Nitric Oxide in Physiologic Concentrations Targets the Translational Machinery to Increase the Proliferation of Human Breast Cancer Cells: Involvement of Mammalian Target of Rapamycin/eIF4E Pathway

Nitric oxide (NO), a potent bioactive molecule, acts as a diffusible messenger in intercellular communication and intracellular signaling (1–3). NO interacts with different molecular targets from superoxide anion to protein macromolecules, which can be activated or inhibited through oxidation of thiols, hemes, Fe-S clusters, or other non-heme iron prosthetic groups of macromolecules (4, 5). The apparent outcome of NO signaling is dependent on the level of NO production, which is dictated by the isoforms of NO synthase (NOS)/endothelial (eNOS), neuronal (nNOS), or inducible (iNOS) NO synthases (6). Physiologic, low (nanomolar, nmol/L) concentrations of NO have been shown to increase tumorigenesis (7–9), whereas pathophysiologic high (micromolar, μmol/L) concentrations of NO, as produced by activated macrophages, induce tumoricidal effects (10). We have found that micromolar concentrations of NO, partially dependent on mitogen-activated protein kinase (MAPK) phosphatase-1 (MKP-1) induction, can either induce prolonged cytostasis or cytostasis followed by apoptosis in a number of human breast cancer cell lines. We have previously reported that micromolar concentrations of NO up-regulates MKP-1 and induces apoptosis in breast cancer cell line MDA-MB-468 by mitochondria-regulated pathway (11). In another study using the MDA-MB-231 cell line, we observed that NO (micromolar concentrations) was unable to up-regulate MKP-1, and apoptosis was not induced; however, cells underwent prolonged cytostasis by down-regulating cyclin D1 protein and hypophosphorylation of retinoblastoma protein (11, 12).

Recent studies have implicated low physiologic concentrations of NO (nmol/L) in regulating different processes of tumorigenesis. Clinical and experimental studies support a positive relationship between tumor malignancy and NOS activity in certain tumors (13–15). Solid tumors have been found to produce sustained levels of NO, which is produced by tumor cells themselves or macrophages that infiltrate these tumors (16). NO production is stimulated by vascular endothelial growth factor (VEGF), a strong angiogenic factor that is up-regulated in tumor tissues (17). Cytokines and chemokines in tumor microenvironment may provide appropriate signals for recruitment and activation of macrophages that are capable of NO production (18). Although NO is an important component of tumor microenvironment, its role in tumor biology remains to be elucidated. Detailed investigation is required to understand cellular targets of NO, which could facilitate elucidation of signaling pathways involved in tumor growth, invasion, and metastasis. Extracellular signal-regulated kinases (ERK1/2), which play a vital role in VEGF induced endothelial cell proliferation, is the most well studied downstream target of NO (19). The purpose of this study was to determine the key signaling cascades targeted by physiologic (nanomolar) concentrations of NO in malignant breast cancer cell lines, including MDA-MB-231 and MCF-7, which could give us an insight into the possible mechanisms by which NO increases proliferation of tumor cells.

In this study, we have used human breast cancer cell lines MDA-MB-231, MCF-7, and MDA-MB-468 to study the mechanism by which DETA-NONOate, an NO donor, increases cell proliferation at physiologic concentrations. These human breast cancer cell lines with undetectable to low endogenous NOS levels (20) were exposed to various concentrations of DETA-NONOate, a long acting NO donor with a half-life of 20 h. We found that nanomolar concentrations of NO released by 30 to 60 μmol/L DETA-NONOate significantly increased proliferation in all these breast cancer cells. Due to comparable % increase in proliferation with NO treatment, we used MDA-MB-231 and MCF-7 to study the mechanism for increased proliferation. Nanomolar NO treatment increased the rate of synthesis of some proliferation-related proteins like cyclin D1 and ornithine decarboxylase (ODC) without altering the mRNA levels or half-lives. Nanomolar concentrations of NO, as released by 30 to 60 μmol/L DETA-NONOate, increased phosphatidylinositol 3-kinase (PI-3 kinase)/Akt signaling to activate mammalian target of rapamycin (mTOR) and its downstream effectors p70 S6 kinase (p70^s6k) and eukaryotic translation initiation factor (eIF-4E), which are critical components in the translational machinery. Pharmacologic inhibition of PI-3 kinase/Akt or Raf/MAP/ERK kinase (MEK)/ERK1/2 signaling reduced the levels of phosphorylated eIF4E (peIF4E) and cyclin D1. NO-induced up-regulation of PI-3 kinase/Akt and Raf/MEK/ERK signaling was found to be mediated by Ras as inactivation of Ras by farnesyl transferase inhibitor (FTase inhibitor 111) or K-Ras si RNA attenuated the effect of NO on these signaling cascades. The effects of NO on cell proliferation were not attenuated in the presence of 1H-[1,2,3]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ), a guanylate cyclase inhibitor. Our studies therefore suggest that NO induced proliferation in these breast cancer cells was due to cyclic guanosine 3′,5′-monophosphate (cGMP)–independent up-regulation of PI-3 kinase/Akt and Raf/MEK/ERK1/2 signaling cascades.

