Mechanistic studies of anticancer aptamer AS1411 reveal a novel role for nucleolin in regulating Rac1 activation

Abstract

AS1411 is a G‐rich quadruplex‐forming oligodeoxynucleotide that binds specifically to nucleolin, a protein found on the surface and in the cytoplasm of most malignant cells but absent from the surface/cytoplasm of most normal cells. AS1411 has shown promising clinical activity and is being widely used as a tumor‐targeting agent, but its mechanism of action is not fully understood. Previously, we showed that AS1411 is taken up in cancer cells by macropinocytosis (fluid phase endocytosis) and subsequently stimulates further macropinocytosis by a nucleolin‐dependent mechanism. In the current study, we have investigated the significance and molecular mechanisms of AS1411‐induced macropinocytosis. Our results indicate that the antiproliferative activity of AS1411 in various cell lines correlated with its capacity to stimulate macropinocytosis. In DU145 prostate cancer cells, AS1411 induced activation of EGFR, Akt, p38, and Rac1. Activation of Akt and p38 were not critical for AS1411 activity because Akt activation was not observed in all AS1411‐responsive cell lines and knockdown of p38 had no effect on AS1411's ability to inhibit proliferation. On the other hand, activation of EGFR and Rac1 appeared to play a role in AS1411 activity in all cancer cell lines examined (DU145, MDA‐MB‐468, A549, LNCaP) and their inhibition significantly reduced AS1411‐mediated macropinocytosis and AS1411 antiproliferative activity. Interestingly, downregulation of nucleolin expression by siRNA also produced a substantial increase in activated Rac1, revealing a previously unknown role for nucleolin as a negative regulator of Rac1 activation. Our results are consistent with a model whereby AS1411 binding to nucleolin leads to sustained activation of Rac1 and causes methuosis, a novel type of nonapoptotic cell death characterized by hyperstimulation of macropinocytosis. We speculate that methuosis is a tumor/metastasis suppressor mechanism that opposes the malignant functions of Rac1 and that cancer cells may overexpress nucleolin to surmount this barrier.

Abbreviations

CRO

cytosine-rich oligonucleotide (an inactive control sequence)

G-LISA

GTPase-linked immunosorbent assay

G-rich

guanosine-rich

MP

macropinocytosis

NCL

nucleolin

NIH

National Institutes of Health

RTK

receptor tyrosine kinase

1. Introduction

AS1411 is a 26‐base G‐rich DNA oligonucleotide that functions as a nucleolin‐binding aptamer and has antiproliferative activity against a wide range of cancer cells, but little or no effect on nonmalignant cells (Bates et al., 2009b). AS1411 was the first anticancer aptamer to be tested in humans and the results from Phase 1 and 2 clinical trials show that AS1411 has an excellent safety profile and indications of clinical activity, with many examples of disease stabilization and a few cases of dramatic and long‐lasting objective responses (Bates et al., 2009b; Laber et al., 2005; Rizzieri et al., 2010; Rosenberg et al., 2013). AS1411 is also being widely used as a tumor‐targeting aptamer that can direct a variety of attached molecules or nanomaterials specifically to tumor tissues and to the inside of cancer cells (for examples, see Dam et al., 2014; Shieh et al., 2010).

Nucleolin, the molecular target of AS1411, is a remarkably multifunctional protein that binds selectively to G‐rich DNA or RNA (reviewed in Abdelmohsen et al., 2011; Bates et al., 2009b; Haeusler et al., 2014). Nucleolin is expressed abundantly in the nucleoli of proliferating cells, but can also be found in the nucleoplasm, the cytoplasm, the inner leaflet of the plasma membrane, and on the cell surface, under certain conditions (Borer et al., 1989; Daniely and Borowiec, 2000; Hovanessian et al., 2010; Inder et al., 2009; Semenkovich et al., 1990). The cytoplasmic and cell surface forms of nucleolin have been implicated in cancer progression and are found at constitutively high levels in most cancer cells, but not in non‐transformed cells (Abdelmohsen and Gorospe, 2012; Bates et al., 2009b). AS1411 appears to target primarily the extranuclear functions of nucleolin because the aptamer does not localize to the nucleus in viable cells (Reyes‐Reyes et al., 2010).

We are seeking a better understanding of the mechanism of action of AS1411 in order to explain its ability to selectively target and kill cancer cells. It has been suggested that tumors with certain genetic aberrations or mutations may be especially sensitive to AS1411 (Rosenberg et al., 2013), so mechanistic studies of this agent may ultimately lead to predictive biomarkers that could be used prospectively in a clinical setting to identify patients who are most likely to respond to AS1411. Studying AS1411 has also led to further insights into nucleolin biology; the aptamer has been used by our group and by independent researchers to reveal or confirm a number of new nucleolin functions (Girvan et al., 2006; Pichiorri et al., 2013; Shen et al., 2014; Soundararajan et al., 2008; Teng et al., 2007). These nucleolin‐dependent effects of AS1411 likely contribute to its antiproliferative effects, but none can fully explain AS1411's cancer‐selectivity and ability to target a broad range of malignant cells.

We recently discovered that AS1411 is taken up in cancer cells via macropinocytosis (Reyes‐Reyes et al., 2010), a form of clathrin‐independent endocytosis in which cells “gulp” the surrounding medium and internalize any molecules it contains (Mercer and Helenius, 2009). Furthermore, we found that internalization of AS1411 stimulates additional macropinocytosis, inducing further uptake of molecules from the extracellular medium. Remarkably, AS1411‐mediated stimulation of macropinocytosis requires nucleolin and occurs only in cancer cells and not in non‐malignant cells (Reyes‐Reyes et al., 2010). Although macropinocytosis occurs constitutively in some immune cells, it can also be induced in response to growth factors or due to oncogenic signaling via activation of Ras, PI3K, and Rac1 (Lim and Gleeson, 2011; Mercer and Helenius, 2009; Mooren et al., 2012). Moreover, we and others have noted that macropinocytosis occurs selectively in malignant cells (Commisso et al., 2013; Reyes‐Reyes et al., 2010), and recent evidence suggests that macropinocytosis plays a critical role in sustaining neoplastic growth by providing nutrients via internalization of extracellular proteins (Commisso et al., 2013). On the other hand, hyperstimulation of macropinocytosis might provide a useful method to selectively destroy cancer cells because it can induce a novel form of non‐apoptotic cell death, which has been referred to variously as methuosis, death by macropinocytosis, or catastrophic vacuolation (Kitambi et al., 2014; Li et al., 2010; Maltese and Overmeyer, 2014; Nara et al., 2012; Overmeyer et al., 2008; Overmeyer and Maltese, 2011).

