayed any effect on phosphorylation of Akt. PD98059 at 25μ m only slightly inhibited activation of p70S6K. On the other hand, reduction of p70S6K phosphorylation by U0126 was much more apparent. Although U0126 did show selectivity toward MAPK over p70S6K, it clearly blocked p70S6K activation at high concentrations. Reduction of p70S6K phosphorylation was further manifested by the increased electrophoretic mobility of total p70S6K.
As shown above, anchorage-independent growth of v-src/NRK cells was not preferentially inhibited by U0126. However, U0126 also inhibited activation of MAPK and p70S6K of the v-src transformant, although p70S6Kwas less attenuated than in Ki-ras/NRK (Fig. 3 B). Phosphotyrosine content of v-src/NRK cells did not change(data not shown).
PD98059 plus Rapamycin Blocks Transformation by Ki-ras.
The above results raised a possibility that U0126 normalizes Ki-ras-mediated transformation by blocking both MAPK and p70S6K pathways. The p70S6Kpathway is selectively blocked by the immunosuppressant rapamycin(7, 10). Activation of p70S6K is mediated by the phosphatidylinositol kinase-related kinase mTOR/FK506-binding protein rapamycin-associated protein(11). Rapamycin inhibits mTOR and hence p70S6K by forming a stable complex with FK506-binding protein 12, which binds to mTOR (12). We tested whether simultaneous blockade of MAPK and p70S6K pathways by combination of PD98059 and rapamycin would result in similar consequences.
Rapamycin alone did not selectively inhibit anchorage-independent growth (data not shown) or alter morphology (Fig. 2,A) of Ki-ras/NRK cells. As expected, rapamycin completely blocked activation of p70S6K in Ki-ras/NRK cells without any effect on MAPK or Akt, and the combination of PD98059 and rapamycin reduced phosphorylation of both MAPK and p70S6K (Fig. 3 A).
We examined the effect of PD98059 plus rapamycin on anchorage-independent growth and morphology of Ki-ras/NRK cells. At 25 nm, rapamycin by itself reduced the growth of Ki-ras/NRK cells to 57.8 ± 3.1 and 64.8 ± 4.8% of control in plastic plates and polyHEMA plates, respectively. As shown above, PD98059 did not show a marked effect on anchorage independence. However, when PD98059 was combined with 25 nm rapamycin, anchorage-independent growth on polyHEMA was noticeably inhibited over anchorage-dependent growth on plastic (Fig. 1,A, bottom panel). Ki-ras/NRK cells treated with PD98059 plus rapamycin appeared morphologically normal, similar to those treated with U0126(Fig. 2 A).
PD98059 plus rapamycin also inhibited MAPK and p70S6K in v-src/NRK (Fig. 3,B). Cells treated with the combination were flatter than cells treated with U0126, which probably reflects the more potent inhibition of p70S6K phosphorylation. Despite the clear morphological alterations, anchorage-independent and -dependent growth was still equally inhibited (Fig. 1 B, bottom panel).
U0126 Is Effective against Other ras-transformed Cells.
We next examined whether U0126 could inhibit anchorage-independent growth of other ras-transformed fibroblasts. Results of two representative cell lines, Ki-ras/3Y1 and Ha-ras/3Y1, are shown in Fig. 4. U0126 suppressed anchorage-independent growth (Fig. 4) and flattened morphology (data not shown) of these two cell lines as well as of pMAM-Ki-ras, another Ki-ras transformed fibroblast tested (data not shown). On the contrary, PD98059 was ineffective in normalizing transformed phenotypes of Ki-ras/3Y1 or pMAM-Ki-ras (Fig. 4; data not shown). The results further support the hypothesis that normalization of Ki-ras-induced transformation requires inhibition of both MAPK and p70S6K pathways. PD98059, however,inhibited anchorage-independent growth (Fig. 4) and induced morphological reversion (data not shown) of Ha-ras/3Y1 cells.
