Antitumor compounds and tumor diagnosis

11788086 · 2023-10-17

Assignee

Inventors

Cpc classification

International classification

Abstract

The present invention relates to a eukaryotic initiation factor (eIF) modulating compound for use in the treatment of a tumor and in the diagnosis of a cancer. The present invention also relates to a method of diagnosing lung cancer in an individual and to a method of providing a prognosis to an individual suffering from lung cancer. Furthermore, the present invention relates to a method of diagnosing colorectal cancer in an individual, a method of differentiating between colon cancer (CC) and rectum cancer (RC) in an individual, a method of determining whether an individual responds to a therapeutic treatment of colorectal cancer, and to a method of determining the course of colorectal cancer in an individual. Furthermore, the present invention relates to a method of diagnosing a glioma in an individual, a method of grading a glioma in an individual, a method of differentiating between a low-grade glioma and a high-grade glioma in an individual, a method of determining whether an individual responds to a therapeutic treatment of a glioma, and a method of determining the course of a glioma in an individual. In addition, the present invention relates to a kit for conducting the above mentioned methods.

Claims

1. A method for treating lung adenocarcinoma, comprising the step of: administering a eukaryotic initiation factor 6 (eIF6) small interfering RNA (siRNA) to a patient having lung adenocarcinoma, wherein the eIF6 siRNA comprises SEQ ID NO:5 or 6.

2. The method according to claim 1, wherein the eIF6 siRNA comprises SEQ ID NO:5.

3. The method according to claim 1, wherein the eIF6 siRNA comprises SEQ ID NO:6.

4. The method according to claim 1, wherein the eIF6 siRNA consists of SEQ ID NO:5.

5. The method according to claim 1, wherein the eIF6 siRNA consists of SEQ ID NO:6.

6. The method of claim 1, wherein the eIF6 siRNA is administered by intravenous, intramuscular, intrathecal, subcutaneous, transdermal or aerosol administration.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) FIG. 1 shows protein expression levels of eIFs and mTOR pathway members in adenocarcinoma (lung cancer) samples compared to healthy tissue samples. eIF6 and eIF1A are upregulated in adenocarcinoma tissue compared to healthy lung tissue.

(2) FIG. 2 shows protein expression levels of eIFs and mTOR pathway members in squamous cell carcinoma (lung cancer) samples compared to healthy tissue samples. The factors eIF1A, eIF6 and eIF4A are upregulated in squamous cell carcinoma compared to healthy lung tissue.

(3) FIG. 3 shows mRNA expression levels of eIFs in adenocarcinoma and squamous cell carcinoma (lung cancer) samples compared to healthy tissue samples using qRT-PCR. The mRNA levels of eIF6 are upregulated in tumor tissue compared to healthy control. eIF1A is not altered in tumor samples compared to healthy controls.

(4) FIG. 4A shows silencing experiments in lung carcinoma cell line A549, where eIF6 and eIF1aX were knocked down.

(5) FIG. 4B shows reproduced silencing experiment in lung carcinoma cell line A549, where eIF6 and eIF1AX were knocked down. Both experiments inhibited proliferation. Knock down of eIF6 leads to higher apoptosis in A549.

(6) FIG. 5 shows the REMBRANDT gene expression analyses of eIFs in astrocytomas grade II-IV compared to non tumor controls. eIF gene expression represented in bars+SD was analysed for eIF1 (a), eIF1AX (b), eIF2S1 (c), eIF3A (d), eIF3B (e), eIF3C (f), eIF3D (g), eIF3H (h), eIF3I (i), eIF3J (j), eIF3K (k), eIF3M (1), eIF4A2 (m), eIF4B (n), eIF4E (o), eIF4G1 (p), eIF4H (q), eIF5 (r) and eIF6 (s). Numbers: non tumor: n=21, grade II: n=61, grade III: n=47, grade IV: n=191.

(7) FIGS. 6A-6D shows the correlation of eIF gene expression with survival probabilities of GBM patients using the TCGA database. Statistical decision trees were generated for eIF gene expression without any exclusion criteria (FIG. 6A) and with the exclusion of TMZ treated patients (FIG. 6C). For each decision tree, survival curves for all patients (FIG. 6B) und and patients without TMZ treatment (FIG. 6D) were prepared. Numbers: Total patients: n=535, patients without TMZ: n=230.

(8) FIGS. 7A-7L shows mRNA expression of eIFs in gliomas compared to non tumor controls (Ctrl) using qRT-PCR. eIF mRNA expression is represented in x-fold change to the non tumor control for eIF3A (FIG. 7A), eIF3B (FIG. 7B), eIF3C (FIG. 7C), eIF3I (FIG. 7D), eIF3M (FIG. 7E), eIF4A1 (FIG. 7F), eIF4E (FIG. 7G), eIF4EBP1 (FIG. 7H), eIF4G1 (FIG. 7I), eIF4H (FIG. 7J), eIF5 (FIG. 7K) and eIF6 (FIG. 7L). As housekeeping genes GAPDH and SDHA were used. Bars represent group means+SEM. Statistical analyses: One-way ANOVA followed by DUNN's or Bonferroni post-hoc test. Significance levels: *p<0.05; **p<0.01, ***p<0.001. Numbers: Control (Ctrl): n=7, grade I: n=7, grade II: n=9, grade III: n=6, grade IV: n=14.

(9) FIG. 8 shows eIF protein expression in astrocytomas (grade I-IV) compared to healthy brain tissue (Ctrl) using immunoblot analyses. eIF subunits up-regulated in tumor tissue compared to healthy control tissue are underlined, down-regulated ones are highlighted in grey. Actin was used as loading control. Numbers: Control (Ctrl): n=3, grade I: n=2, grade II-IV: n=3.

(10) FIGS. 9A-9E shows the immunohistochemical evaluation of eIF protein expression in astrocytomas (grade I-IV) compared to healthy control tissue. Graphs show the TIS of eIF3C (FIG. 9A), eIF4G (FIG. 9B), eIF4H (FIG. 9C), eIF5 (FIG. 9D) and eIF6 (FIG. 9E). Bars represent group means+SEM. Statistical analyses: One-way ANOVA followed by DUNN's post-hoc test. Significance levels: *p<0.05; **p<0.01, ***p<0.001. Numbers: Control: n=11, grade I: n=19, grade II: n=24, grade III: n=21, grade IV: n=13.

(11) FIG. 10 shows eIF protein expression after chemosensitivity testings in 2 different murine GBM xenografts models (Patient X1 and X2) using immunoblot analyses. Chemosensitivity testings were performed with the following cytostatics: Everolimus (B), Sorafenib (C), Bevacizumab (Avastin) (D), Irinotecan (E), Salinomycin (F) and Temozolomide (G). As control treatment for murine xenografts phosphate buffered saline (PBS, A) was used. Actin was used as loading control. Number: n=2/group.

(12) FIG. 11 shows eIF protein expression after chemosensitivity testings in 2 different murine GBM xenografts models (Patient X3 and X4) using immunoblot analyses. Chemosensitivity testings were performed with the following cytostatics: Temozolomide (B), Thalidomide (C), Everolimus (E), Regorafenib (G), Sorafenib (I), Bevacizumab (Avastin)(J) and Irinotecan (K). Temozolomide treatment was combined with Thalidomide (D), Everolimus (F) and Regorafenib (H). As control treatment for murine xenografts phosphate buffered saline (PBS, A) was used. Actin was used as loading control. Number: n=1-2/group.

(13) FIG. 12 shows eIF protein expression after Temozolomide treatment in the U-87 MG glioma cell line using immunoblot analyses. U-87 MG cells were treated with Temozolomide for 1 (d1), 3 (d3) and 5 days (d5). 3 distinct concentrations were used for the treatment: 20, 50 and 100 μM. As control, cells without treatment (d0) and cells treated with DMSO were taken. GAPDH was used as loading control.

(14) FIG. 13 shows the expression/survival correlation for the 1.sup.st quartile cut off level between high and low eIF expression, in particular of eIF2C 3, eIF3g, eIF-4G1 and eIF-5, in DLBCL. Analyzed was the correlation between the expression of the mRNA coding for the respective indicated protein and patient outcome.

(15) FIG. 14 shows the expression/survival correlation for the Median cut off level between high and low eIF expression, in particular of eIF2AK3/HsPEK, eIF2B4/eIF-2B subunit delta, eIF3c and 4E-BP1, in DLBCL. Analyzed was the correlation between the expression of the mRNA coding for the respective indicated protein and patient outcome.

(16) FIG. 15 shows the expression/survival correlation for the 3.sup.rd quartile cut off level between high and low eIF expression, in particular of eIF2AK4, eIF2d, eIF-2a and eIF-2-beta/eIF2S2, in DLBCL. Analyzed was the correlation between the expression of the mRNA coding for the respective indicated protein and patient outcome.

(17) FIG. 16 shows the expression/survival correlation for the 3.sup.rd quartile cut off level between high and low eIF expression, in particular of eIF3b, eIF3d, eIF3f and eIF3l, in DLBCL (continued). Analyzed was the correlation between the expression of the mRNA coding for the respective indicated protein and patient outcome.