Although our previous research suggests that stimulation of macropinocytosis likely plays a role in AS1411 activity, it is difficult to test directly if blocking macropinocytosis inhibits AS1411‐induced cell death. This is because the biological effects of AS1411 occur over a relatively long time period (induced macropinocytosis at 24–48 h, substantial cell death at 96 h) and macropinocytosis inhibitors such as amiloride are toxic when are applied to cells for more than a few hours. Thus, the goal of the current research was to identify the molecular pathways involved in AS1411‐induced macropinocytosis in order to indirectly investigate the role of macropinocytosis in AS1411 activity and to learn more about the role of nucleolin in this process.

2. Materials and methods

2.1. Materials

Oligodeoxynucleotides, AS1411 (5′‐GGTGGTGGTGGTTGTGGTGGTGGTGG, nucleolin aptamer) and CRO (5′‐CCTCCTCCTCCTTCTCCTCCTCCTCC, inactive control) in the desalted form, and dextran (10,000 Da molecular weight) anionic fixable (dextran‐10K) were purchased from Life Technologies (Grand Island, NY). Anti‐rabbit IgG and anti‐mouse IgG antibodies linked to horseradish peroxidase (HRP), and anti‐GAPDH monoclonal antibody were from Santa Cruz Biotech (Dallas, TX). Anti‐nucleolin monoclonal (MS‐3) and rabbit (H‐250) antibodies were from Santa Cruz. Mouse anti‐rabbit IgG conformation specific (L27A9) linked to HRP, and rabbit anti‐Akt (C67E7), anti‐EGFR (D38B1), anti‐phospho Akt (Thr308) (C31E5E), anti‐p38 (D13E1), anti‐ERK1/2 (137F5), anti‐ubiquitin (P4D1), anti‐phospho p38 (Thr180/Try182), and anti‐phospho ERK1/2 (Thr202/Tyr204) antibodies were from Cell Signaling Technology (Beverly, MA). Anti‐Rac1 (23A8), and anti‐phospho Tyr (4G10^®) monoclonal antibodies were from EMD Millipore (Billerica, MA). Inhibitors for ErbB2 (Cat No. 324732), IGF‐1R (PPP, Cat No. 407247), Met Kinase (SGX523, Cat No. 448106), Tyrosine Kinases (Genistein, Cat No. 345834), Src family of protein tyrosine kinases (PP2, Cat No. 529573), JAK‐2 (AG490, Tyrphostin B42, Cat No. 658401), PDGFR (AG1296, Cat No. 658551), and EGFR (AG1478, Cat No. 658552) were purchased from EMD Millipore. DMSO was from American Type Culture Collection (ATCC).

2.2. Cell culture and treatments

Cell lines were purchased from the American Type Culture Collection (ATCC), except for RCC4, which was a gift from Dr. Wayne Zundel. Cells were grown in the appropriate medium supplemented with 62.5 μg/mL penicillin and 100 μg/mL streptomycin (Life Technologies, Grand Island, NY) in a humidified incubator at 37 °C with 5% CO2. DMEM, RPMI, and F‐12K media were additionally supplemented with 10% fetal bovine serum (FBS, Hyclone Laboratories, Logan, Utah). DMEM medium was used for DU145, A549, MDA‐MB‐231, MDA‐MB‐468, RCC4, Hs27, and MCF‐7 (with 0.01 mg/ml human recombinant insulin for MCF‐7 cells). F‐12K medium was used for PC3 and CHO‐K1 cells. Bronchial Epithelial Cell Growth Medium (BEGM™) and Mammary Epithelial Growth Medium (MEGM^®) supplemented with all of the components of their corresponding Bullet kit, except for GA‐1000 (Lonza), were used respectively for BEAS2B and MCF‐10A cells. Keratinocyte serum free K‐SFM medium supplemented with bovine pituitary extract and EGF (Invitrogen, Carlsbad, CA) was used for RWPE‐1 cells. Cells were plated and incubated for 18 h to allow adherence. Cells at 50% confluence were treated by addition of oligodeoxynucleotides directly to the culture medium to give the final concentration and time indicated in the figure legends. As a positive control for EGFR signaling, cells were starved for 24 h in serum free medium and treated with EGF (100 ng/mL) for 10 min. Signaling pathway inhibitors were dissolved in DMSO and cells were treated in serum free medium before addition of EGF or after AS1411 or CRO treatment, as indicated in the figure legend. For biochemical analyses, cells were lysed in cell lysis buffer [150 mmol/L NaCl, 2 mmol/L EDTA, 50 mmol/L Tris–HCl, 0.25% deoxycholic acid, 1% IGEPAL CA‐630 (pH 7.5)] containing protease and phosphatase inhibitor cocktails (EMD Millipore) for 5 min at 4 °C and then cleared by centrifugation at 16,000 × g for 10 min at 4 °C. All protein concentrations were determined using the bicinchoninic acid assay (Thermo Fisher Scientific, Waltham, MA).