Materials. DETA-NONOate, ODQ, and 2-(4-carboxyphenyl)-4,4,5,5-tetramethyl-1-oxyl-3-oxide (cPTIO) were purchased from Cayman Biochemicals (Ann Arbor, MI). Actinomycin D, cycloheximide, rapamycin, 8-bromo-cGMP, leupeptin, aprotinin, phenylmethylsulfonyl fluoride (PMSF), and protein A-Sepharose beads were purchased from Sigma (St. Louis, MO); PD 98059, LY294002, and farnesyl transferase inhibitor (FTI) were from Calbiochem (San Diego, CA). All of the cell culture media were purchased from Life Technologies, Inc. (Gaithersburg, MD). Rabbit polyclonal anti–phosphorylated Akt (anti-pAkt; 559029) was from PharMingen, BD Biosciences (San Diego, CA). ERK1/2 MAPK (9102) was from New England Biolabs (Beverly, MA). PI-3 kinase, ODC, cyclin D1, Akt, ERK1/2 were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Phosphorylated mTOR, peIF4E, phosphorylated p70^s6k (pp70^s6k), pAkt, phosphorylated ERK1/2 (pERK1/2), phosphorylated Raf (pRaf), and MBP kit were from Cell Signaling Technology (Beverly, MA). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was purchased from Chemicon (Temecula, CA). Tran³⁵S-label and [³⁵S]cysteine were from MP Biomedicals (Solon, OH). [α-³²P]dCTP and [γ-³²P]ATP were from Amersham Pharmacia (Piscataway, NJ).

Cell culture. Human breast cancer cell lines MDA-MB-231, MCF-7, and MDA-MB-468 were obtained from the American Type Culture Collection (Manassas, VA). MDA-MB-231, MCF-7 cells, and MDA-MB-468 were cultured in DMEM containing 10 mmol/L nonessential amino acids, 2 mmol/L l-glutamine, 1 μg/mL insulin, and 10% fetal bovine serum (FBS). For experimental purpose, cells were grown in 5% FBS, allowed to seed overnight, and treated with drugs for various time periods.

Measurement of rate of NO release by DETA-NONOate. The rate of NO released from 10 to 0 μmol/L DETA-NONOate in DMEM was assessed over time using Sievers chemiluminescent NO analyzer (21).

Measurement of total protein synthesis. Cells (2 × 10⁶) were plated overnight in media containing DMEM and 5% FBS. The cells were kept starved for 3 h in methionine/cysteine–free media (Life Technologies), after which they were labeled for further 3 h with 100 μCi/mL of a mixture of [³⁵S]methionine and [³⁵S]cysteine. After removal of the radioactive media, the cells were collected, washed twice with PBS, and lysed on ice for 30 min in 1 mL of lysis buffer [10 mmol/L Tris/HCl buffer (pH 7.4), 150 mmol/L NaCl, 1 mmol/L EDTA (pH 8), 0.1% (v/v) Triton X-100] containing 0.2 mmol/L PMSF, 1 μmol/L pepstatin A, and 1 μmol/L leupeptin. Cell lysates were clarified by centrifugation at 2,200 × g for 5 min, and the protein concentrations of the supernatant were determined using Protein Assay Dye Reagent Concentrate (Bio-Rad Laboratories, Inc., Hercules, CA). Equal amounts of lysates were spotted on Whatman No. 3MM and washed with TCA, and incorporation of labeled [³⁵S]methionine/[³⁵S]cysteine in protein was measured in a scintillation counter.

Cell viability. Cells seeded in six-well plate (7.5 × 10⁵ per well) were allowed to grow overnight. The cells treated with various concentrations of DETA-NONOate for 24 h were collected, and viability was determined by trypan blue exclusion method. The number of viable cells at each concentration and time point was determined in triplicate with a hemacytometer.

Western analysis. Cells were lysed in cell lysis buffer containing 50 mmol/L HEPES (pH 7.5), 1 mmol/L DTT, 150 mmol/L NaCl, 1 mmol/L EDTA, 0.1% Tween 20, 10% glycerol, 10 mmol/L β-glycerophosphate, 1 mmol/L NaF, 0.1 mmol/L orthovanadate, 10 μg/mL leupeptin, 10 μg/mL aprotinin, and 0.1 mmol/L PMSF. Western blot was done as previously described (11).

Determination of ODC activity. MDA-MB-231 cells (5 × 10⁶ per plate) were treated with DETA-NONOate (0–60 μmol/L) for 24 h, washed twice with homogenization buffer [0.1 mmol/L pyridoxal-5-phosphate, 0.1 mmol/L EDTA, 2.5 mmol/L DTT, and 50 mmol/L sodium phosphate pH (7.2)], followed by homogenization in 1 mL of homogenization buffer and centrifugation at 13,000 × g for 15 min at 4°C. The supernatant, free from mitochondria, was used for ODC assay (22).

Immunoprecipitation of labeled cyclin D1 and ODC. [³⁵S]methionine- and [³⁵S]cysteine-labeled cyclin D1 was immunoprecipitated from the total cell lysates, which was prepared for measuring total protein synthesis. Radiolabeled total cell lysate (500 μg) was incubated overnight with 1 μg of monoclonal anti-cyclin D1 and anti-ODC antibodies in modified radioimmunoprecipitation assay buffer (50 mmol/L Tris-HCl, 150 mmol/L NaCl, 1% NP40, and 0.5% sodium deoxycholate). The immune complex was precipitated with 30 μL of Protein A-agarose (Santa Cruz Biotechnology), washed four times with lysis buffer, and analyzed by 10% SDS-PAGE.