(18) FIG. 17 shows the expression/survival correlation for the 3.sup.rd quartile cut off level between high and low eIF expression, in particular of eIF-4B, eIF-4E3 and eIF-5A, in DLBCL (continued). Analyzed was the correlation between the expression of the mRNA coding for the respective indicated protein and patient outcome.

(19) FIG. 18 shows the eIF expression comparison between an immortalized B-cell line (MUG-CC1-LCL) and eight lymphoma cell lines (Ri-1, U-2932, NU-DUL-1, SU-DHL-4, KARPAS 422, OCI-LY1, Raji and BL2). Actin expression was used as a loading control.

(20) FIG. 19 shows Kaplan Meier Curves, in particular of HCC patients. Survival curve according to t-stage; 14 patients with a score of 3, 135 patients with score 2 and 85 patients with a score of 1. The survival is better with a lower score compared to a score of 3.

(21) FIGS. 20A-20B shows Kaplan Meier Curves, in particular of HCC patients, to various eIFs. [FIG. 20A] Association of eIF2α expression with survival time in HCC. The eIF3h expression in HCC shows in patients with high score of eIF2α. [FIG. 20B] The eIF3h expression shows in patients with high score of eIF3h.

(22) FIG. 21 shows Kaplan Meier Curves, in particular of HCC patients, to various eIFs. FIG. 21A shows the overall survival of patients with a score of 2 or 3 for eIF5 is better than with a score below 2. FIG. 21B Overall survival of patients with a high score of eIF6 is better than patients with a sore below 2 for eIF6.

(23) FIGS. 22A-22B shows Kaplan Meier Curves, in particular of HCC patients, to various eIFs. [FIG. 22A] Association of eIF3p (eIF3c) expression with overall survival time in HCC patients. The eIF3p expression in HCC shows in 37 patients with high score of eIF3p, 37 patients represents a score below 3, 72 patients shows a score of 1 and no staining intensity for 88 patients. FIG. 22B Overall survival for eIF4e in patients with HCC. 19 patients with a high eIF4e expression level, 11 patients with a sore of 2, and 81 patients with a score of 1.

(24) FIGS. 23A-23C shows expression of eIF3 subunits in CRC. [A and B] Significant increase of protein level for eIF3A, eIF3B, eIF3B, eIF3D and eIF3M in CRC samples compared to normal mucosa. Protein expression of eIF3C, eIF3j and eIF3K is significantly upregulated in RC compared to CC and healthy tissue. [C] mRNA expression of eIF3A, eIF3B and eIF3j show an overexpression in RC compared to CC samples. eIF3H and eIF3M overexpressed in CC compared to RC. mRNA expression of eIF3C and eIF3j show an overexpression in RC compared to CC samples. eIF3C show no significant differences on mRNA level in CRC. In particular, FIG. 23A shows IHC of eIF3 subunits in CRC, FIG. 23B shows protein expression of eIF3 subunits in CRC, and FIG. 23C shows mRNA expression of eIF3 subunits in CRC.

(25) FIGS. 24A-24C shows expression of eIF4 subunits in CRC. [FIGS. 24A and 24B] Significant upregulation on protein level for peIF4B, eIF4B and eIF4G in CRC samples compared to normal mucosa. eIF4E shows high expression in RC samples compared to normal mucosa. [FIG. 24C] Upregulation of eIF4B on mRNA level in CRC samples. Significant upregulation of eIF4G in RC samples. No changes of eIF4E on mRNA level. In particular, FIG. 24A shows IHC of eIF4 subunits in CRC, FIG. 24B shows protein expression of eIF4 subunits in CRC, and FIG. 24C shows mRNA expression of eIF4 subunits in CRC.

(26) FIGS. 25A-25C shows expression of peIF2α, eIF2α, eIF5 and eIF6 in CRC. [FIGS. 25A and 25B] Significant upregulation on protein level for peIF2α and eIF2α in CC samples compared to RC and normal mucosa. Increase of eIF5 and eIF6 on protein level in CRC samples compared to normal mucosa. [FIG. 25C] Upregulation of eIF2α and eIF5 on mRNA level in RC samples. Significant increase of eIF6 on mRNA in CRC compared to normal mucosa. In particular, FIG. 25A shows IHC of, eIF2α and eIF6 in CRC, FIG. 25B shows Protein expression of p eIF2α, eIF2α, eIF5 and eIF6 in CRC, and FIG. 25C shows mRNA expression of eIF2α, eIF5 and eIF6 in CRC.

(27) FIG. 26 to 28 show the eIF1, eIF5 and eIF6 silencing and FIG. 29 shows polysomal profiles of eIF1, eIF5 and eIF6. [FIG. 26] Significant reduction of the protein expression after eIF1 silencing in HCT116 cells. Significant reduction of the mRNA expression after eIF1 silencing in HCT116 cells. Significant reduction of the cell viability after 24h, 48h and 72h of eIF1 silencing in HCT116 cells. Significant reduction of apoptosis after 24h, 48h and 72h of eIF1 silencing in HCT116 cells. [FIG. 27] Significant reduction of the protein expression after eIF5 silencing in HCT116 cells. Significant reduction of the mRNA expression after eIF5 silencing in HCT116 cells. Significant reduction of the cell viability after 24h, 48h and 72h of eIF5 silencing in HCT116 cells. Significant reduction of apoptosis after 24h, 48h and 72h of eIF5 silencing in HCT116 cells. FIG. 28 Significant reduction of the protein expression after eIF6 silencing in HCT116 cells. Significant reduction of the mRNA expression after eIF6 silencing in HCT116 cells. Significant reduction of the cell viability after 24h, 48h and 72h of eIF6 silencing in HCT116 cells. Significant reduction of apoptosis after 24h, 48h and 72h of eIF6 silencing in HCT116 cells. [FIGS. 29A-29C] Polysome associated fraction analysis of eIF1, eIF5 and eIF6. Polysomal profiles of HCT116 cells transfected with eIF1 (Si eIF1) and control (MOK). Increased formation of functional 60S and reduced 80S ribosomes recorded after knockdown of eIF1. Polysomal profiles of HCT116 cells transfected with eIF5 (Si eIF5) and control (MOK). Increased formation of functional 40S and 60S ribosomes recorded after knockdown of eIF5. Polysomal profiles of HCT116 cells transfected with eIF6 (Si eIF6) and control (MOK). Increased formation of functional 40S and 60S and reduced 80S ribosomes recorded after knockdown of eIF6.

(28) FIGS. 30A and 30B shows colony assays of eIF1 and eIF5. In particular, FIG. 30A shows the effect of eIF1 knockdown on invasiveness and clonogenicity in HCT 116 cells, and FIG. 30B shows the effect of eIF5 knockdown on invasiveness and clonogenicity in HCT 116 cells. Clonogenicity is dramatically reduced after eIF1 knockdown in HCT 116 cells compared to the scrambled control cell. Clonogenicity is dramatically reduced at all three time points after eIF5 knockdown in HCT 116 cells compared to the scrambled control cell.

(29) FIG. 31 shows colony assays of eIF6. In particular, it shows the effect of eIF6 knockdown on invasiveness and clonogenicity in HCT 116 cells. Clonogenicity is dramatically reduced at all three time points after eIF6 knockdown in HCT 116 cells compared to the scrambled control cell. The decrease in transmigrating cells after transfection with eIF6 siRNA constructs is significant in HCT116 cells. Statistical analyses: 1-way ANOVA with Bonferroni post-test *p<0.05, **p<0.01 and ***p<0.001.

(30) FIG. 32 shows eIF1, eIF5 and eIF6 expression levels in low and high grade colon and rectum carcinomas. [A] Western blot of eIF1, eIF5 and eIF6 from low grade (LG) and high grade (HG) colon carcinomas (CC) and rectum carcinomas (RC) compared to non-neoplastic tissues (NNT). Equal amounts of protein from each pair are resolved on SDS PAGE and immunoblotted with antibodies directed against eIF1, eIF5, eIF6 and R-actin (loading control). [B] qRT-PCR of eIF1, eIF5 and eIF6 from LG and HG CC and RC compared to non-neoplastic tissues (NNT). Error bars show SEM.

(31) FIG. 33 shows single eIFs and a set of eIFs which level is preferably determined in a method of diagnosing a lung cancer in an individual.

(32) FIG. 34 shows single eIFs and sets of eIFs which level is preferably determined in a method of diagnosing a glioma in an individual. These sets comprise 3, 4, 5, or 6 eIFs.

(33) FIG. 35 shows Affimetrix gene expression analysis of eIF6 in NSCLC, ADC and SQC in correlation to patient overall survival. Survival is significantly influenced by eIF6 expression all three analysis. For NSCLC with a p-value of 1.5*10.sup.−13, ADC with a p-value of 3.8*10.sup.−13 and SQC with a p-value of 0.14.

(34) FIGS. 36A-36F shows the impact of eIF gene expression on patients overall survival in LGG and GBM using data from the TCGA database. Kaplan-Meier survival curves were calculated based on eIF3I (FIGS. 36A, 36D) eIF4G (FIGS. 36B, 36E) and eIF4H (FIGS. 36C, 36F) gene expression for patients with LGG (FIGS. 36A-36C) and GBM (FIGS. 36D-36F). Numbers: LGG: n=389, GBM: n=123.