2.3. Cell proliferation assays

The antiproliferative activity was tested using a previously published 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) assay protocol (Bates et al., 2009a). Briefly, 1000 cells were seeded in quadruplicate wells in 96‐well plates and allowed to adhere overnight. Cells were treated with different concentration of oligodeoxynucleotides and incubated for five days, during which the cell culture medium was not changed. The background corresponding to medium alone (no cells) was subtracted. Cell viability was determined by normalizing to the proliferation of untreated cells for each cell type.

2.4. Measurement of macropinocytosis

Cells (6 × 10⁴) in fresh complete medium were plated into six‐well plates. After complete adhesion, cells were incubated as indicated in the figures or legends. After treatment, medium was replaced with fresh medium containing 0.2 mg/mL dextran‐10K labeled with Alexa Fluor 488 for 30 min at 37 °C. Cells were washed once with ice‐cold PBS, incubated with 1 μg/mL propidium iodide (PI) for 1 min on ice to allow exclusion of nonviable cells, and washed twice with ice‐cold PBS. To harvest, cells were treated with 0.01% trypsin/0.5 mmol/L EDTA prior to the addition of ice‐cold supplemented culture medium. Cells were then centrifuged and resuspended in 0.5 mL of 1% paraformaldehyde for analysis by flow cytometry using a FACSCalibur cytometer (BD Biosciences). The % macropinocytosis is defined as the MFI (mean fluorescence intensity) for each condition relative to the MFI for the relevant control cells (specified in figure legends). The term “% increase in MP” refers to the AS1411‐induced increase over baseline, i.e. the % macropinocytosis in the presence of AS1411 minus 100%.

2.5. Immunoprecipitation

DU145 cell lysates (600–1000 μg) were incubated with anti‐EGFR (1:100, Cell Signaling #4267), anti‐Rac‐1 (2.5 μg, Santa Cruz # sc‐95), or anti‐nucleolin (2 μg, Santa Cruz #sc‐13057) antibodies overnight at 4 °C, and then protein complexes were pull down by incubating with 20 μl of Protein A/G‐UltraLink Resin (Thermo Fisher Scientific) for 2 h at 4 °C. The immunoprecipitates were washed once with cold cell lysis buffer and three times with cold PBS.

2.6. Immunoblotting

Total cell lysates or immunoprecipitates of EGFR, Rac1, or nucleolin were resolved by SDS‐Tris PAGE and then electrotransferred onto polyvinylidene fluoride membranes (Millipore) in Tris‐glycine buffer containing 20% methanol. Proteins were detected by immunoblotting. Where indicated, membranes were stripped of bound antibodies using 62.5 mmol/L Tris–HCl (pH 6.7), 100 mmol/L 2‐mercaptoethanol, and 2% SDS for 30 min at 60 °C and then reprobed as described in figure legends.

2.7. Rac1 and Ras activation assays

Measurements of Rac1 activation and Ras activation were performed using the Rac1 G‐LISA Activation Assay Biochem kit (Cat No. BK126, Cytoskeleton, Denver, CO) and the Ras Activation Assay kit (Cat No. 17‐218, Millipore), respectively, according to the manufacturer's recommendations.

2.8. RNA interference

Small interfering RNA (siRNA) duplexes sequences were chemically synthesized and annealed by Life Technologies. The siRNA duplexes for nucleolin (NCL), p38, and Rac1 corresponded to the following sequences: 5′‐GGUCGUCAUACCUCAGAAGtt (NCL#1, ID#144014), 5′‐CGGUGAAAUUGAUGGAAAUtt (NCL#2, ID#144016), 5′‐CCUAAAACCUAGUAAUCUAtt (p38#1, ID#s3585), 5′‐GAAGCUCUCCAGACCAUUUtt (p38#2, ID#s3586), 5′‐CCUUUGUACGCUUUGCUCAtt (Rac1 #1, ID#120600), and 5′‐GGAACUAAACUUGAUCUUAtt (Rac1 #2, ID#s11713). BLAST analysis showed no homology to any sequence in the Human Genome Database, other than the intended target. The scrambled siRNA used were Silencer^® Negative Control #1 siRNA (AM4635) and Silencer^® Select Negative Control #2 siRNA (4390847). The siRNAs were transfected using Lipofectamine™ RNAiMAX (Life Technologies), according to the manufacturer's directions. Briefly, cells were plated and incubated overnight to allow adherence, then transfected with siRNAs for 6–8 h. Cell medium was replaced with fresh complete medium and cells were incubated overnight before being treated as described in the figure legends.

2.9. Immunofluorescence microscopy

Cells (2 × 10⁴) in fresh complete culture medium were plated on 18‐mm diameter glass coverslips for 18 h. Cells were treated as described in the figure legends. After incubation, cells were washed three times with ice‐cold PBS, fixed in 4% paraformaldehyde in PBS for 20 min at room temperature, and washed three times with PBS. Cells were permeabilized for 10 min in PBS containing 0.01% digitonin and washed three times with PBS. Nonspecific binding sites were blocked for 60 min with 3% BSA in PBS, and the fixed cells were incubated for 60 min with primary antibody. Cells were washed three times and incubated with the appropriate secondary antibody for 45 min. All washes and antibody dilutions were carried out using PBS containing 1% BSA. After washing, the cover slips were mounted on glass slides with ProLong Antifade (Molecular Probes) according to the manufacturer's directions to inhibit photobleaching. Immunofluorescence was documented with a Nikon A1R Resonant Scanning Confocal System microscope (Nikon Instruments Inc., Melville, NY) equipped with an Omnichrome argon–krypton laser. Images were obtained with a Plan‐Apo 100× oil immersion objective (1.4 NA).

2.10. Densitometry and statistical analysis

Densitometry was used to measure band intensities by scanning autoradiographic films and using ImageJ software (NIH). Band intensities were normalized as indicated in the figure legends. Where indicated, statistical comparisons between AS1411‐treated and control groups were performed using Student's t test.