Northern analysis. Cells were treated with DETA-NONOate for various time points, and total cellular RNA was extracted using RNeasy (Qiagen, Chatsworth, CA) according to the manufacturer's instruction. Equal amounts (15 μg per lane) of RNA were resolved by electrophoresis on a 1.2% formaldehyde-agarose gel, blotted onto nylon membrane, and probed using standard protocol (12).

Immunoprecipitation and kinase assay. Cells were lysed in lysis buffer containing 50 mmol/L Tris/HCl (pH 8), 300 mmol/L NaCl, 10 mmol/L MgCl₂, 0.5% Igepal Ca-630 (Sigma), 1 mmol/L EDTA, and antiprotease cocktail (Roche, Indianapolis, IN); 500 μg of supernatant were precleared by incubation (4 h at 4°C) once with Protein A/G-agarose (Santa Cruz Biotechnology) and once with Protein A/G-agarose preincubated with preimmune serum. The precleared lysate was adjusted to 150 mmol/L NaCl and 0.25% Igepal Ca-630 and subjected to overnight immunoprecipitation on a rotating wheel at 4°C using rabbit antiserum or anti-Akt antibody. The mixtures were incubated further for 4 h at 4°C with 50 μL of Protein A/G-agarose. After five washes in 50 mmol/L Tris/HCl (pH 8), 0.25% Igepal Ca-630, 1 mmol/L EDTA, and 150 mmol/L NaCl, the immunoprecipitates were analyzed by kinase assay using myelin basic protein as the substrate.

Inhibition of K-Ras by small interfering RNA. Synthetic small interfering RNA (siRNA) targeting activated K-Ras (GGAGCUGUUGGCGUAGGCAA) present in MDA-MB-231 cells, and random oligo VIII as a control for nonspecific siRNA effects were purchased from Dharmacon Research (Lafayette, CO) as described (23). Amaxa nucleofecter was used for the transfection (∼85% efficiency) for 24 h, after which the cells were treated with and without DETA-NONOate for another 24 h, and cell proliferation was measured.

cGMP assay. cGMP was measured using Enzyme Immunoassay kit from Assay Designs, Inc. (Ann Arbor, MI).

Statistics. Data are presented as mean ± SE. Differences between the groups were analyzed by ANOVA. If overall ANOVA revealed significant differences, then pairwise comparisons between groups were done using Student's t test. Ps < 0.05 were considered statistically significant. The experiments were repeated three to five times, and data from representative experiments are shown.

Nanomolar NO increased proliferation of breast cancer cells. In this study, we used lower, physiologic concentrations of NO (nanomolar; DETA-NONOate at 10–60 μmol/L) to elucidate the cellular targets and the possible mechanisms by which it promotes tumor growth. We have determined that 10, 30, and 60 μmol/L DETA-NONOate releases 2, 33, and 52 nmol/L NO in the culture media as described in Materials and Methods. We initially assessed the effect of different concentrations of NO on cell proliferation in human breast cancer cell lines MDA-MB-231, MCF7, and MDA-MB-468 (data not shown). A significant increase in viable cells after exposure to 30 μmol/L (26 ± 2% for MDA-MB-231, P < 0.05 and 27 ± 2% for MCF7, P < 0.05) and 60 μmol/L (53 ± 3% for MDA-MB-231, P < 0.01 and 63 ± 4% for MCF7, P < 0.01) of DETA-NONOate was observed compared with control cells (Fig. 1A). Because the NO-induced % increase in cell proliferation of MDA-MB-231, MCF-7, and MDA-MB-468 was very comparable, we have used MDA-MB-231 and MCF-7 for further experiments. To assess whether the increased proliferation was due to NO released from DETA-NONOate, MDA-MB-231 and MCF-7 cells were coincubated with DETA-NONOate and cPTIO (50 μmol/L), a well-known NO quencher (24). Coincubation with cPTIO attenuated NO-induced increase in proliferation of these cells (Fig. 1A), suggesting that the increase with DETA-NONOate (30–60 μmol/L) was due to the NO released from the chemical compound. To examine whether nanomolar NO affected the rate of total protein synthesis in MDA-MB-231 and MCF-7 cell lines, DETA-NONOate–treated cells were pulse labeled with [³⁵S]methionine/[³⁵S]cysteine for 6 h. Equal amounts of lysates were spotted on Whatman No. 3MM and washed with TCA, and incorporation of labeled [³⁵S]methionine/[³⁵S]cysteine in protein was measured in a scintillation counter. We observed a significant increase in [³⁵S]methionine/[³⁵S]cysteine–labeled proteins in cells exposed to 30 μmol/L (57 ± 3% for MDA-MB-231, P < 0.05 and 46 ± 16% for MCF7, P < 0.05) and 60 μmol/L (118 ± 6 for MDA-MB-231, P < 0.01 and 146 ± 8 for MCF7, P < 0.01) of DETA-NONOate, compared with the control cells (Fig. 1B).