(35) FIGS. 37A-37R shows eIF protein expression in murine xenograft models after chemosensitivity testings. The effect of Everolimus (FIG. 37B), Sorafenib (FIG. 37C), Bevacizumab (FIG. 37D), Irinotecan (FIG. 37E), Salinomycin (FIG. 37F) and Temozolomide (FIG. 37G) on eIF protein expression was analyzed in comparison to the PBS control group (FIG. 37A) using immunoblot analyses. Six different murine xenograft models were investigated (X5, X6, X7, X8, X9, X11). In all xenograft models except for X5, Temozolomide drastically reduced tumor growth. Densitometric analyses of immunoblots were performed using ImageJ software (NIH, MD, United States). For relative densities, expression of eIF1A (FIG. 37a), p-eIF2α (FIG. 37b), eIF2α (FIG. 37c), eIF3A (FIG. 37d), eIF3B (FIG. 37e), eIF3C (FIG. 37f), eIF3D (FIG. 37g), eIF3H (FIG. 37h), eIF3I (FIG. 37i), eIF3J (FIG. 37j), eIF3K (FIG. 37k), eIF3M (FIG. 37l), eIF4A (FIG. 37m), p-eIF4B (FIG. 37n), eIF4E (FIG. 37o), eIF4G (FIG. 37p), eIF4H (FIG. 37q) und eIF6 (FIG. 37r) was normalized to the loading control (Actin). Scatter dot blot+SEM. Statistical analysis: 1-way ANOVA with Bonferroni posttest (*p<0.05; **p<0.01).

(36) FIGS. 38A-38O shows eIF protein expression in murine xenograft models after chemosensitivity testings using either monotherapies or combination therapies. The effect of Temozolomide (FIG. 38B), Thalidomide (FIG. 38C), Temozolomide/Thalidomide (FIG. 38D), Everolimus (FIG. 38E), Temozolomide/Everolimus (FIG. 38F), Regorafenib (FIG. 38G) und Temozolomide/Regorafenib (FIG. 38H) on eIF protein expression was analyzed in comparison to the PBS control group (FIG. 38A) using immunoblot analyses. Two different murine xenograft models were investigated (X3, X4). Densitometric analyses of immunoblots were performed using ImageJ software (NIH, MD, United States). For relative densities, expression of p-eIF2α (FIG. 38a), eIF2α (FIG. 38b), eIF3A (FIG. 38c), eIF3B (FIG. 38d), eIF3H (FIG. 38e), eIF3I (FIG. 38g), eIF3J (FIG. 38h), eIF3K (FIG. 38i), eIF3M (FIG. 38j), eIF4A (FIG. 38k), p-eIF4B (FIG. 38l), eIF4E (FIG. 38m), eIF4G (FIG. 38n) und eIF4H (FIG. 38o) was normalized to the loading control (Actin). Scatter dot blot+SEM. Statistical analysis: 1-way ANOVA with Bonferroni posttest (*p<0.05; **p<0.01)

EXAMPLES

(37) Materials & Methods

(38) Immunoblot Analyses

(39) Protein expression was analyzed in NP-40 tissue lysates. Therefore normalized protein amounts were loaded onto 8% or 12.5% polyacrylamide gels in consideration of the molecular mass of the protein of interest. Electrophoresis was performed for 1.5 h at 120 V in SDS Running Buffer with the Mini-vertical electrophoresis unit (Amersham Biosciences). Then separated proteins were transferred to PVDF-membranes (Immobilin-P Transfer Membran; Millipore) using a Semi Dry Blotting Unit (JH BioInnovations) at 160 mA for 1.5 hours. Membranes were blocked with 5% non-fat dried milk (AppliChem) in Tris-buffered saline-Tween (TBS-T; 0.2 M Tris, 1.4 M NaCl, pH 7.4, 0.1% Tween) for 1 h, followed by a primary antibody incubation (overnight at 4° C.) and an 1 h secondary antibody incubation (1:3000 dilution, GE Healthcare). All primary antibodies were diluted in 5% bovine serum albumin diluted in TBS-T. Between incubation steps membranes were washed 3 times in TBS-T for 10 min. Membranes were then developed with an enhanced chemiluminescence (ECL) Western blotting system (GE healthcare) using the ChemiImager™ System (Alpha Innotech). Immunoblots were evaluated semi-quantitative using ImageJ (Schneider C A, et al. Nat Meth. 9(2012):671-5) software for densitometric analyses. Relative densities were then calculated by normalizing density values for each protein to the loading control (Actin).

(40) TABLE-US-00001 TABLE 1 Primary antibodies used for immunoblot analyses Primary Antibody Manufacturer Dilution eIF1A Abcam  1:10000 eIF2α (D7D3) XP Cell Signaling 1:1000 Phospho-eIF2α (Ser51)(D9G8) Cell Signaling 1:1000 eIF3A Cell Signaling 1:1000 eIF3B (=eIF3η (D-9)) Santa Cruz 1:1000 eIF3C Cell Signaling 1:1000 eIF3D (=eIF3ζ (H-300)) Santa Cruz 1:1000 eIF3H (D9C1) XP Cell Signaling 1:1000 eIF3J Cell Signaling 1:1000 eIF3K (2313C2a) Santa Cruz 1:1000 eIF3M Santa Cruz 1:1000 eIF3Q (H-300) Santa Cruz 1:200  eIF3θ (H-300) Santa Cruz 1:1000 eIF4A1 Cell Signaling 1:1000 eIF4B Cell Signaling 1:1000 Phospho-eIF4B Cell Signaling 1:1000 eIF4E Cell Signaling 1:1000 eIF4G Cell Signaling 1:1000 eIF4H Cell Signaling 1:1000 eIF5 Gene Tex 1:1000 eIF6 Cell Signaling 1:1000

(41) RNA Isolation and qRT-PCR

(42) Total RNA was isolated from deep-frozen brain tissue using Trizol reagent (Life Technologies). Tissue pieces were homogenized in 1 ml Trizol for 30 seconds at 6500 rpm with the MagNA Lyser (Roche). The lysate was incubated for 10 minutes at RT. Next, 200 μl chloroform were added, mixed, incubated for 3 minutes at RT and centrifuged at 10 000 rpm for 15 minutes at 4° C. The upper phase containing the RNA was carefully transferred into a fresh tube, mixed with 500 μl isopropanol and again centrifuged at 10 000 rpm for 20 minutes at 4° C. The supernatant was discarded and the pellet washed with 1 ml 75% ethanol. The pellet was then dried at 37° C. to completely remove the ethanol and then dissolved in 100-200 μl DEPC treated water at 58° C. RNA concentration and quality were determined with the NanoDrop 1000 Spectrophotometer (PeqLab) and subsequently 1 μg RNA is transcribed from total RNA with the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems) according to the manufacturer's instructions.

(43) The obtained cDNA was used for quantitative real time RT-PCR (qRT-PCR) using the 7900HT Fast Real-Time PCR System cycler (Applied Biosystems). Oligonucleotides were self-designed using Primer-BLAST software (10) and synthesized by Eurofins Genomics. Efficiencies of all self-designed primers were calculated with known cDNA concentrations. For the qRT-PCR reaction the Power SYBR® Green PCR Master Mix (Life technologies) is used. Parameters for the qRT-PCR program are set according to the manuals. Threshold cycles (C(t)) were automatically calculated by the 7900HT Fast Real-Time PCR System software (Applied Biosystems). Relative RNA expression was evaluated using the ΔΔC(t))-method.

(44) Tissue Microarrays (TMA)

(45) All tumor tissue samples were acquired at the time of surgery and immediately frozen in liquid nitrogen and stored at −80° C.

(46) Every sample was stained for haematoxylin-eosin and examined to determine relevant tumor areas which were marked on the slide. Tissue cones of the chosen tumor regions were punched out, assembled in an array structure and embedded into a fresh paraffin block, according to a specific pattern. The sections taken were 5 μm thick, mounted on a specific adhesive-coated glass slide, compatible for immunohistochemical staining and analysis.

(47) Immunohistochemistry (IHC)

(48) A summary of all used antibodies and the dilution to determine the expression of different eIFs is shown in Table 2. Staining was performed using the Ventana Immunostainer XT (Ventana Medical Systems, USA). The endogenous peroxidase activity was inactivated in 3% hydrogen peroxide for 5 minutes. The primary antibodies were applied at different dilutions (Table 1) for 60 minutes, followed by incubation with a peroxidase-labelled secondary antibody for 30 minutes and substrate-chromogen 3.3′-diaminobenzidine tetrahydrochloride for 8 minutes. Counterstaining was performed with haematoxylin.

(49) The intensity of IHC staining was evaluated by light microscopy. Density and intensity of each TMA spot was scored in a semi-quantitative manner by differentiating nuclear and cytoplasmic staining. The Total Immunostaining Score (TIS) was calculated in percent. No staining was termed as 0, weak staining as 1, moderate staining as 2 and strong staining as 3.

