Antisense Oligonucleotide Targeting Eukaryotic Translation Initiation Factor 4E Reduces Growth and Enhances Chemosensitivity of Non-Small-Cell Lung Cancer Cells

Home , Eukaryotic translation

ORIGINAL ARTICLE Antisense oligonucleotide targeting eukaryotic translation initiation factor 4E reduces growth and enhances chemosensitivity of non-small-cell lung cancer cells

SC Thumma1, BA Jacobson1, MR Patel1, BW Konicek2, MJ Franklin1, J Jay-Dixon1, A Sadiq1,ADe1, JR Graff2 and RA Kratzke1

Elevated levels of eukaryotic translation initiation factor 4E (eIF4E) enhance translation of many malignancy-related proteins, such as vascular endothelial growth factor (VEGF), c-Myc and osteopontin. In non-small-cell lung cancer (NSCLC), levels of eIF4E are significantly increased compared with normal lung tissue. Here, we used an antisense oligonucleotide (ASO) to inhibit the expression of eIF4E in NSCLC cell lines. eIF4E levels were significantly reduced in a dose-dependent manner in NSCLC cells treated with eIF4E-specific ASO (4EASO) compared with control ASO. Treatment of NSCLC cells with the 4EASO resulted in decreased cap- dependent complex formation, decreased cell proliferation and increased sensitivity to gemcitabine. At the molecular level, repression of eIF4E with ASO resulted in decreased expression of the oncogenic proteins VEGF, c-Myc and osteopontin, whereas expression of β-actin was unaffected. Based on these findings, we conclude that eIF4E-silencing therapy alone or in conjunction with chemotherapy represents a promising approach deserving of further investigation in future NSCLC clinical trials.

Cancer Gene Therapy (2015) 22, 396–401; doi:10.1038/cgt.2015.34; published online 31 July 2015

INTRODUCTION eIF4E. These structural differences account for why mRNAs differ In 2012, worldwide, an estimated 1.8 million new lung cancer in their dependence on eIF4E. A recent report has also shown that cases occurred and an estimated 1.5 million patients succumbed eIF4E promotes nuclear export of mRNAs containing an eIF4E- 6 to the disease. Lung cancer was the most frequently diagnosed sensitivity element (4E-SE) in their 3′ UTR. These observations cancer and caused the most deaths in 2012.1 New therapeutic indicate that eIF4E enhances the translation of a specific group of agents are needed for this fatal disease. mRNAs by increasing their cytoplasmic levels and the efficiency Deregulation of translation initiation has been implicated in the with which they associate with the eIF4F complex. Importantly, pathogenesis of several cancer types.2 Under normal conditions, these mRNAs encode proteins involved in cellular growth and translation initiation is tightly regulated by the availability of survival, including many proto-oncogenes, such as vascular eukaryotic initiation factor 4E (eIF4E). Free eIF4E binds mRNA via endothelial growth factor (VEGF), c-Myc, survivin, Bcl-2 and cyclin the 5′ 7-methyl-GTP-cap structure, after which it brings these D1.7 It has been shown that eIF4E overexpression causes mRNAs to the eIF4F complex for unwinding of mRNA secondary malignant transformation in several in vitro and in vivo models. structure, which allows the 40S ribosomal subunit to scan through For example, NIH-3T3 cells overexpressing eIF4E develop a the 5′ untranslated region (UTR) until it recognizes the start codon malignant phenotype.8 In addition, overexpression of eIF4E using AUG. The availability of free eIF4E is controlled by eIF4E-binding a β-actin promoter produces tumors in many tissues.9 Also, eIF4E proteins (4E-BPs), which sequester eIF4E and block its association levels have been shown to be elevated in surveys of different with the eIF4F complex. 4E-BP1 binding to eIF4E is regulated by human malignancies of the prostate,10 breast,11 stomach,12 several pathways including the Ras/extracellular signal regulated colon13 and skin14 as well as cancers of the hematopoietic kinase (ERK) and phosphatidylinositol-4,5-bisphosphate 3-kinase system.15 In non-small-cell lung cancer (NSCLC), levels of eIF4E are (PI3 kinase)/AKT/mammalian target of rapamycin (mTOR) path- substantially increased compared with normal lung tissue,16 with ways. Phosphorylation of 4E-BP1 by these signaling pathways increased eIF4E expression being associated with tumor inactivates 4E-BP1 freeing eIF4E to interact with eIF4G to form the invasiveness.17 Repression of eIF4E activity, however, decreases eIF4F complex.3 the tumorigencity of NSCLC cells in xenograft models.18 These Specific RNA motifs related to eIF4E have led to the notion that results suggest that targeting eIF4E and decreasing its availability eIF4E coordinates the translation of proteins that function in cell will lead to decreased eIF4F function and hence suppress NSCLC growth and survival pathways.4 The translation of mRNAs with tumor growth. complex 5′UTRs are particularly dependent on eIF4E because of Given that eIF4E acts as an oncogene when it is overexpressed the inefficiency of scanning for the start codon in long, highly or aberrantly activated, investigators are in the process of structured 5′UTRs.5 In contrast, mRNAs with short, unstructured developing drugs that specifically target eIF4E in an attempt to 5′UTRs are translated efficiently in the presence of low levels of decrease its level or affect its hyperactivity in translation initiation.