To investigate the mechanism by which nanomolar NO increased proliferation of MDA-MB-231 and MCF-7 cells, we examined some of the cell cycle proteins involved in G₁-S progression. ODC, which is strongly implicated in G₁ progression, was also assessed in nanomolar NO–treated MDA-MB-231 and MCF-7 cells. We observed that nanomolar NO treatment increased the levels of cyclin D1 and ODC in MDA-MB-231 (Fig. 1C) and MCF-7 cells (Fig. 1D). Densitometric scans showed a significant increase in cyclin D1 protein in MDA-MB-231 (1.6 ± 0.3-fold at 30 μmol/L, P < 0.05 and 2.2 ± 0.3-fold at 60 μmol/L, P < 0.05) and in MCF7 (1.5 ± 0.4-fold at 30 μmol/L, P < 0.05 and 2.6 ± 0.3-fold at 60 μmol/L, P < 0.05) after exposure to DETA-NONOate (Fig. 1C and D). We also found a significant increase in ODC protein in MDA-MB-231 (1.7 ± 0.3-fold at 30 μmol/L, P < 0.05 and 1.9 ± 0.2-fold at 60 μmol/L, P < 0.05) and in MCF7 (1.6 ± 0.3-fold at 30 μmol/L, P < 0.05 and 2.1 ± 0.3-fold at 60 μmol/L, P < 0.05) after DETA-NONOate exposure (Fig. 1C and D). We further assessed the enzymatic activity of ODC in MDA-MB-231 cell lysates after treatment with DETA-NONOate (0–60 μmol/L) for 24 h. We observed a significant increase in enzymatic activity of ODC following treatment of these cells with DETA-NONOate (0.98 ± 0.06 nmol/mg protein at 10 μmol/L, P < 0.05; 1.1 ± 0.05 nmol/mg protein at 30 μmol/L, P < 0.05; and 0.99 ± 0.08 nmol/mg protein at 60 μmol/L, P < 0.05) compared with control (0.76 ± 0.04 nmol/mg protein; Fig. 1E).

Nanomolar NO increased the rate of synthesis of cyclin D1 and ODC. To determine whether increased levels of cyclin D1 and ODC proteins were due to up-regulation of their transcription, Northern blot analysis was done. Nanomolar NO treatment of MDA-MB-231 cells did not alter the mRNA levels of cyclin D1 or ODC when compared with control cells (Fig. 2A). In addition, blocking cellular transcription with actinomycin D (1 μg/mL) did not attenuate NO-induced increase in cyclin D1 (Fig. 2B) and ODC (data not shown) proteins, suggesting the involvement of posttranscriptional mechanisms in the process.

Changes in steady-state protein levels can be accomplished by alterations in either the rate of synthesis or by change in protein half-life. To determine the effect of NO on the rate of synthesis of cyclin D1 and ODC, the MDA-MB-231 cells after various treatments were starved in methionine/cysteine–free growth media and further pulse labeled for 3 h with [³⁵S]methionine/[³⁵S]cysteine. Total cyclin D1 or ODC was immunoprecipitated from the labeled cell lysates using anti-cyclin D1 or anti-ODC antibodies and analyzed by SDS-PAGE and autoradiography. Cells exposed to nanomolar NO had increased labeled cyclin D1 (3.1 ± 0.3-fold, P < 0.001) and ODC (3.6 ± 0.4-fold, P < 0.001) proteins at 60 μmol/L DETA-NONOate compared with control cells, suggesting an increased rate of translation of their mRNAs (Fig. 2C). To ascertain the potential mechanisms by which nanomolar NO increased translation of cell cycle proteins, we decided to focus on cyclin D1 as the downstream target of NO for further experiments. To assess whether NO affected protein stability of cyclin D1, we analyzed cyclin D1 turnover before and after NO treatment by Western analysis. Protein degradation was monitored by blocking the de novo protein synthesis in the cells with cycloheximide (100 μg/mL). The half-life of cyclin D1 (∼30 min with or without NO) was unaltered under these conditions, whereas overall cyclin D1 protein levels were significantly increased with NO treatment (Fig. 2D). These results indicate that nanomolar NO did not affect the stability of cyclin D but increased its rate of synthesis.

Nanomolar NO up-regulates PI-3 kinase/pAkt activation in MDA-MB-231 cells. Increased rate of synthesis of cyclin D1 and ODC could reflect a direct effect of nanomolar NO on the translational machinery or result from activation of signaling pathways, such as those initiated by tyrosine kinase activation. The PI-3 kinase/Akt pathway plays an important role in human cancer and has been found to modulate the activity of key translation factors in the translational machinery (25). To analyze whether nanomolar NO–induced increase in rate of synthesis of cyclin D1 and ODC was through the PI-3 kinase/Akt pathway, we examined the levels of total and pAkt by immunoblot analysis using total and phospho-specific antibodies that detected phospho- groups on both Ser⁴⁷³ and Thr³⁰⁸ in Akt. Treatment of MDA-MB-231 cells with 60 μmol/L DETA-NONOate increased the levels of pAkt by 1.8 ± 0.2-fold (P < 0.01) compared with untreated control cells, whereas total Akt level remained unchanged (Fig. 3A). To further confirm whether nanomolar NO–treated cells had increased pAkt activity, a kinase assay was done, where total Akt was immunoprecipitated from cell lysates, and the pellets were subjected to kinase assay using myelin basic protein as substrate. We observed a 3.1 ± 0.3-fold (P < 0.001) increased phosphorylation of myelin basic protein (Fig. 3B) in cells treated with 60 μmol/L DETA-NONOate compared with control, suggesting an increase in pAkt activity after treatment with nanomolar NO.