(50) TABLE-US-00002 TABLE 2 Primary antibodies used for immunohistochemistry Primary Antibody Company Dilution Second Antibody Phospho-mTOR Cell Signalling (#5536) 1:1000 Rabbit mTOR Cell Signalling (#2983) 1:1000 Rabbit Phospho-PTEN Cell Signalling (#9551) 1:1000 Rabbit PTEN Cell Signalling (#9559) 1:1000 Rabbit Phospho-P70S6K Cell Signalling (#9204) 1:1000 Rabbit P70S6K Cell Signalling (#9202) 1:1000 Rabbit Phospho Akt Cell Signalling (#4058) 1:1000 Rabbit Akt Cell Signalling (#9272) 1:1000 Rabbit GAPDH Cell Signalling (#2118) 1:3000 Rabbit Phospho 4E-BP1 Cell Signalling (#9456) 1:1000 Rabbit 4E-BP1 Cell Signalling (#9452) 1:1000 Rabbit Anti-Actin Sigma (A2103) 1:1000 Rabbit eIF1 Sigma (HPA043003) 1:500  Rabbit Phospho-eIF2α (Ser51)(D9G8) Cell Signalling (#3398) 1:1000 Rabid eIF2α (D7D3) XP Cell Signalling (#5324) 1:1000 Rabbit eIF3A Cell Signalling (#2538) 1:1000 Rabbit eIF3β(A-8) = eIF3I Santa Cruz (sc-374155) 1:1000 Mouse eIF3C Cell Signalling (#2068) 1:1000 Rabbit eIF3H (D9C1)XP Cell Signalling (#3413) 1:1000 Rabbit eIF3J Cell Signalling (#3261) 1:1000 Rabbit eIF3K (2313C2a) Santa Cruz (sc-81262) 1:1000 Mouse eIF3M (V-21) Santa Cruz (sc-133541) 1:500  Rabbit eIF3B = eIF3ηD-9 Santa Cruz (sc-137215) 1:1000 Mouse eIF3P110 (B-6) Santa Cruz (sc-74507) 1:500  Mouse eIF3θ (H-300) Santa Cruz (sc-30149) 1:1000 Rabbit eIP3ζ (H-300) = eIF3D Santa Cruz (sc-28856) 1:1000 Rabbit Phospho eIF4b (Ser406) Cell Signalling (#5399) 1:1000 Rabbit eIF4B Cell Signalling (#3592) 1:1000 Rabbit eIF4E Cell Signalling (#9742) 1:1000 Rabbit eIF4G Cell Signalling (#2498) 1:1000 Rabbit eIF5 GeneTex (GTX114923) 1:500  Rabbit eIF6 GeneTex (GTX63642) 1:1000 Rabbit

(51) Generation of Xenograft Models

(52) Samples of patients suffering from colon cancer or GBMs were transplanted into 3 to 6 immunodeficient NOD/SCID mice. The tumor growth was monitored in a daily rhythm. At a size of about 1 cm.sup.3, the tumors were removed and transferred to naive NMRI:nu/nu mice for chemosensitivity testing. Xenotransplanted carcinomas and metastases were treated with different standard and novel chemotherapeutic drugs. During chemosensitivity testing the tumor volume was measured regularly and used to generate growth curves. After a time period of 30-40 days the tumors were excised and analyzed by Immunoblot and Real-Time-PCR. Chemosensitivity data were kindly provided by EPO Berlin-Buch GmbH. Tumor volume of treatment in comparison to control (T/C) was calculated in percent.

(53) Cell Culture

(54) siRNA transfection in: We targeted the gene of interest by using small interfering RNAs (siRNAs) from QIAGEN (Hilden, Germany). For each gene of interest, two target sequences were used. For eIF1; 5′-GACCAGACATATCCTAGCTAA-3′ (SEQ ID NO: 1) and 5′-AAGCAATACCGTCATGTTTCA-3′ (SEQ ID NO: 2), for eIF5; 5′-AGGCGCTTAATCGGCCTCCAA-3′ (SEQ ID NO: 3) and 5′-CAGCCAGAAGTGCAACATGTA-3′ (SEQ ID NO: 4); for eIF6; 5′-CTGCTTTGCCAAGCTCACCAA-3′ (SEQ ID NO: 5) and 5′-CTGGTGCATCCCAAGACTTCA-3′ (SEQ ID NO: 6). For eIF3I: 5′-ATAAATTGGTTTGGTAATAAA-3′ (SEQ ID NO: 7) and 5′-AAGGACCCTATCGTCAATGTA-3′(SEQ ID NO: 8). For eIF1AX; 5′-CCGAGACTACCAGGATAACAA-3′ (SEQ ID NO: 9) and 5′-ATCAATGAAACTGATACATTT-3′ (SEQ ID NO: 10).

(55) Transfection experiments for: Were performed using MetafecteneRsi+transfection reagent (Biontex, Munich, Germany) according to manufacturer's instructions. For the transfection, 1×SI buffer, Metafectene SI+ and siRNA were mixed into a drop. After an incubation of 15 min at room temperature 500 μl cells (80 000 cells/well) were seeded onto a 24-well plate. Cells with transfection mix were cultured at 37° C. in a humidified atmosphere of 5% CO.sub.2. The cells were collected after incubation for 24h, 48h and 72h.

(56) Proliferation Assay for: Transfected cells and MOK were seeded in 96-well plates (80 000 cells/well) and cultivated under low serum conditions (1% FCS) for 24h, 48h and 72h. Viable cell number was determined on the basis of mitochondrial conversion of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma Aldrich, Missouri, USA) to formazine. Cells were incubated with MTT for 2h at 37° C., the medium supernatant was removed and the cells were lysed with sodium dodecyl sulfate for 15 min at room temperature. The MTT formazan crystals were resolved with isopropanol/HCl under shaking for 15 min at room temperature. Optical density was measured at 570 nm (SynergyTM4, BioTek, Winooski, USA). Each assay was executed in six-fold determination and three independent experiments were performed.

(57) Apoptosis for: Apoptotic cells were detected using YO-PRO®-1 (Thermo Fisher Scientific, Massachusetts, USA) reagent. siRNA-transfected and control cells were seeded onto 96-well plates (80 000 cells/well). After 24h, 48h and 72h, cells were incubated with YO-PRO®-1 for 15 min at 37° C., the supernatant was removed, cells were washed with PBS and then measured 485 nm to 535 nm. Each assay was performed in six-fold determination and three independent experiments were carried out.

(58) Invasion Assay for: For analysis of invasiveness of CRC cells, the CytoSelect™ 24-Well Cell Invasion Assay (Cell Biolabs, USA) was used according to the manufacturer's instructions. 1×10.sup.5 siRNA transfected cells and control cells were suspended in medium with 10% FCS, placed in the upper chamber and incubated for 48h at 37° C. The cells that had invaded to the lower surface of the filter inserts were stained with crystal violet. The optical density was measured at 560 nm (SynergyTM4, BioTek, USA).

(59) Colony forming assay for: cells transfected with siRNA and scrambled siRNA as control were collected and seeded in six-well plates at a density of 500 cells/well. The medium was changed every three days. After two weeks of culture, cells were washed three times with PBS and fixed in 4% paraformaldehyde (Sigma-Aldrich, Missouri, USA). Fixed cells were stained by adding freshly prepared diluted Giemsa solution (Sigma-Aldrich, USA) for 20 min. Then the cells were rinsed with distilled water and colonies were analysed using a microscope (Nikon™S—Inverted Microscope, Japan).

(60) Sucrose-gradient fractionation, polysome associated fraction analysis: Sucrose density-gradient centrifugation was used to analyzed the cellular distribution of polysomes, 80S ribosomes and free 40S and 60S subunits. Cells were cultured in 100 mm dishes and transfected with siRNA and control for 24h, 48h and 72h. 15 minutes prior to lysis, cells were incubated with 100 μg/ml cycloheximide (Sigma-Aldrich, Missouri, USA) to stall ribosomes on the mRNA strand. Lysis was performed on ice by washing cells in ice-cold PBS containing 100 μg/ml cycloheximide followed by suspension in lysis buffer (20 mM HEPES pH 7.4, 15 mM MgCl.sub.2, 200 mM KCl, 1% Triton X-100, 2 mM DTT and 100 μg/ml cycloheximide), and nuclei were removed by centrifugation (14000g, 10 min, 4° C.). The supernatant was layered onto 15%-40% sucrose gradients (50 mM NH4Cl, 50 mM Tris-acetate pH 7.0, 12 mM MgCl.sub.2, 100 μg/ml cycloheximide and freshly added 1 mM DTT) and centrifuged in a SW41Ti rotor (Beckman, Villepinte, France) for 150 min at 160000g, 4° C. without breaking. Polysomal profiles were analysed via an ISCO density gradient analyser unit, which analyses and simultaneously blots ribosomal distribution measured by an UA6 detector with 254 nm filter (Teledyne ISCO, Nebraska, USA).

(61) All fractions were TCA (Trichloroacetic acid) precipitated over night at −20° C. to concentrate proteins for gel electrophoresis.

(62) Statistical Analysis

(63) All values are represented as means±standard error of the mean if not indicated otherwise. Statistical significance was evaluated using one-way ANOVA followed by the Bonferroni's multiple comparison tests. All calculations were performed using GraphPad Prism software (La Jolla, USA).

Example 1: Lung Cancer

(64) Non Small Cell Lung Cancer (NSCLC) is among the most frequently diagnosed cancer entities and is the leading cause of cancer related death worldwide. In recent years, protein synthesis has turned out to be tightly linked to cancer. Translating mRNA to the corresponding protein is one of the major activities in each cell and is a strictly regulated process. Crucial for this translation process are eukaryotic initiation factors (eIFs), which are themselves regulated by the mammalian target of Rapamycin (mTOR)-pathway. Dysregulation of translation initiation may lead to alterations in protein synthesis resulting in uncontrolled cell growth and cancer formation. Thus, eIFs represent crossroads for carcinogenesis.