1Division of Hematology, Department of Medicine, Oncology and Transplantation, University of Minnesota Medical School, Minneapolis, MN, USA and 2Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN, USA. Correspondence: Dr RA Kratzke, Division of Hematology, Oncology and Transplantation, University of Minnesota Medical School, 420 Delaware Avenue SE, MMC 480, Minneapolis, MN 55455, USA. E-mail: [email protected] Received 17 April 2015; revised 30 June 2015; accepted 10 July 2015; published online 31 July 2015 4EASO as a therapy for non-small-cell lung cancer SC Thumma et al 397 Disruption of the eIF4E-to-eIF4G interaction is one approach eIF4E-specific ASO (4EASO) targets the 3′ UTR of human and to attenuate eIF4E-mediated tumorigenicity. A compound desig- murine eIF4E mRNA using the sequence, 5′-TGTCATATTCCTGGA nated as 4EGI-1, which mimicked the inhibitory properties of 4E- TCCTT-3′, where the underlined nucleotides are 2’-methoxyethyl BP1, was identified following a high-through put screen.19 4EGI-1 modified. The ASO control was designed with the same base was found to suppress cancer cell growth along with attenuation composition as the 4EASO but with a mismatched sequence of of c-Myc and Bcl-XL oncoprotein expression. Another tactic to 5′-TCTTATGTTTCCGAACCGTT-3′. target the eIF4E-to-eIF4G association is to design 4EBP-based For ASO cell treatment, Oligofectamine (Invitrogen, Life Techno- eIF4E-binding peptides that block eIF4E from binding eIF4G. logies, Waltham, MA, USA) was mixed 1:16 with Opti-MEM I (Gibco, A compound based upon a 19-residue fragment of 4E-BP1 was Life Technologies) and incubated at room temperature for 5 min. generated and shown to inhibit xenograft tumor growth in an The appropriate concentration of ASO was then diluted in Opti- ovarian cancer model.20 4E1RCat, a small-molecule inhibitor of the MEM I and mixed with the Oligofectamine-Opti-MEM I mixture eIF4E-to-eIF4G interaction was developed recently and shown to and incubated for 45 min. Cells were washed with RPMI 1640 impede eIF4E-to-eIF4G interaction by binding to a region on eIF4E (Gibco, Life Technologies) followed by addition of fresh RPMI 1640 that also binds both eIF4G and 4EBP1. This compound was shown on adherent cells. This transfection solution was added to cells to to reverse tumor chemoresistance in a genetically engineered achieve a 1:4 ratio between the transfection solution and RPMI lymphoma mouse model.21 1640 followed by incubation of this mixture for 4 h at 37 °C in a Another approach to directly target eIF4E hyperactivity is to CO2 incubator. After transfection, RPMI 1640 containing sufficient antagonize the eIF4E-mRNA–cap interaction. Therapies that either serum was added to cells until a final serum concentration of decrease eIF4E abundance or hinder the association of eIF4E with 10% was reached. Cells were harvested or counted after 72 h the 5’ mRNA cap would accomplish this objective. Two teams of of incubation. For cell proliferation analysis, cells were seeded investigators have developed cap-analog libraries that function to onto six-well plates at 125 000 cells per well 1 day before ASO either prevent binding of eIF4E to capped mRNAs or dislodge transfection. For all remaining studies, 10-cm plates were seeded capped mRNAs from eIF4E and therefore the eIF4E complex.22,23 with 1 × 106 cells per plate one day before ASO transfection. One promising cap-analog, 4Ei1,24 is a prodrug that is converted to its functionally active derivative, 7-BnGMP (7-benzyl guanosi- Cell viability assay with ASO and gemcitabine nemonophosphate) in human breast carcinoma cells where it Cells seeded onto six-well plates for 1 day were then transfected has been shown to downregulate cap-dependent translation. with ASOs as described above and incubated overnight. After 24 h When breast carcinoma cancer cell lines were treated with 4Ei1 of incubation with 4EASO or its control, gemcitabine (Eli Lilly) was fl combined with the drug 5- uorouracil there was enhanced added to the appropriate wells at a final concentration of 8 nM cytostatic and/or cytotoxic effects