We used pharmacologic inhibitors of PI-3 kinase to examine whether increase in pAkt activity regulated cyclin D1 translation in NO-treated cells. Treatment of MDA-MB-231 cells with LY294002 (30 μmol/L), a PI-3 kinase inhibitor, decreased the basal levels of pAkt (50 ± 6%, P < 0.05) and cyclin D1 (60 ± 7%, P < 0.02) and attenuated DETA-NONOate (60 μmol/L)–induced increase of pAkt (60 ± 8-fold, P < 0.001) and cyclin D1 (54 ± 11-fold, P < 0.02; Fig. 3C). Similar results were obtained when wortmannin, another PI-3 kinase inhibitor, was used to inhibit pAkt (data not shown). These results indicated that nanomolar NO–induced increased cyclin D1 synthesis was possibly via up-regulation of pAkt in MDA-MB-231 cells. Activated PI-3 kinase enhances the synthesis of phosphatidylinositol 3′-phosphates, which binds to pleckstrin homology domain of Akt and increases its activation by targeting it to the cell membrane (26). To assess whether nanomolar NO–induced activation of Akt was due to increased activity of upstream PI-3 kinase, lysates from control and nanomolar NO–treated cells were immunoprecipitated with antibody against regulatory p85 subunit of PI-3 kinase and Western blotted with antibody against its catalytic p110 subunit. We found that nanomolar NO treatment increased the association of p110 subunit and p85 subunits of PI-3 kinase (3.2 ± 0.3-fold at 60 μmol/L DETA-NONOate, P < 0.001) compared with control cells (Fig. 3D), suggesting a possible mechanism for increase in its activity. In addition, for loading control, we immunoprecipitated p85 from the same treated lysates, which was blotted with anti-p85 antibody (Fig. 3D).

Nanomolar NO increased the activity of mTOR and its downstream targets eIF4E and p70^s6k. The mTOR is a downstream effector of PI-3 kinase/Akt signaling pathway, which mediates crucial aspects of Akt-induced oncogenesis (27). To determine the effect of NO on mTOR activity, Western blot analysis was done using phospho-specific antibody to Ser²⁴⁴⁸, which has been shown to be important in the control of mTOR activity. We observed that nanomolar NO treatment increased the levels of phosphorylated mTOR (pmTOR) by 1.8 ± 0.3-fold (P < 0.05) in MDA-MB-231 and 2.1 ± 0.4-fold (P < 0.05) in MCF-7 compared with control cells (Fig. 4A and B). Two translational components, ribosomal p70^s6k and eIF4E binding protein 1 (4E-BP-1), are the best-characterized downstream effectors of mTOR. 4E-BP-1 binds to and represses the function of eIF4E, and phosphorylation of 4E-BP-1 dissociates it from eIF4E complex, enhancing a cap-dependent translation (28). Effect of nanomolar NO treatment on the phosphorylation state of p70^s6k was assessed using phospho-specific antibody to Thr³⁸⁹ and Thr²²⁹, sites whose phosphorylation are vital to p70^s6k activation. Effect of NO on the phosphorylation state of eIF4E was assessed by Western blot analysis using a phospho-specific antibody to Ser²⁰⁹ whose phosphorylation was important to eIF4E activation. There were increased phosphorylation levels of eIF4E by 1.7 ± 0.3-fold (P < 0.05) in MDA-MB-231 (Fig. 4A) and 1.8 ± 0.2-fold (P < 0.05) in MCF-7 cells after 60 μmol/L DETA-NONOate treatment. This was also accompanied by a significant increase in pp70^s6k (1.7 ± 0.2-fold, P < 0.05 in MDA-MB-231 and 1.6 ± 0.2-fold, P < 0.05 in MCF-7) at a similar concentration of the NO donor (Fig. 4B). To elucidate whether NO-induced activation of mTOR was involved in the increase of cyclin D1 in MDA-MB-231 cells, we further inhibited mTOR with rapamycin, a specific inhibitor of mTOR kinase. A reduction (53 ± 8%) in cyclin D1 levels in rapamycin-treated cells was observed compared with control, suggesting the involvement of the pAkt/mTOR pathway in nanomolar NO–induced increase in cyclin D1 synthesis (Fig. 4C).