3. Results

3.1. AS1411‐mediated macropinocytosis correlates with antiproliferative activity and depends on EGFR activity

Previously, we showed that AS1411 could stimulate macropinocytosis in several cancer cell lines (DU145, MDA‐MB‐231, MCF7), but not in non‐malignant cells (Hs27, MCF‐10A), suggesting a relationship between AS1411‐induced macropinocytosis and antiproliferative activity (Reyes‐Reyes et al., 2010). To confirm and extend these findings, we compared AS1411‐induced macropinocytosis and antiproliferative activity under the same conditions in a larger collection of cancer and non‐cancer cell lines. Induced macropinocytosis was measured at 48 h after adding AS1411 because this is generally when we observed the maximum increase in macropinocytosis in our previous work (Reyes‐Reyes et al., 2010) and because there was negligible AS1411‐induced cell death at this time point (Supplementary Figure S1). Antiproliferative activity was measured 5 days after adding AS1411 to reflect both the cytostatic effect of the aptamer (which was evident from early time points, Supplementary Figure S1A) and its cytotoxicity (which was not apparent until 4 days, Supplementary Figure S1B). Consistently, we observed that AS1411 induced an increase in macropinocytosis in cells that were sensitive to AS1411 (i.e., cancer cells), but AS1411 did not stimulate macropinocytosis in resistant cells (i.e., non‐cancerous cells) (Figure 1A and B, Supplementary Figures S2 and S3). These results confirm that stimulation of macropinocytosis correlates with AS1411 antiproliferative activity.

Figure 1.

AS1411‐induced macropinocytosis correlates with its antiproliferative effects and depends on EGFR. (A) Table showing how various non‐malignant (NM) and cancer (Ca) cell lines respond to 10 μM AS1411 in terms of their percent inhibition of cell proliferation after 5 days (relative to vehicle‐treated cells) and their percent increase in macropinocytosis (MP) over baseline levels at 48 h (relative to vehicle‐treated cells). (B) Graph showing correlation between inhibition of proliferation and macropinocytosis induction. (C) DU145 cells treated with AS1411 (10 μM) or without (no DNA) for 48 h, then treated with various inhibitors or the corresponding vehicle (DMSO) in serum‐free medium for 1 h. Inhibitors were used as follows: general tyrosine kinase inhibitor (60 μM genistein), EGFR (10 μM AG1478), IGF1R (2 μM picropodophyllin, PPP), MET (2 μM SGX523), HER2 (10 μM 4‐[3‐Phenoxyphenyl]‐5‐cyano‐2H‐1,2,3‐triazole), PDGFR (25 μM AG1296), SRC (10 μM PP2), and JAK2 (50 μM AG490). Cellular uptake of the macropinocytic indicator (dextran 10K‐Alexa 488) was analyzed by flow cytometry and shown relative to vehicle‐treated cells. Data are the mean and SE of three independent experiments. (D) Macropinocytosis in A549, LNCaP, and MDA‐MB‐468 cells that were untreated (control) or treated with 10 μM AS1411 followed by incubation with or without 10 μM AG1478 (EGFR inhibitor), as described above. Histograms are representative of at least two independent experiments.

We began our search for regulators of AS1411‐induced macropinocytosis by screening pharmacological inhibitors of receptor tyrosine kinases (RTKs) that have been reported to activate macropinocytosis (Anton et al., 2003; Dowrick et al., 1993; Haigler et al., 1979; Miyata et al., 1988) and their target pathways. We found that pre‐incubating DU145 prostate cancer cells with a selective inhibitor of EGFR (AG1478) led to a pronounced and highly significant (p < 0.001) decrease in AS1411's ability to stimulate macropinocytosis (Figure 1C), whereas inhibiting other RTKs (HER2, MET, PDGFR, IGFR‐1) or blocking Src or JAK2 had little or no effect on macropinocytosis stimulated by AS1411. The EGFR inhibitor was also able to block AS1411‐mediated macropinocytosis in A549 lung cancer, LNCaP prostate cancer, and MDA‐MD‐468 breast cancer cells (Figure 1D), suggesting that activation of EGFR by AS1411 might be involved in the induced macropinocytosis.

3.2. AS1411 activates EGFR, p38, Akt, and Rac1 in DU145 cells

To test the hypothesis that AS1411 treatment activates EGFR, we tested whether AS1411 induces tyrosine phosphorylation of EGFR. Immunoblotting of EGFR immunoprecipitates obtained from lysates of DU145 cells showed that AS1411 induces a modest but reproducible increase in tyrosine phosphorylation of EGFR after 48 h treatment. No changes in EGFR tyrosine phosphorylation levels were seen in untreated cells or cells treated with CRO control oligonucleotide (Figure 2A), or in cells treated with AS1411 for shorter periods of time (2–12 h, data not shown).

Figure 2.

Effect of AS1411 on growth factor signaling pathways in DU145 cells. Cells were treated without added oligonucleotide (–), with 10 μM CRO control (C), or with 10 μM AS1411 (AS) for 24 or 48 h. Where indicated, cells were serum‐starved for 24 h and then stimulated with 100 ng/ml epidermal growth factor (EGF) for 10 min (10′) as a positive control. (A) EGFR immunoprecipitates (IP) obtained from whole cell lysates were tested with antibodies for phospho‐tyrosine (p‐Tyr) or EGFR. (B) Whole cell lysates were analyzed by immunoblotting using antibodies for phospho‐Akt T308 (p‐Akt) and total Akt. (C) As in panel b, blotted for phopho‐p38 (p‐p38) and total p38. (D) As in panel b, blotted for phospho‐ERK1/2 (p‐ERK1/2) and total ERK1/2. (E) The top panel indicates activated Ras (Ras‐GTP), which was precipitated using Raf‐1‐RBD‐agarose and detected by immunoblotting with anti‐pan‐Ras antibody (top panel). The bottom panel shows total Ras in 20 μg of whole cell lysate. (F) Rac1 activity was analyzed using the G‐LISA assay. Data are the mean and SE of three independent experiments.