compared with each compound 25 and incubated for an additional 48 h. Cells were then counted using alone. One tactic that was employed to decrease eIF4E a hemacytometer with trypan blue exclusion. Each experiment was abundance in breast carcinoma subtypes was with treatments performed in triplicate, with data normalized to that of untreated 26 that used short-interfering RNA against eIF4E. eIF4E knockdown cells. Group comparisons were performed using two-sided Student’s led to growth inhibition of breast cancer cell lines and also t-test and a P value of o0.05 was taken as significant. decreased the expression of the oncogenic proteins, Cyclin D1, Bcl-2 and Bcl-xL. A second approach to reducing eIF4E abundance Preparation of cell lysates is by using a nuclease resistant, second-generation, antisense oligonucleotide (ASO) that recognizes eIF4E mRNA (4EASO) by Cells were washed with phosphate-buffered saline, scraped from Watson-Crick base pairing and triggers RNase-H-mediated RNA the plate in phosphate-buffered saline, centrifuged (14 K r.p.m. for 27 14 s), washed with ice-cold phosphate-buffered saline, centrifuged destruction. This strategy effectively uses ASOs to downregulate fi eIF4E protein expression in breast and prostate tumor xenografts, and resuspended in ve times the pellet volume with freeze-thaw notably reducing tumor growth.27 4EASO was also shown to lysis buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 50 mM NaF, 1 mM EDTA, 10 mM tetrasodium pyrophosphate) supple- reduce eIF4E levels, eIF4F complex formation and proliferation of mented with a protease inhibitor cocktail (Roche, Penzberg, mesothelioma cells.28 In addition, results from a phase I/II clinical Germany) and a phosphatase inhibitor cocktail (Sigma-Aldrich). trial of 4EASO in patients with advanced cancer has recently been Following three freeze-thaw cycles (15 min at –80 °C, 2 min at published.29 A reduction of up to 80% of eIF4E mRNA was realized 37 °C), protein concentrations of cell lysate were determined by in patient tumors but did not result in tumor response. It was Bradford assay (Bio-Rad, Hercules, CA, USA) before storing lysate suggested that 4EASO be examined with combined therapies. In at –80 °C. the present study, we use 4EASO to reduce eIF4E levels in NSCLC cell lines and assess its effect on expression of oncogenic proteins, cap-dependent complex formation, tumor cell proliferation and Cap-dependent complex assessment sensitivity to chemotherapy. The potency of cap-dependent complex formation was determined as before;18 briefly, 300 μg of lysate diluted in 300 μloffreeze-thaw lysis buffer were incubated while mixing with 50 μlofsuspended MATERIAL AND METHODS (50% mixture) 7-methyl GTP-SepharoseTM4B (Amersham Bios- Cell culture and ASO treatment ciences, Piscataway, NJ, USA) for 2 h at 4 °C. The captured eIF4E μ μ NSCLC cell lines H838 and H2009 and H522 were obtained from was eluted with 35 l of freeze-thaw lysis buffer containing 100 M ′ the American Tissue Culture Collection (Manassas, VA, USA). of 7-methylguanosine 5 -triphosphate (Sigma-Aldrich) followed by heating at 95 °C and subsequent sample removal from 7-methyl H2009 and H522 cells were grown in RPMI (Gibco, Life Technol- TM ogies, Waltham, MA, USA) containing 10% calf serum (Sigma- GTP-Sepharose 4B. Samples were then prepared for immunoblot Aldrich, St Louis, MO, USA), whereas H838 cells were grown in analysis. The density of protein bands was measured using ImageJ RPMI supplemented with 10% calf serum, 10 mM HEPES, 1 mM (US National Institutes of Health, Bethesda, MD, USA), a public sodium pyruvate, 4.5 g l − 1 glucose and 1.5 g l − 1 sodium bicarbo- domain Java image processing program. nate. All cells were maintained at 37 °C in 5% CO2. The ASOs were provided by Eli Lilly and Company (Indianapolis, IN, USA) and were Immunoblot analysis 20 nucleotides in length containing five 2’-methoxyethyl-modified Cap-affinity assay samples were separated by SDS-polyacrylamide bases flanking both sides of 10 central, non-modified bases.27 The gel electrophoresis utilizing 8–15% gradient gels, whereas