FTase inhibitor 111 inhibits nanomolar NO–induced activation of PI-3 kinase/Akt and pERK1/2. Farnesylation of Ras is required to anchor it to the cell membrane and trigger signaling through the PI-3 kinase/Akt and Raf/MEK/ERK pathways (29). To explore the possible role of Ras in nanomolar NO–mediated activation of membrane-bound PI-3 kinase/Akt, Ras was rendered inactive in MDA-MB-231 cells by FTI (FTase inhibitor 111). FTI (25 μmol/L) treatment alone was sufficient to reduce the levels of pAkt in these cells. FTI treatment also attenuated NO-induced increase in pAkt, suggesting requirement of activated Ras in NO-mediated increase in PI-3 kinase/Akt signaling (Fig. 5A). A decline in the level of cyclin D1 in FTI and NO plus FTI–treated MDA-MB-231 cells (Fig. 5A) further emphasizes the role of Ras/PI-3 kinase/Akt signaling in cyclin D1 synthesis. Because it is reported that Ras also increases the signaling through the Raf/MEK/ERK pathway, we examined the levels of pERK1/2 in FTI and NO plus FTI–treated cells. To examine the effect of nanomolar NO on pERK1/2 in MDA-MB-231 cells, Western blot using a phospho-specific antibody against Thr²⁰²/Tyr²⁰⁴ of pERK1/2 was done. Whereas DETA-NONOate (60 μmol/L) increased the levels of pAkt (2.15 ± 0.07-fold, P < 0.001), pERK1/2 (1.6 ± 0.2-fold, P < 0.02), and cyclin D1 (1.82 ± 0.12-fold, P < 0.05) compared with control cells. In DETA-NONOate–treated cells after simultaneous treatment with FTI (25 μmol/L), the respective increases in pAkt (P < 0.001) and cyclin D1 (P < 0.001) were abolished (Fig. 5A). We also observed that DETA-NONOate (60 μmol/L)–induced increase in cell proliferation was abolished when the cells were simultaneously treated with FTI (Fig. 5B). It is reported that K-Ras is highly activated in MDA-MB-231 cells (30). To further assess the involvement of K-Ras in NO-induced proliferation, we treated the cells with K-Ras siRNA, which reduced the level of K-Ras by 65% to 70%. We observed that K-Ras siRNA significantly attenuated NO-induced increase in cell proliferation (P < 0.05; Fig. 5C) in these cells. The above data suggest that nanomolar NO–mediated increase of PI-3 kinase/Akt and pERK1/2 was most likely at least in part mediated through activation of Ras.

To examine the mechanism by which nanomolar NO increased the levels of pERK1/2 in MDA-MB-231 cells, the levels of pMEK1/2 and pRaf, which are kinases upstream to ERK1/2, were examined. Lysates from nanomolar NO–treated and control cells were Western blotted with phospho-specific antibodies to Ser²⁵⁹ for pRaf and Ser^217/221 for pMEK1/2 and pERK1/2. We observed that DETA-NONOate (60 μmol/L)–treated cells had increased levels of pRaf (1.6 ± 0.1, P < 0.02), pMEK (1.9 ± 0.1, P < 0.001), and pERK1/2 (2.1 ± 0.2, P < 0.001) compared with control group, whereas the levels of ERK1/2 remained unchanged (Fig. 5D), suggesting increased Raf/MEK/pERK1/2 signaling in MDA-MB-231 cells.

To examine whether Raf/MEK/ERK1/2 signaling was involved in nanomolar NO–induced increase in cyclin D1 levels, PD 98059, an inhibitor of MEK1/2, was used to inhibit ERK1/2 activity in MDA-MB-231 cells. After treatment with PD 98059 (30 μmol/L) inhibitor for 48 h, cells were further exposed to nanomolar NO for 24 h. We observed that DETA-NONOate (60 μmol/L) increased cyclin D1 (2.1 ± 0.7-fold, P < 0.02) expression compared with control cells, whereas simultaneous treatment of cells with PD 98059 (30 μmol/L) led to a significant inhibition of DETA-NONOate–induced increase in cyclin D1, suggesting involvement of Raf/MEK/ERK1/2 signaling in cyclin D1 synthesis (Fig. 5E). Because pERK1/2 is known to increase the activity of cap-binding translation factor eIF4E, we assessed the levels of peIF4E and in the presence of PD 98059. DETA-NONOate (30 μmol/L) induced increase in peIF4E (1.8 ± 0.12, P < 0.02) compared with control (1.08 ± 0.2) and was abolished in cells simultaneously treated with PD 98059 (Fig. 5F).

Because nanomolar NO has been found to increase Raf/MEK/ERK1/2 and PI-3 kinase/Akt signaling in MDA-MB-231 cells, we assessed the possibility of a cross-talk between these pathways after nanomolar NO treatment. In presence of PD 98059 inhibitors when the levels of pERK1/2 was lowered at 48 h, nanomolar NO induced increase in the levels of pAkt was not attenuated, suggesting the increased signaling through these pathways occurred independent of each other (data not shown).