(65) Methods

(66) Paired NSCLC and Non Neoplastic Lung Tissue (NNLT) from 28 individuals were studied on protein expression level for eIFs and mTOR pathway members by Immunoblot analyses. qRT-PCR As housekeeping gene (HKG) for adenocarcinoma a combination of importin 8(IP08) and succinate dehydrogenase A (SDHA) was used. For squamous cell carcinoma succinate dehydrogenase A (SDHA) and beta actin was used. Therefore the mean of the C(t) values of both HKGs was calculated and used for the ΔΔC(t))-calculation.

(67) Transfection of A549

(68) Carcinoma cell line A549 was transfected with Oligofectamin™ tranfection reagent form Invitrogen. Cells were transfected according to manufacturer's instructions. Cells were serum straved 24h before seeding in 12 well plates. We seeded 20.000 cell per well and transfected them after 24h. Cells with tranfection solution were cultured at 37° C. in a humidified atmosphere of 5% CO.sub.2. The cells were collected after incubation for 72h and 120h.

(69) For A549 we used small interfering RNAs (siRNAs) from QIAGEN (Hilden, Germany). For each gene of interest, two target sequences were used. For eIF6; 5′-CTGCTTTGCCAAGCTCACCAA-3′(SEQ ID NO: 5) and 5′-CTGGTGCATCCCAAGACTTCA-3′ (SEQ ID NO: 6). For eIF1AX; 5′-CCGAGACTACCAGGATAACAA-3′ (SEQ ID NO: 9) and 5′-ATCAATGAAACTGATACATTT-3′ (SEQ ID NO: 10).

(70) Proliferation Measurements

(71) A549 viable cells were counted at 72h and 120h with Guava ViaCount Reagent for Flow Cytometry from Merck Millipore according to manufacturer's instructions.

(72) Apoptosis Measurements

(73) Apoptosis in eIF silenced A549 cells were measured with Annexin V: FITC Apoptosis Detection Kit I from Becton Dickinson Austria GmbH according to manufacturer's instructions, at 72h and 120h.

(74) Affimetrix Data Analysis

(75) The gene expression data of eIF6 in NSCLC, ADC and SQC were downloaded from RMA (Robust Multi-Array Average) normalized counts from Affymetrix platform for NSCLC on 20 Feb. 2017. The number of NSCLC was 1926 patients, for ADC 720 patient and for SQC 524 patients. The patient survival data was also obtained from Affimetrix data set and analyzed with log-rank test. To analyze the association between gene expression (stratified by median) and survival gene expression values were used in all cancer types.

(76) Immunohistochemistry of eIF6

(77) Immunohistochemistry on an adenocarcinoma (ADC) and squamous cell carcinoma (SQCC) tissue microarray (TMA) was performed. Following antibody was used in a 1:100 dilution, eIF6 Antibody, Bethyl, A303-030A.

(78) Results

(79) Adenocarcinoma

(80) In FIG. 1 a basic characterization of eIFs and mTOR pathway members in 9 adenocarcinoma patients is shown. The factors eIF1A, eIF6 and eIF4A are upregulated compared to healthy lung tissue. This finding is not described in literature. In FIG. 3 mRNA levels of eIF6 and eIF1A was investigated in adenocarcinoma (ADC) patients. The mRNA levels of eIF6 are upregulated in tumor tissue compared to healthy control. eIF1A is not altered in tumor samples compared to healthy controls.

(81) Immunohistochemistry data from an ADC TMA show that eIF6 is significantly (p>0,001) upregulated in ADC patient tissue (n=307) compared to healthy parenchymal tissue (n=156) (data not shown).

(82) Squamous Cell Carcinoma

(83) In FIG. 2 a basic characterization of eIF protein expression levels in squamous cell carcinoma patients is shown. eIFs are altered in squamous cell carcinoma tissue compared to non neoplastic tissue (FIG. 2). The factors eIF1A, eIF4A, Rps6 and eIF6 are upregulated which is a new finding. In FIG. 3 mRNA levels of eIF6 and eIF1A was investigated in squamous cell carcinoma (SQCC) patients. The mRNA levels of eIF6 are upregulated in tumor tissue compared to healthy control. eIF1a is not altered in tumor samples compared to healthy controls.

(84) Immunohistochemistry data from a SQCC TMA show that eIF6 is significantly (p>0,001) upregulated in SQCC patient tissue (n=61) compared to healthy parenchymal tissue (n=31) (data not shown).

(85) Knock Down of eIF6 and eIF1A

(86) In FIGS. 4A and 4B results of eIF6 and eIF1A knockdown in A549 cells are displayed. Cell proliferation is significant reduced in silenced cells. eIF6 knock down in A549 cells leads to apoptosis. eIF1A knock down in A549 does not show significant more apoptosis then controls.

(87) Affimetrix Gene Expression Analysis of eIF6

(88) In FIG. 35, the gene expression analysis and the influence on patient overall survival of eIF6 in NSCLC, ADC and SQC is displayed. High expression in NSCLC of eIF6 shows a significant worse outcome for patient overall survival (p=1.5*10.sup.−13). The detailed analysis of eIF6 gene expression in ADC patients also significantly correlates with patient overall survival, high eIF6 expression leads to worse survival (p=3.8*10.sup.−13). Also in SQC eIF6 gene expression shows a significant impact on patient overall survival (p=0.14).

(89) Conclusion

(90) As eIFs are significantly altered in lung tumors they potentially represent an oncological biomarker. The present data indicate a major contribution of eIFs and mTOR signaling to the development and progression of lung carcinomas. 2 eIFs which are upregulated in NSCLC and not mentioned in literature have been identified. Knock down of eIF6 and eIF1A leads to proliferation inhibition. eIF6 knock down leads to apoptosis. High gene expression of eIF6 in NSCLC has significantly worse outcome for patient overall survival. Analysing the subgroups ADC and SQC, it also influences SQC significantly but has more significant impact on ADC.

(91) It is possible to stain eIF6 on immunohistochemistry (IHC) and determine eIF6 expression on routinely pathological bases and determine eIF6 protein expression level of Non Small Cell Lung Cancer (NSCLC) patients.

Example 2: Glioma

(92) Gliomas are brain tumors deriving from glial cell origin. They are classified in four tumor grades according to their neoplastic behavior by the World Health Organization (WHO): pilozytic astrocytoma (grade I), diffuse astrocytoma (grade II), anaplastic astrocytoma (grade III) and glioblastoma multiform (GBM, grade IV). Out of these grades, low-grade gliomas (LGG, grades I and II) account for rather benign neoplasias, whereas high-grade gliomas (HGG, grades III and IV) represent malignant tumor forms.

(93) GBMs have with more than 50% of all gliomas the highest occurrence of all gliomas. Although current GBM treatment strategies have improved over the past years by combining surgical resection, adjuvant radiotherapy and chemotherapy, the outcome is still very poor. The median survival of patients is approximately one year. One reason for the bad prognosis is the highly infiltrative nature of malignant gliomas which also leads to frequent recurrences. Therefore the development of novel therapeutic targets and treatments is strongly required.

(94) Materials & Methods

(95) qRT-PCR

(96) As housekeeping gene (HKG) a combination of glycerinaldehyd-3-phosphat-dehydrogenase (GAPDH) and succinate dehydrogenase A (SDHA) was used. Therefore the mean of the C(t) values of both HKGs was calculated and used for the ΔΔC(t))-calculation.

(97) In Silico Analyses

(98) For bioinformatical analyses two different online databases were evaluated regarding eIF gene expression in glioma patients: REMBRANDT (Repository of Molecular Brain Neoplasia Data) and The Cancer Genome Atlas (TCGA). REMBRANDT is an online data portal that comprises molecular research and clinical trial data related to brain cancers, including gliomas (https://rembrandt.nci.nih.gov/). It allows a molecular classification of tumors based on gene expression and genomic data from tumors of patients. eIF gene expression was then analyzed using R package ‘stats’ (version 2.15.3) software (http://www.r-project.org/). Numbers: Non tumor n=21, WHO grade II n=61, grade III n=47, grade IV n=191.

(99) The TCGA database is a collaborative project between the National Cancer Institute (NCI) and the National Human Genome Research Institute and comprises a collection of biomolecular investigations and clinical studies in the field of brain tumors (Tomczak K, et al. Contemporary Oncology. 19(2015):A68-A77). eIF gene expression in GBM patients was analysed in the TCGA database and correlated to the total patient survival in low grade glioma (LGG) and GBM patients. Additionally, the impact of temozolomide treatment was included in survival analyses. Statistical analyses were performed using the log-Rank test with defined significance values. Numbers: LGG n=389, GBM n=123, total n=535, without TMZ n=230.