all other samples were subjected to separation by 12% Mismatch ASO 4EASO SDS-polyacrylamide gel electrophoresis. Following electrophoresis, proteins were transferred to Hybond-P, poly(vinylidene eIF4E fl 1 0.63 0.53 0.45 Band intensity uoride) membrane (Amersham Biosciences). Blocking of mem- -actin branes was done in 5% non-fat dry milk for 1 h at room tem- H2009 perature in Tris-buffered saline-Tween 20 (0.15 M NaCl; 0.01 M 0 50 100 200 0 50 100 200 nM ASO Tris-HCl, pH 7.6; 0.05% Tween 20). The membranes were then rinsed, and incubated for 1 h at room temperature or overnight at eIF4E 1 0.22 0.23 0.28 0.32 Band intensity 4 °C with the appropriate primary antibody. Blots were probed separately with rabbit α-eIF4E antibody (Cell Signaling Technol- H838 -actin ogy, Danvers, MA, USA; catalog number 9742) at a 1:1000 dilution, 0 50 100 200 400 0 50 100 200 400 nM ASO rabbit α-c-Myc antibody Cell Signaling Technology; catalog α number 9402) at a 1:1000 dilution, -Osteopontin (ThermoFisher eIF4E fi Scienti c, Waltham, MA, USA, catalog number RB-9097-P1) at a 1 0.3 0.18 0.04 0.03 0.04 Band intensity α fi 1:100 dilution, mouse -VEGF antibody (ThermoFisher Scienti c, H522 -actin catalog number MS-1467-P0) at a 1:200 dilution and mouse α-β- actin (Sigma-Aldrich, St Louis, MO, USA, catalog number A1978) at 25 50 100 200 400 0 25 50 100 200 400 nM ASO a 1:10 000 dilution. Blots were washed three times for 5 min in Tris-buffered saline-Tween 20 before and following an hour Mismatch ASO 4EASO incubation at room temperature with the appropriate horseradish peroxidase-labeled secondary antibodies. Detection was carried eIF4E out using ECL Plus Western Blotting System (Amersham Bios- ciences) to visualize bands of interest. The protein band density 1 0.64 0.42 Band intensity was determined employing ImageJ, a Java image processing 0 50 400 0 50 400 program from the public domain. eIF4E