Nanomolar NO–induced increase in PI-3 kinase/Akt and Raf/MEK/ERK signaling is cGMP independent in MDA-MB-231 cells. Because cGMP influences a number of crucial NO-regulated signaling events, we assessed whether low concentrations of nanomolar NO–mediated increase in cell proliferation of MDA-MB-231 was cGMP dependent. To determine the role of cGMP in nanomolar NO–induced increase in Raf/MEK/ERK and PI-3 kinase/Akt signaling, cells after various treatments were analyzed for cGMP levels as described in Materials and Methods. Pharmacologic inhibitor of guanylate cyclase ODQ (30 μmol/L; ref. 31) was used to block the production of cGMP in NO-treated cells. The level of cGMP was found to peak at 3 h of NO treatment in MDA-MB-231 cells, after which the levels of cGMP declined (Fig. 6A). In our study, DETA-NONOate (60 μmol/L) treatment significantly increased cGMP (14.4 ± 1.3, P < 0.003 after 3 h and 10.1 ± 1.1, P < 0.003 after 6 h) compared with control, whereas simultaneous treatment of DETA-NONOate treated cells with ODQ (30 μmol/L) led to a significant decrease (0.8 ± 0.2, P < 0.001 at 3 h and 0.7 ± 0.18, P < 0.001 at 6 h) in cGMP levels (Fig. 6A). Higher concentrations of NO (μmol/L), which is associated with cytostasis and apoptosis, led to an increase in cGMP levels to 40 to 50 pmol/mL at 3 h (data not shown). Even in the presence of ODQ, when NO-induced increase in cGMP was abrogated, nanomolar NO increased the proliferation of MDA-MB-231 cells (Fig. 6B). Pretreatment of DETA-NONOate (60 μmol/L)–treated cells with ODQ did not attenuate nanomolar NO–induced increase in the levels of pERK1/2 (1.6 ± 0.2-fold versus 1.9 ± 0.3-fold), cyclin D1 (1.8 ± 0.3-fold versus 1.7 ± 0.2-fold), or pAkt (2.1 ± 0.3-fold versus 2.0 ± 0.3-fold) compared with control group, suggesting that cGMP-independent mechanisms were operative in MDA-MB-231 cells (Fig. 6C). These experiments suggest that nanomolar NO–mediated Ras activation in MDA-MB-231 cells seems to be cGMP independent.

Tumor-associated NO, depending on its local concentration and duration of exposure, can affect tumor growth positively or negatively by regulating specific signal transduction pathways (32, 33). Although controversies still persist regarding concentrations of NO, there are substantial experimental evidences that support a positive association between nanomolar concentrations of NO and tumor progression (7–9). In this study, we observed that low concentrations (nanomolar) of NO target the translational machinery to increase the rate of synthesis of cyclin D1 and ODC (Fig. 2C), without altering their mRNA levels or half-lives. Translational control has been predicted for many proliferation-related proteins after sequence comparison of their 5′ untranslated regions (UTR) with other mRNA species (34). Although malignant cells harbor potentially perturbed translational machinery, there are also evidences that manipulation of this machinery leads to cancer development and progression (33, 34). Studies have identified cyclin D1 and ODC as two integral cell cycle transit proteins whose cap-dependent transcripts are subject to strong translational control (33, 34). Recent studies show that posttranscriptional events play a critical role in maintaining high protein levels of cyclin D1, a potential oncogenic element in tumors (35). Considerable increase in ODC activity due to up-regulation in the translation of the ODC mRNA into protein has been found in Ras-overexpressed cells (36). ODC, a key regulatory enzyme in the biosynthesis of polyamines, is associated with carcinogenesis, and breast tumor tissues have been found to have 2- to 3-fold higher polyamine levels than surrounding normal tissues (36). Both cyclin D1 and ODC mRNA contain highly structured 5′ UTRs, rendering them to be less efficiently translated in normal cells (37). In our study, NO treatment of MDA-MB-231 and MCF-7 cells increased the activity of key components in the translational machinery. eIF4E is an important regulatory component of the translation initiation complex that controls the activation of proto-oncogenes and is present in limiting concentration in the cells. We observed increased levels of peIF4E in nanomolar NO–treated cells, which could explain the increased translation of cyclin D1 and ODC, both of which have lengthy highly structured 5′ UTR. To our knowledge, this is the first report where NO, at least, partially up-regulates the rate of synthesis of cyclin D1 and ODC through the mTOR/eIF4E pathway.

Biological activities of NO can be mediated by either cGMP-dependent (38) or cGMP-independent pathways (39–41). The most important physiologic effects of NO are mediated by binding to guanylyl cyclase–coupled specialized heme group, which results in activation of guanylyl cyclase and production of cGMP (38). In our study, cells pretreated with ODQ, an inhibitor of guanylate cyclase, did not attenuate nanomolar NO–induced increase in pERK, pAkt, and cyclin D1, which was assessed at 24 h. In addition, nanomolar NO–induced increase in cell proliferation was also not attenuated in presence of ODQ, suggesting that NO initiated cGMP-independent mechanisms, which were operative in this malignant breast cancer cell line. In our study, we observed that NO increased cell proliferation by up-regulating PI-3 kinase/Akt activity by a cGMP-independent mechanism. Our results are contrary to the study where NO promotes endothelial cell migration and neovascularization by increasing PI-3 kinase/Akt activity in a cGMP-dependent mechanism (42). Our results are also inconsistent with the study where NO in low concentrations increases survival of neurons and endothelial cells by increasing the activity of PI-3 kinase/Akt in a cGMP-mediated manner (43). Inconsistency with our results could be due to utilization of other NO donors (GSNO and SNAP) with shorter half-life in other studies, which is in contrast to our study, where DETA-NONOate, with the longest half-life of the entire known NO donors, was used. The inconsistency could also be due to other studies using normal cell types (endothelial and neuronal cells), which have very different biological settings than the malignant human breast cancer cells used in our study. In our studies, we were also analyzing longer-lasting perturbations in the signaling pathways that lead to proliferation, a sharp contrast to the transient changes detected in most studies. However, to our knowledge, this is the first report where NO increases PI-3 kinase/Akt signaling in a cGMP-independent manner by activation of membrane-bound Ras. This increase in the levels of pAkt is only with lower concentrations of NO, whereas higher concentrations (micromolar) decrease the levels of pAkt-inducing apoptosis (11).