We next examined whether AS1411 activates pathways that are typically downstream of EGFR, such as Ras, Rac1, p38, Akt, and ERK1/2. Lysates from DU145 cells that were untreated or treated with AS1411 or CRO were analyzed by various methods. Immunoblotting showed that AS1411 was able to increase phosphorylation of Akt (T308) (Figure 2B) and p38 (Figure 2C) after treatment for 24–48 h. Levels of Akt and p38 phosphorylation were not significantly changed compared to untreated, in the control‐treated cells (Figure 2B and C), or in cells treated with AS1411 for shorter times (2, 6, and 12 h, data not shown). High levels of baseline ERK activation were observed in this cell line, but AS1411 did not further increase p‐ERK1/2 and exogenous EGF (as a positive control) did so only modestly (Figure 2D). Ras activity was determined by pull down assay using the Ras‐binding domain (RBD) of Raf to pull‐down Ras‐GTP. All samples showed similarly high levels of active Ras, and AS1411 did not significantly increase Ras activity (Figure 2E). It was recently reported that DU145 cells express a variant KRAS (LC S6) that is not repressed by let‐7 miRNA (Kundu et al., 2012), which would account for the constitutive activation of Ras and its downstream effector, ERK1/2, in this cell line. Activation of Rac1, a small GTPase that plays a critical role in macropinocytosis (Ridley et al., 1992), was measured using the G‐LISA Rac1 activation assay. A significant increase in the levels of activated Rac1 (Rac1‐GTP) was observed after 24 h of AS1411 stimulation and persisted until at least 48 h of stimulation, whereas the control DNA (CRO) had no effect (Figure 2F).

To establish whether AS1411‐mediated activation of Akt, p38 and Rac1 in DU145 cells could be downstream consequences of AS1411‐mediated EGFR activation, cells were incubated with the EGFR inhibitor (AG1478) after AS1411 treatment. Immunoblotting showed that the EGFR inhibitor completely blocked AS1411‐mediated phosphorylation of Akt at T308 (Figure 3A), but did not change AS1411‐induced phosphorylation of p38 (Figure 3B). Surprisingly, the AS1411‐mediated Rac1 activation was only partially dependent on EGFR, with AG1478 producing a slight but significant effect (11.4% decrease, p < 0.05) (Figure 3C). In control experiments, we confirmed that the selected pathways showed a similar dependence on EGFR activity following EGF‐induced activation in this cell line (Supplementary Figure S4).

Figure 3.

Effect of EGFR inhibition on AS1411‐induced signaling. DU145 cells were treated without added oligonucleotide (–), with 10 μM CRO control (C), or with 10 μM AS1411 (AS) for 48 h. Cells were washed with serum‐free medium and treated with DMSO vehicle (Veh) or 10 μM EGFR inhibitor (AG1478) for 1 h. Whole cell lysates were analyzed as follows: (A) Immunoblotting for p‐Akt (T308) and total Akt. (B) Immunoblotting for p‐p38 and total p38. (C) Rac1 activation determined by G‐LISA. Immunoblots are representative of two or three independent experiments; G‐LISA data are the mean and SE for three independent experiments.

3.3. Activation of p38 and Rac1 (but not Akt) is a consistent feature of AS1411 activity

Having identified pathways activated by AS1411, our next objective was to determine what role these play in AS1411 activity. Towards this end, the capacity of AS1411 to induce activation of p38, Akt, and Rac1 was measured in three other AS1411‐responsive cancer cell lines (LNCaP, A549, and MDA‐MB‐468) in order to identify which could be considered universal features of AS1411 activity. We found that AS1411 induced Akt phosphorylation (T308) in A549 cells, but not in LNCaP and MDA‐MB‐468 (Figure 4A). However, p38 activation (Figure 4A) and Rac1 activation (Figure 4B) were observed in all of the cell lines following AS1411 treatment, suggesting that these molecules are potential mediators of AS1411 activity.

Figure 4.

Effect of AS1411 on growth factor signaling pathways in various cancer cell lines. Whole cell lysates from cells treated without added oligonucleotide (–), with 10 μM CRO control (C), or with 10 μM AS1411 (AS) for 48 h were analyzed as follows: (A) Immunoblotting for p‐Akt (T308) and total Akt, and for p‐p38 and total p38. (B) Rac1 activation determined by G‐LISA. Immunoblots are representative of two or three independent experiments; G‐LISA data are the mean and SE for three independent experiments.

3.4. AS1411‐induced macropinocytosis requires Rac1 but not p38

To further define the roles of Rac1 and p38 in the activity of AS1411, we first investigated whether p38 and Rac1 are required for AS1411‐stimulated macropinocytosis. Expression of each of these molecules was knocked down by transfection of the corresponding siRNAs in DU145 cells and immunoblot analysis confirmed >80% reduction in levels compared with controls (Figure 5B and D). As shown in Figure 5A, inhibition of p38 expression caused a small but significant increase in AS1411‐induced macropinocytosis (p < 0.05 for both p38 siRNAs). Inhibition of Rac1 expression caused a substantial decrease in AS1411‐induced macropinocytosis (p < 0.05), as well as a slight reduction in baseline macropinocytosis (Figure 5C). Our results indicate that Rac1 is required for AS1411‐mediated macropinocytosis, but that p38 activation is not and may be a protective stress response to AS1411 treatment.

Figure 5.

Role of Rac1 and p38 in AS1411‐induced macropinocytosis. DU145 cells were transfected as described without siRNA (mock), with two different target‐specific siRNAs, or with a control siRNA (scramble), then treated without added oligonucleotide (no DNA), with 10 μM CRO control, or with 10 μM AS1411 for a further 48 h. Analysis of macropinocytosis (cellular uptake of macropinocytic marker) was then performed by flow cytometry. (A) Macropinocytosis in cells transfected cells with or without p38 siRNAs. (B) Immunoblotting of whole cell lysates for p38 or GAPDH antibodies (loading control) to confirm target knockdown. Relative band intensities were determined using ImageJ software. (C) Macropinocytosis in cells transfected cells with or without Rac1 siRNAs. (D) Immunoblotting of whole cell lysates for Rac1 or GAPDH antibodies (loading control) to confirm target knockdown. Relative band intensities were determined using ImageJ software.