We have shown here that the MEK inhibitor U0126 normalizes morphology and inhibits anchorage-independent growth of Ki-ras-transformed rat fibroblasts. U0126 has been introduced as a potent and selective inhibitor of the MAPK pathway(5). However, our immunoblot analyses with phosphospecific antibodies suggested that U0126 also blocks the p70S6K pathway, and the reversion of transformation appeared to coincide with concurrent inhibition of MAPK and p70S6K activation. The MEK inhibitor PD98059 did not reverse Ki-ras-induced transformation by itself but essentially restored the normal phenotype in conjunction with the p70S6K pathway inhibitor rapamycin. We conclude that simultaneous inhibition of the MAPK and p70S6K pathways is necessary and sufficient to block transformation induced by Ki-ras.
Although our results demonstrate that U0126 is more effective in blocking the MAPK pathway than the p70S6Kpathway, it is obvious that data obtained from experiments that rely on U0126 as a MEK inhibitor must be cautiously interpreted. The mechanism by which U0126 inhibits p70S6K activation is not clear at present. PD98059 was less potent than U0126 in blocking the MAPK pathway but appeared to be more selective, as far as p70S6K is concerned.
Penuel and Martin (13) recently reported that simultaneous inhibition of MAPK and p70S6K pathways blocked transformation by v-src. PD98059 plus rapamycin reduced various parameters of transformation, including morphology and anchorage-independent growth in soft agar. However, their soft agar colony formation assay was not done in parallel with colony formation on a solid support. In our experiments, reversal of v-srcmediated transformation was partial; i.e., inhibition of MAPK and p70S6K pathways flattened morphology to some extent but equally suppressed anchorage-dependent and -independent growth. The possibility that this is due to incomplete blockade of the two pathways remains, but we have thus far been unable to selectively reduce anchorage-independent growth with increased concentrations of the inhibitors and have obtained similar results with another v-src transformant. Intervention of other pathway(s), such as Stat3 signaling (14, 15), might be required to antagonize transformation by v-src completely.
In contrast, U0126 or PD98059 plus rapamycin was effective against all four ras transformants we tested (Fig. 4; data not shown). In the case of a Ha-ras transformant, inhibition of the MAPK pathway alone appeared to normalize transformation. U0126 also blocked anchorage-independent growth of several human cancer cell lines. Effects of U0126 on human cancer cell lines and cells transformed by other oncogenes will be reported elsewhere.
Activating mutations of the ras genes are among the frequently found abnormalities in human cancers (16). MAPK is thought to be the key component in ras-mediated transformation (9) and is elevated in various human tumors(17, 18). Consequently, the MAPK cascade has been the focus of special attention as a target for a new generation of noncytotoxic antitumor drugs. Recently, a highly potent and selective MEK inhibitor was shown to suppress colon tumor growth in vivo without noticeable side effects, substantiating the concept of the MAPK cascade targeting as a mechanism-based cancer therapy(19).
In Ki-ras transformed cells, however, inhibition of the MAPK pathway was not sufficient to repress disordered growth but required additional inhibition of the p70S6K pathway. Given the fact that uncontrolled growth of cancer cells involves aberrations of several mechanisms, the requirement of intervention of multiple downstream pathways in many other tumors can be envisaged. Because complete blockade of all elevated growth-promoting machineries may be harmful to normal cell activity, it would be important to identify the minimum subset of signaling pathways required to control oncogenic growth. Our results suggest that absolute inhibition of Ras function may not be necessary to remedy abnormal growth properties. Blocking of all Ras function by Ras antagonists such as farnesyltransferase inhibitors (20) might be overcorrection. Knowledge of necessary and sufficient pathways for reestablishment of regulated growth would open a way to more selective,noncytotoxic signal transduction-based chemotherapy. In cases in which blockade of a single pathway by a single agent is ineffectual,targeting of multiple, but a minimal number, of pathways by combinations of selective pharmacological agents could be a competent strategy for regulation of cancers.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Supported by a grant-in-aid for Cancer Research from the Ministry of Education, Science, Sports and Culture of Japan,by the Social Institute Agency Contract Fund of the Japan Health Science Foundation, and by the Promotion of Fundamental Studies in Health Sciences of the Organization for Pharmaceutical Safety and Research of Japan.
The abbreviations used are: HEMA,2-(hydroxyethyl methacrylate); MAPK, mitogen-activated protein kinase;MEK, MAPK/extracellular signal-regulated kinase kinase; NRK, normal rat kidney; mTOR, mammalian target of rapamycin.
We thank Drs. S. Mizuno and Y. Murakami for helpful discussions and Y. Zakabi for technical assistance.