In this study, we observed that NO up-regulates ERK1/2 signaling at least in part by activating Ras in a cGMP-independent manner. Our results are inconsistent with report where NO mediates cGMP-dependent activation of ERK1/2 to promote migration of mouse mammary tumor cells (19). There is probability that in these NOS-overexpressing mouse mammary tumors, Ras or its associated adaptor proteins involved in amplification of Ras signaling were not adequately expressed, thereby preventing the effects of NO on Ras modifications to be manifested. Our results are also inconsistent with those of another study conducted in human breast cancer cell line MCF-7, where NO-mediated cGMP-dependent ERK1/2 activation was very transient (1 h) in nature (33). The inconsistency with our results where we get persistent ERK1/2 activation is probably because other studies used short half-life NO donors (DEA/NO and Spermine/NO), whereas we used DETA-NONOate, which has a much longer half-life.

Recent studies show that NO regulates a wide range of biological functions via posttranslational modifications, leading to nitrosylation of critical cysteine residues of proteins in signaling cascade (40, 41). The reversible regulation of proteins by nitrosothiols has been established as critical signaling mechanism by nNOS (40). Nanomolar NO has been reported to activate Ras posttranslationally, due to S-nitrosylation of critical Cys¹¹⁸ residue, which stimulates guanine nucleotide exchange (44). Ras, a membrane-bound proto-oncogene, plays a key role as a molecular switch that links the activation of cell surface receptors with signaling cascades that control proliferation. Data from our study suggest that nanomolar NO treatment of MDA-MB-231 cells led to Ras-mediated increase in PI-3 kinase/Akt and Raf/MEK/ERK signal transduction pathways. Posttranslational modification of Ras protein, by farnesyl transferase, is critical for its attachment to the cell membrane for proper functioning. FTase inhibitor 111 alone decreased the basal levels of pERK1/2, pAkt, and cyclin D1 and induced cytostasis (Fig. 5A and B), which suggests the dependence of these cells on functional Ras for proliferation. Pretreatment with FTase inhibitor 111 also attenuated nanomolar NO–induced increase in pERK1/2, pAkt, or cyclin D1 and cell proliferation (Fig. 5A and B). Furthermore, K-Ras siRNA attenuated NO-induced cell proliferation (Fig. 5C), indicating a critical role of Ras in nanomolar NO–mediated cell proliferation signaling. MDA-MB-231 is one of the most malignant breast cancer cell lines, which contains numerous genetic mutations leading to overexpression of potential oncogenes. Growth factor receptor binding protein 2, which complexes with adaptor proteins involved in Ras signaling and HER-2, which increases the sensitivity of Ras signaling, are overexpressed in breast cancer cells (44–47). In this biological setting, low concentration of NO could amplify Ras signaling by inducing conformational changes of the membrane-bound Ras protein. Because NO has been shown to promote guanine nucleotide exchange on Ras to promote Ras-GTP levels, there is also a possibility that NO might be modulating Rheb/mTOR activity. The major difference between mode of action of Ras and that of Rheb is that in case of Ras, GTP charging is necessary for ability of Ras to bind and activate its effectors, whereas Rheb GTP charging is necessary only for effector function (48).

In summary, our data show that low concentrations of NO increases the rate of synthesis of cyclin D1 and ODC in MDA-MB-231 and MCF-7 cell lines by activating mTOR, eIF4E, and p70^s6k in the translational machinery via PI-3 kinase/Akt and RAF/MEK/ERK1/2 signaling cascades. The up-regulation of Akt and ERK1/2 activity was due to NO-induced cGMP-independent Ras activation. Results from our studies may have important clinical implications. Most authors have found increased expression of iNOS (18, 49) and eNOS expression (50) compared with weak to undetectable expression seen in benign breast epithelium. Although iNOS is expressed in breast cancer, the activity is at least one to two orders of magnitude lower than the enzyme activity associated with cytotoxicity and apoptosis (51). Furthermore, epithelial and stromal expression of eNOS has been shown in phyllodes tumor, and there is a positive correlation between eNOS expression and tumor grade (50). On this basis, it has been suggested that NO may provide a positive growth signal within the tumor. Our studies have elucidated the potential molecular mechanisms by which this may occur. Further studies are being planned to confirm these in vitro findings in an animal model in vivo.

Grant support: Palomba Weingarten, Allegra Charach Cancer Research Fund, USPHS grant CA-78357 (G. Chaudhuri), grant 1S06-GM068510-03 (R. Singh), RCMI grant G12RR03026, and the Entertainment Industry Foundation Women's Cancer Research Fund.

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.

We thank Janis Cuevas and Svetlana Arutyunova for their technical assistance.