3.5. AS1411 antiproliferative activity depends on Rac1 and EGFR, but not on p38

Next, we determined whether p38 and Rac1 knockdown could affect the antiproliferative activity of AS1411. DU145 cells were transfected with p38 or Rac1 or control siRNAs, as in Figure 5, and treated with varying concentrations of AS1411 for 5 days, then evaluated using the 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) cell proliferation assay. The antiproliferative activity of AS1411 was significantly reduced in cells with Rac1 knockdown expression compared with control‐transfected cells (Figure 6A). Knockdown expression of p38 had no effect on AS1411 antiproliferativity activity (Figure 6B). We attempted similar approaches to assess the role of EGFR, but long‐term knockdown of EGFR expression (using siRNAs) or treatment with AG1478 were highly cytotoxic to DU145 cells (data not shown). Therefore, we assessed the effect of inhibiting EGFR utilizing cetuximab, a recombinant human/mouse chimeric monoclonal antibody that binds the extracellular domain of the receptor and prevents binding to endogenous growth factors (Bou‐Assaly and Mukherji, 2010). We observed that AS1411 activity was reduced by cetuximab, though only when cells were treated with low concentration of AS1411 (Figure 6C). The data in Figure 6 indicate that activation of Rac1 and EGFR likely contribute to AS1411 activity, whereas p38 activation does not. We noted that inhibition of Rac1 or p38 or EGFR in these experiments caused a slight decrease (15–25%) in cell proliferation compared to controls (Figure 6C and Supplementary Figure S5). However, we do not believe that the lower fraction of proliferating cells in these samples can account for the reduced activity of AS1411 in the presence of EGFR antibody or Rac1 siRNAs because p38 siRNAs had no effect on AS1411 activity even though they caused a similar decrease in proliferation.

Figure 6.

Role of Rac1 and p38 in AS1411 antiproliferative effects. DU145 cells were plated at low density and incubated to allow adherence, then treated with various concentrations of AS1411, as indicated. After 5 days of treatment, cell proliferation was measured by MTT assay. Points represent the mean with SE for individual samples from two to three independent experiments. (A) Cells were first transfected without siRNA (mock), with two different Rac1 siRNAs, or with control siRNA (scramble), then treated with AS1411 at two days post‐transfection. (B) Cells were first transfected without siRNA (mock), with two different p38 siRNAs, or with control siRNA (scramble), then treated with AS1411 at two days post‐transfection. (C) Cells were incubated in the absence or presence of cetuximab (50 μg/mL), a monoclonal antibody inhibitor of EGFR activity, which was added at the same time as AS1411 (2.5 μM).

3.6. Nucleolin is negative regulator of Rac1 activation

G‐rich oligonucleotides such as the AS1411 aptamer bind directly and selectively to nucleolin protein (Bates et al., 1999, 2009). Given that Rac1 activation appears to be central to AS1411 activity, it is of interest to determine whether there is a direct or indirect interaction between nucleolin and Rac1. Therefore, we transfected DU145 cells with two different nucleolin siRNAs, a control siRNA (scramble), or without siRNA (mock) and proceeded to assess Rac1 activation. As expected, in mock‐ or control‐transfected cells, Rac1 activity was increased more than two‐fold by 48 h of AS1411 treatment. Surprisingly, knockdown of nucleolin provoked a dramatic increase in Rac1 activation under all conditions (Figure 7A), without affecting Rac1 protein levels (Figure 7B). This result indicates that nucleolin is a negative regulator of Rac1 activation, which is a novel finding. Furthermore, knockdown of nucleolin expression—although it increased baseline Rac1 activation—abrogated further Rac1 activation by AS1411 (Figure 7A), suggesting that AS1411‐stimulated Rac1 activation is nucleolin‐dependent. Additional experiments (Supplementary Figure S6) showed that nucleolin and Rac1 do not co‐immunoprecipitate, either with or without AS1411 treatment, so there does not appear to be a direct interaction between nucleolin and Rac1.

Figure 7.

Effect of nucleolin knockdown on Rac1 activation. DU145 cells were transfected as described without siRNA (mock), with two different nucleolin (NCL) siRNAs, or with control siRNA (scramble), then treated without added oligonucleotide (no DNA), with 10 μM CRO control, or with 10 μM AS1411 for a further 48 h. (A) Cell lysates were processed to analyze Rac1 activation by G‐LISA. Data represent the mean and SE for three independent experiments. (B) Immunoblotting of whole cell lysates for nucleolin (NCL), Rac1, and GAPDH. Relative band intensities are indicated and were determined using ImageJ software. Immunoblots are representative of two independent experiments.

3.7. AS1411 induces methuosis‐like cellular vacuolation

Sustained activation of Rac1 in malignant cells has been previously reported to induce a novel form of non‐apoptotic cell death known as methuosis (Bhanot et al., 2010). Rac1‐induced methuosis is characterized by hyperstimulation of macropinocytosis combined with aberrant vesicle trafficking, which leads to intense cellular vacuolation and ultimately to cell lysis (Bhanot et al., 2010). To explore the possibility that AS1411 induces methuosis, we examined AS1411‐treated cancer cells for evidence of characteristic cytoplasmic vacuolation. Most of the cancer cell lines treated with AS1411 exhibited morphological changes consistent with methuosis, as shown in Figure 8A. The one exception was LNCaP cells, which displayed neuron‐like differentiation in response to AS1411 (Supplementary Figure S7).

Figure 8.