(100) Chemosensitivity Testings in Murine Xenograft Models

(101) TABLE-US-00003 TABLE 3 Chemotherapeutic drugs used for chemosensitivity testings in vivo. Application Function Temozolomide First-line treatment GBM Alkylating agent (Temodal) Second-line treatment astrocytoma Bevacizumab Second-line treatment for GBM Anti-VEGF monoclonal (Avastin) antibody (angiogenesis inhibitor) Irinotecan First-line treatment colon carcinomas Topoisomerase Clinical trials for GBM treatment inhibitor Sorafenib Clinical trials for GBM treatment Protein kinase inhibitor Everolimus Clinical trials for LGG treatment mTOR inhibitor Salinomycin Experimentally tested for glioma therapy Polyether antibiotic Regorafenib Clinical trials for relapsed GBMs Tyrosine kinase inhibitor Thalidomide Clinical trials for recurrent GBMs Immunomodulatory Clinical trials in combination with te- drug mozolomide and irinotecan (VEGF inhibitor)

(102) Additionally to single treatments with each compound, combination therapies for Temozolomide/Everolimus, Temozolomide/Regorafenib and Temozolomide/Thalidomide were investigated.

(103) Chemosensitivity Testings in Human Glioma Cell Lines

(104) For in vitro chemosensitivity testings, same compounds were used as for in vivo testings. Additionally, three compounds were added to the in vitro testing panel (listed in table 4).

(105) TABLE-US-00004 TABLE 4 Chemotherapeutic drugs additionally used for in vitro chemosensitivity testings. Application Function Etoposide Clinical trials for GBM treatment Topoisomerase inhibitor SAHA (Vorinostat) Experimentally tested for glioma therapy Anti-epileptic drug Valproic acid Clinical trials for GBM treatment Histone-deacetylase inhibitor

(106) 3 distinct human neuroglioma cell lines were used for chemosensitivity testings: A172, U-87 MG (purchased from LGC Standards, Germany) and U251 MG (purchased from Sigma, Germany). A172 and U-251 MG cells were cultured at 37° C., 5% CO.sub.2 and 95% humidity in DMEM medium containing 10% fetal bovine serum and 10 mg/ml Penicillin/Streptomycin (growth medium, all Lonza, Belgium). U87-MG were cultured at 37° C., 5% CO.sub.2 and 95% humidity in EMEM medium containing 10% fetal bovine serum and 10 mg/ml Penicillin/Streptomycin (growth medium, all Lonza, Belgium).

(107) Chemosensitivity testings were performed over three different time points (1, 3 and 5 days) using three distinct concentrations of each compound (see table 5). As most of the compounds were dissolved in dimethyl sulfoxide (DMSO), DMSO treatment was used as control treatment. Three days before treatment, cells were seeded into 100 mm dishes (2×10.sup.4 cells/cm.sup.2). On treatment day, growth medium was removed and replaced with growth medium containing the respective compound. During treatment period, media containing the respective compounds were changed every second day. Cells were harvested after 1, 3 and 5 days and cell number and viability were determined. Cell pellets were then snap frozen and analyzed on protein (Immunoblot, immunofluorescence) and mRNA (qRT-PCR) level.

(108) TABLE-US-00005 TABLE 5 Concentrations for chemosensitivity testings in vitro. SAHA Valproic Temozolomide Irinotecan Etoposid Sorafenib Everolimus (Vorinostat) acid Salinomycin Concentration 1 20 μM 1 μM 10 nM 5 μM 50 nM 1 μM 1 mM 500 nM Concentration 2 50 μM 20 μM 100 nM 10 μM 100 nM 5 μM 5 mM 5 μM Concentration 3 100 μM 50 μM 10 μM 20 μM 1 μM 10 μM 10 mM 10 μM

(109) Results

(110) In Silico Analyses of eIF Gene Expression in Glioma Patients

(111) Before biochemical analyses, eIF expression was either evaluated in all 4 tumor grades or in grade 2-4 in silico. REMBRANDT data analyses revealed up-regulation in tumor grades 2-4 compared to control samples of following genes: eIF3C (FIG. 5f), eIF3D (FIG. 5g) A gradual increase in eIF gene expression over all three tumor grades could be observed for eIF3B (FIG. 5e), eIF3I (FIG. 5i), eIF4G1 (FIG. 5p) and eIF6 (FIG. 5s). An interesting expression pattern was identified for eIF3A (FIG. 5d) and eIF3H (FIG. 5h) as gene expression was up-regulated compared to controls, but grade II and III tumors showed a higher gene expression compared to grade IV. eIF4E (FIG. 5o) gene expression was down-regulated compared to controls.

(112) Additionally, eIF gene expression in GBM patients was statistically evaluated in the TCGA data base and correlated with patient survival probabilities. TCGA analyses were performed for LGG and GBM patients (FIG. 36). Interestingly, different eIF subunits significantly affected the survival of LGG and GBM patients. In LGG, low eIF3S2 (=eIF3I, p<0.01; FIG. 36a) and eIF4G (p<0.02; FIG. 36b) expression lead to a significantly increased overall survival in LGG patients. In GBM patients, low eIF4H expression levels (p<0.001) revealed a significantly higher overall survival (FIG. 36f).

(113) Taking all patients of the TCGA database without any exclusion criteria, statistical analyses revealed a higher survival probability for patients with increased eIF4EBP2 expression (leaf 2) in comparison to patients with lower eIF4EBP2 expression (p-value=2e−08). If additionally eIF4H expression is included, survival probabilities of GBM patients increased with low (leaf 6) compared to high eIF4H levels (leaf 7, FIG. 6a, b).

(114) Same analyses were performed with the exclusion of patients treated with temozolomide (TMZ), one of the standard chemotherapeutics for GBMs. Patients treated without TMZ have higher survival probabilities revealing low eIF4H and eIF1AX gene expression (leaf 4, p-value=9.07e−06). High eIF4H levels showed the same outcome in survival (leaf 3) as low eIF4H in addition to high eIF1AX expression (leaf 5, FIG. 6c, d).

(115) eIF mRNA Expression Patterns in Glioma Tissue

(116) eIF mRNA expression was examined using quantitative RT-PCR. Regarding the mRNA expression pattern, an up-regulation of the following genes was observed in all tumor grades compared to healthy control samples: eIF3A (FIG. 7a), eIF3B (FIG. 7b), eIF3C (FIG. 7c), eIF4A1 (FIG. 7f), eIF4G1 (FIG. 7i), eIF4H (FIG. 7j), eIF5 (FIG. 7k) and eIF6 (FIG. 7I). eIF3M (FIG. 7e) solely showed increased mRNA expression in grade II and III tumors. A stepwise increase over all tumor grades was observed for eIF4E (FIG. 7g), 4EBP1 (FIG. 7h) and eIF3I (FIG. 7d).

(117) Protein Expression Pattern in Glioma Patients

(118) eIF protein expression pattern was analyzed using two distinct methods. Protein expression analyses using immunoblot revealed an up-regulation of various eIF subunits (FIG. 8). eIF3B, eIF3I, eIF3M, eIF4A, eIF5 and eIF6 levels were elevated upon all four tumor grades. eIF3A levels seemed to be increased only in grade III astrocytomas. eIF4H showed decreased protein expression in all tumor tissues compared to healthy controls (FIG. 8).

(119) Additionally to immunoblot analyses, protein expression pattern was investigated immunohistochemically. Besides grade III, eIF3C protein levels are increased in all tumor types compared to healthy control tissue. The difference was even significant between grade II and controls (p<0.05, FIG. 9a. Immunohistochemical analyses of eIF4G protein expression revealed a significant downregulation in all tumor grades (grade II, III p<0.05, grade I p<0.001). eIF4G protein expression though seemed to gradually increase from grade I-IV (FIG. 9b). eIF4H levels were significantly up-regulated in all tumor grades in comparison to the control tissue (grade I, III, IV p<0.01, grade II p<0.001; FIG. 9c), eIF5 levels were slightly elevated in tumor samples, although the difference solely reached a significance level between grade II and controls (p<0.05; FIG. 9d). eIF6 protein expression was very strong within all tested tissue, but no difference between tumor and control tissue was detected (FIG. 9e).

(120) Therapeutic Relevance for eIFs in Glioma

(121) To evaluate the therapeutic relevance of eIFs in glioma, eIF protein expression was analyzed after chemosensitivity testings in murine GBM xenograft models (FIGS. 10, 11, 37 and 38). For the chemosensitivity testings several cytostatics, which are either already in clinical usage or in clinical trials for glioma treatment, were evaluated regarding their influence on eIF expression. After treatment with Everolimus, Sorafenib or Salinomycin and Thalidomide almost no difference in eIF protein expression was observed compared to PBS treated controls except for eIF3C, which was upregulated (FIGS. 10 and 11). Bevacizumab induced the down-regulation of eIF3K and Irinotecan treatment revealed differential effects. Those effects shown in FIGS. 10 and 11 could be also confirmed in higher xenograft numbers (FIG. 37). After Temozolomide treatment, which effectively reduced tumor growth, p-eIF2α (FIG. 37b), eIF3A (FIG. 37d), eIF3B (FIG. 37e), eIF3C (FIG. 37f), eIF3D (FIG. 37g), eIF3H (FIG. 37h), eIF3I (FIG. 37i), eIF3I (FIG. 37j), p-eIF4B (FIG. 37n), eIF4E (FIG. 37o), eIF4G (FIG. 37p) and eIF4H (FIG. 37q, p<0.01) protein levels were significantly reduced in comparison to the PBS control and other treatments. However, eIF6 (FIG. 37r) and eIF4A (FIG. 37m) did not response to Temozolomide treatment. In the Temozolomide resistant xenograft model X5 eIF6 protein expression was even increased compared to the non-resistant models (FIG. 37r).