Statistical analysis H2009 H838 1 0.67 0.53 Band intensity Data are presented as means and error bars are standard devia- 0 100 200 0 100 200 tion. Differences between groups were compared using two-sided t-tests. A P value of less than 0.05 was taken as significant. All of Antisense oligonucleotide [nM] the experiments shown in Figures 2 and 4 were performed in Figure 1. Eukaryotic translation initiation factor 4E (eIF4E) expres- triplicate, with data normalized to that of untreated cells. sion and cap-binding activity in non-small-cell lung cancer (NSCLC) cell lines following treatment with antisense oligonucleotide against eIF4E (4EASO) or its mismatched control. (a) eIF4E protein RESULTS expression. H2009, H838 and H522 cells were transfected with 4EASO or its mismatched control and incubated for 72 h followed by 4EASO suppresses eIF4E expression and cap binding β b fi immunoblot analysis. -Actin served as a loading control. ( ) eIF4E We rst evaluated the ability of 4EASO to suppress eIF4E protein cap-binding. A cap-affinity assay was performed to evaluate the levels in NSCLC cell lines. H2009, H838 and H522 cells, all of which level of eIF4E to bound to the 5’ mRNA cap. The indicated NSCLC cell have been shown to possess activated eIF4F,18 were transfected lines were treated as described followed by cap-analog capture of with the 4EASO or its mismatched control and incubated for 72 h cell lysates using 7-methyl-GTP-Sepharose beads and immunoblot- before assessing eIF4E protein expression by immunoblot analysis. ting for eIF4E. The band strength levels for eIF4E was normalized to Cells treated with the 4EASO showed a dose-dependent decrease untreated cells for each cell line and was measured employing in eIF4E protein level, whereas mismatched ASO showed no such ImageJ. effect on any of the tested NSCLC cell lines (Figure 1a). The band densities indicate that in terms of eIF4E suppression by 4EASO exposure, H2009 is least sensitive, whereas H522 is most sensitive. β-Actin expression was unaffected supporting previous findings H522 (Figure 2). H838 cells appeared to be particularly sensitive to that low eIF4E levels have no effect on the translation of ‘strong’ 4EASO with an ~ 70% decrease in growth at 50 nM concentration mRNAs like that of β-actin. Suppression of eIF4E with 4EASO also compared with untreated cells. Mismatched ASO also had an resulted in a dose-dependent decrease in eIF4E bound to a cap- inhibitory effect upon NSCLC cell growth; however, this effect was analog; however, the amount of eIF4E bound to the cap-analog substantially less than 4EASO at each concentration tested. was unaffected by mismatched ASO (Figure 1b). The eIF4E band densities following treatment are indicated. Therefore, treatment 4EASO suppresses expression of eIF4E-regulated proteins with 4EASO not only reduces eIF4E expression, but also reduces Many key oncogenes and growth factors are regulated at the level eIF4F complex formation. of translation and are repressed in normal cells, when the level and activity of eIF4E are limited. When overexpressed, eIF4E acts 4EASO inhibits NSCLC cell growth as an oncogene and greatly facilitates translation of various eIF4E overexpression promotes cell growth and survival and has mRNAs coding for proteins critical to angiogenesis and growth been shown to enhance tumorigenesis in different tissues.9 including VEGF, c-Myc, cyclin D1 and such antiapoptotic proteins 30 Having shown that we could effectively decrease the expression as Bcl-XL and survivin. We evaluated whether a 4EASO-mediated of eIF4E, we wanted to assess the effects that this intervention decrease in eIF4E levels would influence the expression of would have upon the growth characteristics of NSCLC cells. Cells translationally regulated proteins. We chose to evaluate the treated with the 4EASO in concentrations ranging from 25 to angiogenesis factor VEGF and oncogene c-Myc in the NSCLC cell 400 nM revealed a dose-dependent decrease in cell growth line H838, where a considerable decrease in expression was compared with mismatched ASO, with an estimated IC50 less observed in cells treated with 4EASO compared with cells treated than 50 nM for both H2009 and H838 and close to 100 nM for with mismatched control (Figure 3a). In H838 cells at 400 nM VEGF