Effect of AS1411 and nucleolin knockdown on cancer cell morphology. (A) Photomicrographs showing various cancer cells incubated in the absence (control) or presence of 10 μM AS1411 for 48 h. Note the distinctive cytoplasmic vacuolation, which is characteristic of methuosis. (B) Photomicrographs showing DU145 transfected as described without siRNA (none), with two different nucleolin siRNAs (NCL #1, NCL #2), or with control siRNA (scramble), then incubated for a further 48 h. (C) Photomicrographs showing DU145 transfected as described without siRNA (none), with two different nucleolin siRNAs (NCL #1, NCL #2), or with control siRNA (scramble), then treated with 10 μM AS1411 for a further 48 h.

Based on our finding that nucleolin siRNAs strongly activate Rac1 (Figure 7), we at first anticipated that these too would induce the methuosis‐like vacuolation. However, when we examined DU145 cells with knocked down nucleolin, we did not observe the same morphology as when the cells were treated with AS1411. As shown in Figure 8B and Supplementary Figure S8, nucleolin siRNAs induced a variety of abnormal cell morphologies, including flattened cells, rounded cells, neuron‐like cells with very long projections, multinucleated cells, and cells with a large number of small cytoplasmic/perinuclear vacuoles. Many of the cells had pronounced lamellipodia and filopodia, which is consistent with the expected Rac1 activation, but few had the large vacuoles that are a hallmark of methuosis. When the same experiments were performed in the presence of AS1411, we noted that knocking down nucleolin inhibited AS1411‐induced cellular vacuolation (Figure 8C), suggesting that this process is nucleolin‐dependent. There are several factors that may explain why knocking down nucleolin expression using siRNAs leads to different outcomes than targeting nucleolin with AS1411 and these are discussed below.

4. Discussion

The results of the current study have confirmed our previous work (Reyes‐Reyes et al., 2010) suggesting that stimulation of macropinocytosis by AS1411 is important for its activity. They have also revealed a key role for Rac1 in the mechanism of AS1411 action, as Rac1 was markedly activated by AS1411 in all responsive cell lines examined and inhibition of Rac1 abrogated the antiproliferative effects of AS1411. Moreover, we have identified a previously unknown connection between Rac1 and nucleolin, which is the molecular target of AS1411. Our data indicate that inhibition of nucleolin function—by either siRNA or AS1411—can lead to strong activation of Rac1.

The discovery that an anticancer agent works by increasing macropinocytosis and elevating Rac1 activity may seem paradoxical at first, because both of these actions are typically associated with promoting tumor growth and progression (Bid et al., 2013; Commisso et al., 2013). However, recent research has indicated that excessive Rac1 activation can instead lead to cell death (Bhanot et al., 2010; Overmeyer and Maltese, 2011). Maltese and colleagues found that ectopic expression of activated Ras or Rac1 in some cancer cells can trigger a distinct form of non‐apoptotic cell death caused by hyperstimulation of macropinocytosis (Bhanot et al., 2010; Maltese and Overmeyer, 2014; Overmeyer et al., 2008). This novel phenomenon, dubbed “methuosis”, was found to involve sustained activation of Rac1, leading to Arf6 inactivation and thereby causing a defect in vesicle trafficking (Bhanot et al., 2010). Thus, Rac1‐induced macropinosomes cannot recycle to the surface or fuse with lysosomes as they would normally do; instead, they fuse with each other and late endosomes to create large vacuoles that can be easily observed by light microscopy; these growing vacuoles eventually cause the cells to lyse and die by a necrosis‐like process 4–6 days after Ras/Rac1 induction (Bhanot et al., 2010; Overmeyer et al., 2008). At least two additional independent groups have subsequently described a similar “death by macropinocytosis” (Li et al., 2010; Nara et al., 2012), while Maltese and others have identified small molecules that specifically induce methuosis (Kitambi et al., 2014; Robinson et al., 2012).

Methuosis could also explain the apparent paradox that nucleolin, a malignancy‐associated protein that is overexpressed in cancer cells, is a negative regulator of a process that is typically associated with cancer progression (i.e. Rac1 activation). We speculate that methuosis evolved as a tumor/metastasis suppressor mechanism that serves to limit the pro‐tumorigenic and pro‐invasive properties of Rac1. As with other tumor suppressor mechanisms (e.g. oncogene‐induced senescence), cancer cells can develop adaptive mechanisms to avoid death by methuosis. We hypothesize that upregulation of non‐nuclear nucleolin to limit activation of Rac1 is one such mechanism, and that inhibition of this nucleolin function by AS1411 can restore methuosis, as illustrated in Figure 9A.

Figure 9.

Proposed mechanism for AS1411‐induced cancer cell death. (A) It has been previously shown that transient activation of Rac1 leads to increased cell proliferation and migration, which promotes tumor progression, whereas sustained activation of Rac1 can lead to a novel form of cell death known as methuosis. New data presented herein indicate that nucleolin is a negative regulator of Rac1 activation. We propose that methuosis is a tumor/metastasis suppressor mechanism that opposes the malignant functions of Rac1 and that cancer cells overexpress nucleolin in order to bypass Rac1‐induced methuosis. In the presence of the AS1411 aptamer, nucleolin's ability to limit Rac1 activation is compromised and methuosis is induced. AS1411‐induced activation of EGFR is observed at later time points (compared to Rac1 activation), suggesting that it occurs downstream of Rac1 activation or is an independent effect of AS1411. (B) Confocal microscopy showing EGFR localization in DU145 cells incubated with or without 10 uM AS1411 for 48 h. EGFR was visualized by indirect immunofluorescence (green) and nuclei were stained with DAPI (blue). (C) DU145 cells were treated without added oligonucleotide (–), with 10 μM CRO control (C), or with 10 μM AS1411 (AS) for 48 h, or were serum‐starved for 24 h and then stimulated with 100 ng/ml epidermal growth factor (EGF) for 10 min (10′) as a positive control. EGFR was immunoprecipitated from whole cell lysates and then immunoblotted for ubiquitin or EGFR.