(122) Interestingly, after treatment with Regorafenib, which is in clinical trials for relapsed GBMs, almost all eIF subunits were totally down-regulated compared to the PBS control (FIGS. 11 and 38, Patient X3). A similar effect was observed after Temozolomide treatment, which is the standard first-line therapy in GBM patients (FIGS. 10, 11 and 38). Combination treatments of Regorafenib, Everolimus or Thalidomide with Temozolomide did not reveal any differences in eIF protein expression compared to single Temozolomide treatment.

(123) The down-regulation of eIFs after Temozolomide treatment has been also confirmed in the U-87 MG cell line in vitro (FIG. 12). Several eIF subunits were down-regulated after Temozolomide treatment for 3 and 5 days. Almost no effect was observed after the first day of treatment. This implicates that Temozolomide induced down-regulation of eIF subunits is a process of at least 3 days. Interestingly, the highest Temozolomide concentration (100 μM) seemed to block eIF downregulation.

(124) In near future, also other treatments listed in Table 4 will be analyzed regarding their influence on eIF expression in three different glioma cell lines.

(125) Discussion

(126) Besides the already known eIFs, which might be involved in gliomagenesis, novel subunits altered in all glioma grades were found. Elevated levels of eIF3A, eIF3B, eIF3I, eIF3M, eIF4A, eIF5 and eIF6 were observed on protein and mRNA level. Additionally to the up-regulated eIF subunits, eIF4H was found to be down regulated on protein level using immunoblot analyses, but not immunohistochemistry.

(127) Especially eIF4H and eIF3I seemed to have a special role in gliomagenesis. TCGA analyses revealed an impact of eIF4H and eIF3I gene expression on glioma patient overall survival. Contrary results were found for eIF4H protein expression as it was down-regulated using immunoblot analyses and up-regulated in immunohistochemical analyses. This opposite trend has to be evaluated in more detail.

(128) Potential therapeutic relevance for eIFs in glioma treatment was shown after chemosensitivity testings in the murine xenograft model. After Regorafenib and Temozolomide treatment, eIF expression was totally down-regulated in an indirect manner. Down-regulation of various eIF subunits after Temozoloide treatment was also confirmed in human neuroglioma cell lines in vitro. Temozolomide, the most frequent used first-line therapy in high grade glioma, also showed the most effective tumor growth reduction in murine GBM xenografts. Thus, an eIF directed down-regulation via eIF inhibitors or siRNA in GBM patients might also reveal a positive therapeutic effect as Temozolomide.

Example 3: Lymphoma

(129) Blood cancer is one of the most important cancers in Europe. Within the wide-spread group of blood cancer, malignant lymphomas are a heterogeneous group of neoplastic disorders affecting the lymphatic system. 95% are of B-cell origin. Within B-cell lymphomas a further distinction can be made into Hodgkin's (HL) and non-Hodgkin's lymphoma (NHL). NHL comprise neoplasms with diverse biological and clinical manifestations, including the most common lymphoma subtype, the diffuse large B-cell lymphoma (DLBCL). The prognosis for these lymphatic neoplasms is still bad with 35% of affected patients dying of the disease within the first year after diagnosis. Treatment options are limited, mostly focusing on chemotherapeutic approaches.

(130) So far several publications about the impact of eIFs on lymphomagenesis and lymphoma progression exist. However, the involved research groups have mainly focused on the eIF4F complex, the most intensively studied eIF complex in all tumor entities, and its contribution.

(131) Importantly, there are no publications so far analyzing the relationship between the expression of the whole range of eIF-subunits and patient outcome. As mentioned above, the research focus has been mainly concentrated on the eIF4F-complex. Thus, research studies, investigating the complete range of eIFs, are lacking. Our research group intended to fill this gap.

(132) Materials and Methods

(133) Survival Analysis

(134) The survival analysis between respective eIF expression and patient survival was performed based on the Lenz-dataset which was published by Lenz G et al. (New Engl J Med 359(2008):2313-2323).

(135) Gene expression of 56 eIFs (see table 6) was analyzed in 200 Diffuse Large B-cell Lymphoma (DLBCL) patients which were treated with R-CHOP-chemotherapy (Combination therapy composed of the active ingredients: Rituximab, Cyclophosphamide, Hydroxydaunorubicin, Vincristine, Predniso(lo)ne). R-CHOP-chemotherapy is the standard treatment approach to treat this kind of lymphatic cancers. A panel of expert hemato-pathologists confirmed the diagnosis of DLBCL using current World Health Organization criteria. The gene expression was investigated on mRNA level by using Affymetrix U133 plus 2.0 microarrays (Affymetrix, USA).

(136) TABLE-US-00006 TABLE 6 EIF1 EIF3I EIF2B5 EIF4EBP2 EIF1AD EIF3J EIF2D EIF4EBP3 EIF1AX EIF3K EIF2S1 EIF4ENIF1 EIF1AY EIF3L EIF2S2 EIF4G1 EIF1B EIF3M EIF2S3 EIF4G2 EIF2A EIF4A1 EIF2S3L EIF4G3 EIF2AK1 EIF4A2 EIF3A EIF4H EIF2AK2 EIF4A3 EIF3B EIF5 EIF2AK3 EIF4B EIF3C EIF5A EIF2AK4 EIF4E EIF3CL EIF5A2 EIF2B1 EIF4E1B EIF3D EIF5AL1 EIF2B2 EIF4E2 EIF3E EIF5B EIF2B3 EIF4E3 EIF3F EIF6 EIF2B4 EIF4EBP1 EIF3G EIF3H

(137) DLBCL is caused by the abnormal multiplication of B-cells, which are very important parts of the lymphatic immune system. Like in other human cancers, this abnormal increase in the cell number of specific cells has detrimental effects on the body—leading eventually to the death of affected patients. Because the lymphatic system includes a great variety of different cell types the to be investigated B-cells have to be first of all isolated to be analyzed:

(138) Cell suspensions from three biopsy specimens were separated by means of flow cytometry into a CD19+ malignant subpopulation and a CD19− nonmalignant subpopulation. Before the gene expression of the isolated B-cells could be investigated by microarray analysis, the RNA samples had to be prepared: Gene expression profiling was performed after two rounds of linear amplification from total RNA. To interpret the microarray results regarding gene expression the following adaptations were performed: After normalization to a median signal of 500, provided in the Affymetrix Microarray Suite software, version 5.0 (MAS5.0, Affymetrix, USA), genes were selected that had a signal value greater than 128 in either the CD19+ or CD19− fractions in at least two of the sorted samples.

(139) The 1.sup.st quartile, Median and 3.sup.rd quartile refers to the cut off level for distinguishing high and low expression (see also FIG. 13-17 and Table 6). This means that a patient with an eIF expression higher than the 1.sup.st quartile has a higher eIF expression than the lowest eIF expressing quarter of the complete range of patients tested. In contrast, a patient with an eIF expression higher than the 3.sup.rd quartile has an eIF expression higher than three quarters of the tested patients (therefore a very high expression). To define significance a p-value of 0.05 was defined as significant.

(140) Cell Culture

(141) The lymphoma cell lines U-2932, RI-1, KARPAS 422, SU-DHL-4, OCI-Lyl, SU-DHL-10, NUDUL-1, SU-DHL-6 (all six DLBCL), Raji, BL2 (both Burkitt's lymphoma (BL)) and the spontaneously immortalized B-cell line MUG-CC1-LCL, derived from the tissue of a non-neoplastic donor as normal control, were grown in culture flasks and after one week of culture harvested by centrifugation. The pellets were washed with PBS buffer, the supernatant was discarded and the cell pellets stored at −80° C. until further usage.

(142) Cell pellets were homogenized using NP40-lysis buffer. The lysate was centrifuged for 10 minutes at 10 000 rpm and 4° C. Protein concentration of the resulting supernatant was determined with the Bradford protein assay and was adjusted to 3 μg/μL with SDS-Sample Buffer. Samples were stored at −80° C. until further usage.

(143) Immunohistochemistry

(144) Antibodies for immunohistochemistry were diluted as follows: 1:750 (eIF3c) and 1:500 (eIF-2a). To score the stainings only intensity and no density scores were used. No staining was termed as “0”, weak staining as “1”, moderate staining as “2” and strong staining as “3”. We analyzed the staining in tonsil tissue from patients suffering from chronic tonsillitis (non-neoplastic control) and lymph node tissue from patients suffering from DLBCL. Thereby we distinguished in the case of tonsils between the histological regions “mantle zone” and “germinal center” (with its centroblast and centrocyte cell type). Centroblasts are believed to be the progenitors for lymphoma cells. In the case of the DLBCL tissue we only scored the neoplastic B-cells, which infiltrated the major part of the lymph node destroying the original architecture.

(145) Results

(146) As described above a previously published data set comprising gene expression and patient survival profiles in DLBCL (Lenz G et al.) was analyzed for correlations between eIF mRNA expression and patient survival.

(147) Indeed, we detected for several eIF-subunits a link between altered expression and better or worse patient outcome. The expression/survival correlations are illustrated in FIG. 13-17.