Cancer Gene Therapy (2015), 396 – 401 © 2015 Nature America, Inc. 4EASO as a therapy for non-small-cell lung cancer SC Thumma et al 399 H2009 H838 H522 and c-Myc decreased to 62% of the level of untreated cells. A 1.2 similar decrease was observed in VEGF (27%) and osteopontin

1 (57%) in the H2009 cell line treated with 200 nM 4EASO compared with untreated cells (Figure 3b). 0.8 4EASO sensitizes NSCLC cells to gemcitabine 0.6 We have reported previously that repression of cap-mediated translation results in enhanced cell killing by cytotoxic agents.18 Cell viability 0.4 Gemcitabine is a pyrimidine antimetabolite that inhibits DNA synthesis by inhibition of DNA polymerase and ribonucleotide

[normalized to untreated] 0.2 reductase, and is an active chemotherapy agent in NSCLC and 0 many other cancers. We wanted to evaluate whether 4EASO- 0 0 0 25 50 100 200 400 50 100 200 400 25 50 100 200 400 mediated repression of cap-dependent translation promoted nM antisense oligonucleotide enhanced cell death when combined with gemcitabine in comparison to treatment with either agent alone. We treated Figure 2. Non-small-cell lung cancer (NSCLC) cell viability after NSCLC cell lines H2009 and H838 with 4EASO or mismatched ASO fi treatment with eIF4E-speci c ASO (4EASO) or mismatched antisense with and without gemcitabine (Figure 4). Treatment of cells with oligonucleotide (ASO). NSCLC cell lines H2009, H838 and H522 were 4EASO and gemcitabine resulted in additional cell death treated with the indicated concentration of mismatched ASO (white bars) or 4EASO (black bars). Viable cells were counted after 72 h and compared with either gemcitabine or 4EASO alone, with a 50% are expressed as cell number normalized to untreated cells. Each and 72% total cell death for H838 and H2009, respectively, experiment was performed in triplicate. Columns, means; bars, s.d. compared with untreated cells. For H838 cells, combination treatment resulted in a 26% and 17% increase in cell death compared with treatment with gemcitabine alone or 4EASO alone, respectively. For H2009 cells, combination treatment resulted in a 31% and 35% increase in cell death compared with treatment with H838 gemcitabine alone or 4EASO alone, respectively. The survival of fi Mismatch ASO 4EASO combined treatment was shown to be signi cantly different than that of each drug alone for H2009 cells (Po0.002). Although the eIF4E survival of combined treatment for H838 was shown to be nearly statistically significant (P = 0.05) for 4EASO treatment alone and 0.78 1 0.053 0.09 Band intensity significantly different for gemcitabine alone (P = 0.048). There were changes observed in the survival of mmASO-treated cells, for VEGF both cell lines, compared with untreated cells, however, these differences were not statistically significant (P40.2). Treatment of 0.95 1 0.62 0.44 Band intensity cells with mismatched ASO had little effect upon H838 cells in combination with gemcitabine. However, mismatched ASO c-Myc in combination with gemcitabine did significantly affect the growth of H2009 cells, although this combination was significantly 0.9 1 0.62 0.65 Band intensity less effective than 4EASO in combination with gemcitabine (Po0.007). -actin 400 0 400 800 Antisense oligonucleotide [nM] DISCUSSION Translation initiation is increasingly recognized as an attractive H2009 target for cancer therapy because cap-mediated translation is Mismatch ASO 4EASO frequently overactive in cancer cells. Extensive work has been done establishing eIF4E as an oncogene; hence, different eIF4E strategies are being employed to target eIF4E either directly 1 0.75 0.33 0.23 0.01 0.04 Band intensity through the use of ASOs27 or through the use of small-molecule inhibitors, which inhibit either the eIF4E–eIF4G interaction19–21 or VEGF 23,25 1 0.9 1 0.75 0.27 0.32 Band intensity the eIF4E to mRNA cap association or target eIF4E indirectly like mTOR inhibitors,31 which repress the downstream phosphor- Osteopontin ylation of 4E-BP1 and make eIF4E unavailable for initiation of 1 0.87 0.71 0.56 0.57 0.65 Band intensity translation. In the present study, we show that eIF4E levels were -actin significantly reduced in a dose-dependent manner in NSCLC cells 02550100 200 400 0 25 50 100 200 400 treated with 4EASO. Our results are consistent with those obtained by Graff et al. who developed this second-generation ASO and Antisense oligonucleotide [nM] studied its effects on human breast and prostate xenograft Figure 3. Reducing eukaryotic translation initiation factor 4E (eIF4E) models.27 In this same report, it was also revealed that eIF4E expression in non-small-cell lung cancer (NSCLC) cells suppresses mRNA level was reduced in one NSCLC cell line (A549) and eIF4E expression of eIF4E-regulated oncoproteins. Cells were treated with protein level was diminished in a second cell line (H460). The fi eIF4E-speci c ASO (4EASO) or its mismatched control for 72 h preliminary achievements exhibited in this earlier work propelled followed by immunoblot analysis. (a) H838 cell lysates were probed our group to investigate more thoroughly, 4EASO as a treatment for expression of eIF4E, vascular endothelial growth factor (VEGF) fi and c-Myc. (b) H2009 cell lysates were probed for expression of for NSCLC. A total of ve different cell lines have now eIF4E, VEGF and osteopontin. The intensity of eIF4E bands were demonstrated sensitivity to 4EASO. Importantly, our present normalized to untreated cells for each cell line and was measured results showed that β-actin expression was unaffected by lowering employing ImageJ. β-Actin served as a loading control. eIF4E levels, supporting the notion that translation of