If this model is correct, one might predict that inhibition of nucleolin expression by siRNAs would also lead to sustained Rac1 activation, hyperstimulation of macropinocytosis, and induction of methuosis. However, our studies have shown that whereas nucleolin siRNAs do induce strong Rac1 activation (Figure 7), they lead to neither macropinocytosis stimulation ((Reyes‐Reyes et al., 2010) and Supplementary Figure S9) nor methuosis‐like vacuolation (Figure 8). Although initially puzzling, we believe there are several important differences between AS1411 and nucleolin siRNA that could explain these apparent discrepancies. One possibility is that effects on the nuclear functions of nucleolin are responsible for the morphological changes seen in cells treated with nucleolin siRNA. Our previous work has shown that AS1411 targets only a small subset of the total nucleolin in the cell (Teng et al., 2007), which is presumably limited to the cell surface and/or cytoplasm‐localized forms of nucleolin because AS1411 does not enter the nucleus (Reyes‐Reyes et al., 2010). Thus, only non‐nuclear nucleolin is targeted by AS1411 and only selected functions of nucleolin will be affected by it. Conversely, knocking down nucleolin by siRNA will inhibit all functions of nucleolin, including nuclear functions related to ribosome biogenesis, cell cycle progression, and mitosis. A second possibility is that the methuosis‐like effect of AS1411 also involves another nucleolin‐dependent pathway. AS1411 binds to nucleolin and appears to modulate its interactions with multiple binding partners, which could either increase or decrease the activity of the nucleolin‐containing complex (by contrast, siRNA blocks production of nucleolin and inhibits all activities). It is possible that the effects of AS1411 require sustained activation of Rac1 (by inhibiting nucleolin‐mediated suppression) combined with activation of another pathway—which has not yet been identified—that is promoted when the aptamer binds to nucleolin but not when nucleolin is knocked down by siRNA. Interestingly, knocking down nucleolin inhibits both induced macropinocytosis ((Reyes‐Reyes et al., 2010) and Supplementary Figure S9) and the appearance of the very large vacuoles in cells that are subsequently treated with AS1411 (Figure 8C), suggesting that the presence of nucleolin is critical for these processes.

The role of EGFR in AS1411 activity is not clear‐cut at present. Our initial assumption was that EGFR would lie upstream of Rac1 activation, but the data suggest that AS1411‐induced activation of Rac1 is only partially dependent on EGFR activity and also that Rac1 activation precedes EGFR activation. Therefore, it appears that AS1411‐induced activation of EGFR (which is relatively modest anyway) is either coincidental or a downstream result of Rac1 activation. One possibility is that altered vesicle trafficking (due to Rac1 hyperactivation) leads to internalization of activated EGFR and allows continued signaling from endosomes. This model is supported by our recent experiments, which show that EGFR becomes accumulated in endosomes following treatment with AS1411 (Figure 9B), but is not ubiquitylated to mark it for degradation (Figure 9C). Interestingly, Ceresa and colleagues have found that mislocalization of activated EGFR to endosomes can result in cell death (Hyatt and Ceresa, 2008; Rush et al., 2012), but the relevance of this to AS1411 remains to be seen. In accord with previous reports (Farin et al., 2009), we can detect an association between EGFR and nucleolin, and this is increased in the presence of AS1411 (Supplementary Figure S10), consistent with the increased activation of EGFR under these conditions. These data are in contrast with a recent report that AS1411 inhibits the association of EGFR with nucleolin and decreases EGFR‐mediated signaling (Goldshmit et al., 2014). The inconsistencies between this recent paper and our current findings may be due to the use of different cell lines or conditions, or may add further weight to our suggestion that the activation of EGFR observed in our case may be a coincidental or secondary effect. Nonetheless, it remains that AS1411‐induced macropinocytosis was completely inhibited by a small molecule inhibitor of EGFR (AG1478) in multiple cell lines (Figure 1), so continued investigation of these results—including possible off‐target effects of AG1478 (Douglas et al., 2009; Pan et al., 2008)—appears to be warranted. Indeed, these results may have important clinical implications because they suggest that combining EGFR inhibitors with AS1411 might have an antagonistic effect.

Another high‐priority for future studies is to determine whether Ras plays a role in AS1411 activity. Nucleolin has been shown by several groups to co‐precipitate with Ras and to modulate Ras signaling (Birchenall‐Roberts et al., 2006, 2011, 2010, 2009) and we have previously shown that Ras‐transfected cells are more susceptible to AS1411 than controls (Farin et al., 2011). In the current study, we have found that Ras activation is not significantly increased by AS1411 in DU145 cells. However, as mentioned earlier, Ras is constitutively active in this cell line, which may obscure any AS1411‐induced changes, so more experimentation will be needed to fully elucidate the role of Ras in AS1411 activity.

In conclusion, our study has identified a previously unknown role for nucleolin in regulating Rac1 activation and has provided new information regarding the mechanism of action of the nucleolin‐binding aptamer, AS1411. Further research to confirm AS1411‐induced methuosis and elucidate the molecular mechanisms involved may lead to important discoveries in cancer cell biology and to novel strategies for targeting cancer cells.

Funding statement

This research was sponsored by NIH R01 CA122383 grant to PJB, NIH R25 CA134283 summer fellowship to MF, and the Brown Cancer Center. The sponsors had no role in the study design, data collection/analysis/interpretation, writing the report, or the decision to submit for publication.

Conflicts of interests statement

EMRR and PJB are co‐inventors on issued or pending patents related to this work.

Supporting information

Supplementary data

Supplementary data 1.

1.1.

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.molonc.2015.03.012.

Reyes-Reyes E. Merit, Šalipur Francesca R., Shams Mitra, Forsthoefel Matthew K., Bates Paula J., (2015), Mechanistic studies of anticancer aptamer AS1411 reveal a novel role for nucleolin in regulating Rac1 activation, Molecular Oncology, 9, doi: 10.1016/j.molonc.2015.03.012.