0.4 0.4 Cell viability Cell viability 0.2 0.2 [normalized to untreated] 0 [normalized to untreated] 0 8 nM GEM - - + - + + 8 nM GEM - - + - + + 35 nM 4EASO - - - + + - 50 nM 4EASO - - - + + - 35 nM mmASO - + - - - + 50 nM mmASO - + - - - + Figure 4. eIF4E-speciﬁc ASO (4EASO) sensitizes non-small-cell lung cancer (NSCLC) cells to gemcitabine (GEM). A cell proliferation assay was performed with NSCLC cell lines H838 and H2009. Cells were transfected with mismatched antisense oligonucleotide (mmASO) or 4EASO and treated 24 h later with the indicated concentration of GEM. Viable cells were counted 72 h after transfection using trypan blue exclusion. Cell viability is expressed as cell number normalized to untreated cells. Combination treatment averages were compared with either agent alone by Student's t-test. A P value ⩽ 0.05 is indicated by *. NS, not statistically signiﬁcant. Data are mean ± s.d. of three experiments.

unstructured, simple mRNAs require only a small amount of eIF4E produced a 32% stable disease rate. It is believed that antitumor and are translated with higher efficiency compared with lengthy, effects with 4EASO would be enhanced in combination with highly structured mRNAs.7 Our work also demonstrates that chemotherapy or radiation.29 The results presented here demon- reduction in translation not only induces cell death in NSCLC cells, strate that 4EASO potently sensitizes NSCLC cells to gemcitabine but also regulates key oncogenic proteins involved in malignancy. and support the notion of enhanced chemosensitivity with 4EASO Furthermore, the present study’s finding that repression of cap- in NSCLC. With this in mind, 4EASO is being examined for the dependent translation results in enhanced efficacy of chemo- treatment of patients with stage IV non-small-cell lung cancer therapeutic agents in NSCLC is consistent with our previous work NSCLC in combination with chemotherapy in phase I/II trials.29 using dominant active 4E-BP1 to suppress eIF4E activity.18,32 In addition, a recent study where 4EASO is assessed as a therapeutic agent for mesothelioma cells revealed similar results as in NSCLC CONFLICT OF INTEREST revealing diminished eIF4E levels, repressed eIF4F complex JR Graff and BW Konicek are employees of Eli Lilly and Company and are also on the formation, reduced oncogenic protein levels, reduced meso- patent for 4EASO (LY2275796). US Pat. No. 7601700 is the patent for LY2275796 with thelioma cell survival along with sensitization of cells to both the title ‘Modulation of eIF4E expression’. This does not alter the authors' adherence gemcitabine and pemetrexed.28 These similar results may indicate to any policies on sharing data and materials. the successful and universal nature of targeting deranged translation in cancer cells. If translation inhibition universally enhances sensitivity to cytotoxic chemotherapy, then this ACKNOWLEDGEMENTS therapeutic approach may be applicable to a variety of tumor We thank Ashley McDonald for her excellent technical assistance. Funding that types and should be empirically tested in other cancer types that supported this work was the John Skoglund Lung Cancer Research Endowment. The demonstrate deranged cap-dependent translation. novel drugs 4EASO (LY2275796) and mismatch ASO are not commercially available Although 4EASO exhibited a significant antigrowth effect and were supplied by Eli Lilly and Company. on NSCLC cells, mismatched ASO also suppressed cell growth, albeit to a much lesser extent. This observation could be due to REFERENCES the non-specific effects of ASOs targeting other mRNAs involved 1 Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer in cell growth and survival. However, mismatched ASO had no statistics, 2012. CA Cancer J Clin 2012; 65:87–108. detectable effect on the levels of eIF4E. A previous study has 2 Silvera D, Formenti SC, Schneider RJ. Translational control in cancer. Nat Rev shown that the mechanism by which 4EASOs triggered cell death Cancer 2010; 10:254–266. 27 was due to apoptosis, again reiterating the fact that increased 3 Sonenberg N, Hinnebusch AG. Regulation of translation initiation in eukaryotes: eIF4E levels make cells resistant to apoptosis and promote tumor mechanisms and biological targets. Cell 2009; 136:731–745. growth. 4 Culjkovic B, Topisirovic I, Skrabanek L, Ruiz-Gutierrez M, Borden KL. eIF4E is a Preclinical studies have evaluated the safety of 4EASO. Treat- central node of an RNA regulon that governs cellular proliferation. J Cell Biol 2006; ment of breast cancer cells with 4EASOs had no significant effect 175:415–426. 5 Koromilas AE, Lazaris-Karatzas A, Sonenberg N. mRNAs containing extensive on global protein synthesis, but reduced levels of translationally fi regulated oncogenes.27 In a mouse xenograft model, 4EASO secondary structure in their 5' non-coding region translate ef ciently in cells overexpressing initiation factor eIF-4E. EMBO J 1992; 11: 4153–4158. treatment reduced the level of eIF4E in both tumor and normal fi 6 Culjkovic B, Topisirovic I, Skrabanek L, Ruiz-Gutierrez M, Borden KL. eIF4E pro- tissue, without causing any signi cant toxicity as measured by motes nuclear export of cyclin D1 mRNAs via an element in the 3'UTR. J Cell Biol changing body weight, organ weight or plasma liver transaminase 2005; 169:245–256. 27 levels. 7 De Benedetti A, Graff JR. eIF-4E expression and its role in malignancies and A phase 1 clinical trial of 4EASO (LY2275796) in advanced metastases. Oncogene 2004; 23: 3189–3199. human cancer was carried out owing to the promising findings 8 Lazaris-Karatzas A, Montine KS, Sonenberg N. Malignant transformation by a from these preclinical investigations and the results recently eukaryotic initiation factor subunit that binds to mRNA 5' cap. Nature 1990; 345: published.29 The maximum tolerated dose and biological effective 544–547. dose of 1000 mg 4EASO was determined employing a dose 9 Ruggero D, Montanaro L, Ma L, Xu W, Londei P, Cordon-Cardo C et al. The translation factor eIF-4E promotes tumor formation and cooperates with c-Myc in escalation experimental design. When comparing pre- and post- lymphomagenesis. Nat Med 2004; 10: 484–486. treatment tumor biopsies, a decrease in eIF4E expression was 10 Graff JR, Konicek BW, Lynch RL, Dumstorf CA, Dowless MS, McNulty AM et al. observed in the majority of patients. Even though eIF4E mRNA eIF4E activation is commonly elevated in advanced human prostate cancers along with nuclear protein expression in patient tumors was and significantly related to reduced patient survival. Cancer Res 2009; 69: reduced, tumor responses were poor. 4EASO, as a single agent, 3866–3873.

Cancer Gene Therapy (2015), 396 – 401 © 2015 Nature America, Inc. 4EASO as a therapy for non-small-cell lung cancer SC Thumma et al 401 11 Coleman LJ, Peter MB, Teall TJ, Brannan RA, Hanby AM, Honarpisheh H et al. 22 Ghosh P, Park C, Peterson MS, Bitterman PB, Polunovsky VA, Wagner CR. Synthesis Combined analysis of eIF4E and 4E-binding protein expression predicts breast and evaluation of potential inhibitors of eIF4E cap binding to 7-methyl GTP. cancer survival and estimates eIF4E activity. Br J Cancer 2009; 100: 1393–1399. Bioorg Med Chem Lett 2005; 15: 2177–2180. 12 Chen CN, Hsieh FJ, Cheng YM, Lee PH, Chang KJ. Expression of eukaryotic 23 Grudzien E, Stepinski J, Jankowska-Anyszka M, Stolarski R, Darzynkiewicz E, initiation factor 4E in gastric adenocarcinoma and its association with clinical Rhoads RE. Novel cap analogs for in vitro synthesis of mRNAs with high trans- outcome. J Surg Oncol 2004; 86:22–27. lational efficiency. RNA 2004; 10: 1479–1487. 13 Rosenwald IB, Chen JJ, Wang S, Savas L, London IM, Pullman J. Upregulation of 24 Ghosh B, Benyumov AO, Ghosh P, Jia Y, Avdulov S, Dahlberg PS et al. Nontoxic protein synthesis initiation factor eIF-4E is an early event during colon carcino- chemical interdiction of the epithelial-to-mesenchymal transition by targeting genesis. Oncogene 1999; 18: 2507–2517. cap-dependent translation. ACS Chem Biol 2009; 4:367–377. 14 Salehi Z, Mashayekhi F, Shahosseini F. Significance of eIF4E expression in skin 25 Jia Y, Chiu TL, Amin EA, Polunovsky V, Bitterman PB, Wagner CR. Design, synthesis squamous cell carcinoma. Cell Biol Int 2007; 31: 1400–1404. and evaluation of analogs of initiation factor 4E (eIF4E) cap-binding antagonist 15 Wang S, Rosenwald IB, Hutzler MJ, Pihan GA, Savas L, Chen JJ et al. Expression of Bn7-GMP. Eur J Med Chem 2010; 45: 1304–1313. the eukaryotic translation initiation factors 4E and 2alpha in non-Hodgkin's 26 Soni A, Akcakanat A, Singh G, Luyimbazi D, Zheng Y, Kim D et al. eIF4E knock- lymphomas. Am J Pathol 1999; 155: 247–255. down decreases breast cancer cell growth without activating Akt signaling. 16 Rosenwald IB, Hutzler MJ, Wang S, Savas L, Fraire AE. Expression of eukaryotic Mol Cancer Ther 2008; 7:1782–1788. translation initiation factors 4E and 2alpha is increased frequently in bronchio- 27 Graff JR, Konicek BW, Vincent TM, Lynch RL, Monteith D, Weir SN et al. Therapeutic loalveolar but not in squamous cell carcinomas of the lung. Cancer 2001; 92: suppression of translation initiation factor eIF4E expression reduces tumor growth 2164–2171. without toxicity. J Clin Invest 2007; 117: 2638–2648. 17 Seki N, Takasu T, Mandai K, Nakata M, Saeki H, Heike Y et al. Expression of 28 Jacobson BA, Thumma SC, Jay-Dixon J, Patel MR, Dubear Kroening K, Kratzke MG eukaryotic initiation factor 4E in atypical adenomatous hyperplasia and adeno- et al. Targeting eukaryotic translation in mesothelioma cells with an eIF4E-specific carcinoma of the human peripheral lung. Clin Cancer Res 2002; 8: 3046–3053. antisense oligonucleotide. PLoS One 2013; 8: e81669. 18 Jacobson BA, Alter MD, Kratzke MG, Frizelle SP, Zhang Y, Peterson MS et al. 29 Hong DS, Kurzrock R, Oh Y, Wheler J, Naing A, Brail L et al. A phase 1 dose Repression of cap-dependent translation attenuates the transformed phenotype escalation, pharmacokinetic, and pharmacodynamic evaluation of eIF-4E anti- in non-small cell lung cancer both in vitro and in vivo. Cancer Res 2006; 66: sense oligonucleotide LY2275796 in patients with advanced cancer. Clin Cancer 4256–4262. Res 2011; 17: 6582–6591. 19 Moerke NJ, Aktas H, Chen H, Cantel S, Reibarkh MY, Fahmy A et al. Small-molecule 30 Graff JR, Zimmer SG. Translational control and metastatic progression: enhanced inhibition of the interaction between the translation initiation factors eIF4E activity of the mRNA cap-binding protein eIF-4E selectively enhances translation and eIF4G. Cell 2007; 128:257–267. of metastasis-related mRNAs. Clin Exp Metastasis 2003; 20:265–273. 20 Ko SY, Guo H, Barengo N, Naora H. Inhibition of ovarian cancer growth by 31 Dowling RJ, Topisirovic I, Fonseca BD, Sonenberg N. Dissecting the role of a tumor-targeting peptide that binds eukaryotic translation initiation factor 4E. mTOR: lessons from mTOR inhibitors. Biochim Biophys Acta 2009; 1804: Clin Cancer Res 2009; 15: 4336–4347. 433–439. 21 Cencic R, Hall DR, Robert F, Du Y, Min J, Li L et al. Reversing chemoresistance by 32 Williams BW, Chang JJ, Chi RM, Marker PH, Frethem CD, Le CT et al. small molecule inhibition of the translation initiation complex eIF4F. Proc Natl Cap-dependent translation blockade and fixed dose-rate gemcitabine: interaction Acad Sci USA 2010; 108:1046–1051. in an in vitro bioreactor system. Cancer Lett 2009; 284:37–46.