(148) In the case of 9 eIF-subunits there was even a significant correlation between lower subunit expression and better patient outcome. In contrast, for eIF-4E3 the data indicate that higher gene expression seems to be significantly more beneficial for patient survival. 10 further eIF-subunits showed expression-survival correlations too, which however were not statistically significant (p>0.05).

(149) The results of the survival analysis are summarized in Table 7.

(150) TABLE-US-00007 TABLE 7 Results of the bioinformatic eIF expression-survival analysis. For the eIFs indicated at the left a correlation between altered mRNA expression (higher or lower expression compared to the rest of the patient population) and patient outcome was detected (see also FIG. 13-17). For example, a lower expression of the eIF2AK4-mRNA than the 3.sup.rd quartile seems to be better for patient survival. Survival analysis Gene 1.sup.st quartile Median 3.sup.rd quartile eIF cascade eIF2AK4 p = 0.069; low better eIF2AK3/HsPEK p = 0.096; high better eIF2B4 p = 0.057; low better eIF2C 3 p = 0.021; low better eIF2d p = 0.013; low better eIF-2α p = 0.014; low better eIF2S2 p = 0.009; low better eIF3b p = 0.020; low better eIF3c p = 0.031; low better eIF3d p = 0.031; low better eIF3f p = 0.065; low better eIF3g p = 0.071; low better eIF3l p = 0.043; low better eIF-4B p = 0.075; low better eIF-4E3 p = 0.016; high better 4E-BP1 p = 0.052; low better eIF-4G1 p = 0.060; low better eIF-5A p = 0.080; low better eIF-5 p = 0.077; high better

(151) To investigate eIF expression in lymphoma cells also in comparison with healthy B-cells, we performed cell culture studies comparing the immortalized B-cell line MUG-CC1-LCL (non-neoplastic control) with 8 lymphoma cell lines.

(152) We investigated protein expression of three eIF candidates, which were already in the survival analysis shown to be involved in lymphoma biology: eIF3b, eIF3c and eIF3d. As illustrated in FIG. 18, compared to the non-neoplastic control, the expression of all three tested eIFs was markedly increased in the 8 lymphoma cell lines.

(153) In addition, we also analyzed eIF expression in DLBCL patient samples and non-neoplastic control tissue (tonsils) using immunohistochemistry (preliminary data). The results are depicted in Table 8 and Table 9 and show that the analyzed eIFs (eIF3c and eIF-2α) could be detected by using immunohistochemistry. Regarding the scorings, the eIF staining was far more prominent in the germinal center than in the surrounding mantle zone (non-neoplastic tonsils). Having a closer look on the DLBCL patient tissue, the expression of the eIFs varied between the distinct DLBCL-patients, indicating that there is indeed a person-specific eIF expression pattern.

(154) TABLE-US-00008 TABLE 8 Immunohistochemical staining of eIF3c in non-neoplastic tonsils and neoplastic lymph nodes infiltrated with a DLBCL. The centroblasts are believed to be the progenitor cells for DLBCL and therefore one has to focus on their staining intensity within germinal centers. eIF3c Germinal center Non-neonlastic tonsils Mantle zone Centroblasts Centrocytes T1 0 2 0.5 T2 0 1 0.5 T3 0 1.5 0.5 T4 0 2 1 T5 0 3 1 T6 0 3 1.5 DLBCL Neoplastic cells D1 2 D2 0 D3 3 D4 3 D5 3 D6 2.5 D7 3 D8 2 D9 2 D10 1 D11 2.5 D12 2 No staining was termed as “0”, weak staining as “1”, moderate staining as “2” and strong staining as “3”.

(155) TABLE-US-00009 TABLE 9 Immunohistochemical staining of eIF-2α in non-neoplastic tonsils and neoplastic lymph nodes infiltrated with a DLBCL. The centroblasts are believed to be the progenitor cells for DLBCL and therefore one has to focus on their staining intensity within germinal centers. eIF-2α Germinal center Non-neonlastic tonsils Mantle zone Centroblasts Centrocytes T1 / 3 / T2 / 2 / T3 / 2.5 / T4 / 3 / T5 / 3 / T6 / 2 / DLBCL Neoplastic cells D1 3 D2 1 D3 3 D4 3 D5 2 D6 2 D7 3 D8 3 D9 3 D10 1 D11 2 D12 3 No staining was termed as “0”, weak staining as “1”, moderate staining as “2” and strong staining as “3”. “/” indicates that these tissue parts were not scored - in contrast to the immunohistochemical analysis of eIF3c.

(156) Discussion

(157) Till now, the research in the lymphoma area focused mainly on the eIF4F-complex. Therefore, research studies, investigating the complete range of eIFs, were lacking.

(158) Through our survival analysis, cell culture studies and immunohistochemical investigations we detected a potential for in the field previously unstudied eIFs to serve as treatment targets in DLBCL (7-9 and FIGS. 13-17).

(159) These findings could be directly translated into new treatment approaches for affected patients: First of all specific eIF levels could be analyzed by immunohistochemistry (Our immunohistochemical analysis showed that eIF detection in patient tissue samples is possible). Furthermore, the immunohistochemical analysis also indicated that eIFs seem to be more important for germinal center B-cells than for mantle zones. Germinal centers are the regions within lymphatic tissue where B-cells are trained and they are also believed to be the starting points at which the neoplastic DLBCL-cells develop. Therefore, the neoplastic DLBCL-cells could be more sensitive to induced expression alterations of eIFs than other lymphatic cellular subtypes.

(160) Thus, the expression levels of specific eIFs, which we found to be worse for the patient's survival when higher expressed, could for example be reduced as a therapeutic approach. The results of our cell culture studies additionally strengthened the use of such treatment approaches by showing an abnormally increased eIF3b, eIF3c and eIF3d expression in lymphoma cells when compared to non-neoplastic B-cells. Also in the survival analysis, eIF3b, eIF3c and eIF3d were shown to be worse for the patient's outcome when higher expressed. Due to financial constraints we could only test the referred three eIFs of the survival analysis also on protein level in cell culture. However, the promising results indicate that lymphoma cells could be highly sensitive to expression reduction of these eIFs by being dependent on respective high expression levels. Normal, non-neoplastic B-cells with a lower basal eIF expression could be left undamaged by such approaches (at least with regard to this specific eIF subgroup). Investigations on silencing particular eIFs in cell culture are currently already planned in our laboratory.

(161) Other eIFs, which are more beneficial for the patient at higher expression levels, could in contrast be upregulated. (Unfortunately, we could not test any member of this eIF group in our studies so far).

(162) Such course of action would add an additional treatment strategy and thus broaden the treatment options of affected patients.

Example 4: Hepatocellular Carcinoma (HCC)

(163) Materials and Methods

(164) Human Hepatocellular Carcinoma Samples

(165) Formalin fixed paraffin embedded HCC samples and respective healthy control tissues from a total of 234 patients were collected and were used to generate 10 tissue microarrays (TMAs). The histological diagnosis, differentiation, and stage were classified according to the WHO classification (Hamilton S et al. Pathology and Genetics of Tumors of the Digestive System. World Health Organization Classification of Tumours International Agency for Research on Canccer (IARC) 2000; IARC Press).

(166) Results

(167) Immunohistochemistry of Tissue Microarrays

(168) Density of the IHC staining was predominantly evaluated as 100%. In comparison to healthy liver tissue, several eIFs were highly upregulated in HCC tissue.

(169) IHC staining for eIF2α, eIF3H, eIF3C, eIF4E and eIF6 revealed a weak to strong staining in the Healthy liver tissue and also in the HCC tissue.

(170) For eIF5 the IHC staining displayed a high to moderate staining intensity in the HCC samples, whereas the intensity in healthy liver tissue was weak.

(171) Expression of eIF2α, eIF3C, eIF4E and eIF5 in HCC

(172) Protein expression of eIF2α, eIF3C, eIF4E and eIF5 was significantly upregulated in HCC samples and HCC samples with a HCV infection compared to healthy liver tissue. The p-value for this calculation was 0.051.

(173) Statistical Analysis

(174) The Survival Curve according to t-Stage displayed a better survival with a lower score ≤2 than with a score of 3 with a p-value of 0.179. The survival is also better when the patient has one tumor compared to patients with two or more. Microvessel invasion is associated with a poor clinical outcome compared to patients without a microvessel invasion with a p-value of 0.067. The survival according to sex is better for women compared to man with a p-vale of 0.6. The differences between patients with and without a HBV (Hepatitis B virus) infection showed no changes in the survival of these patients. In comparison with HCV (Hepatitis C virus) patients the survival is poor compared to patients without HCV infection with a p-value of 0.036.

(175) Discussion

(176) Liver cancer is the second leading cause of cancer mortality worldwide, with approximately 600,000 cancer related deaths. Altered translation initiation and abnormal gene expression increase the risk of cancer development. Previous studies displayed, that deregulation along the eIF cascade disassociated with malignant transformation and progression of cancer. The goal in this example was to analyze the contribution of various eIFs and their relating upstream mTOR targets in CRC (colorectal carcinoma), to find a link between translation initiation and carcinogenesis. A constitutive activation of mTOR signaling is shown to be a hallmark of cancer and is associated to cell growth and cell cycle progression. mTOR is a downstream target of AKT, which is highly overexpressed in CRC. AKT can be inhibited by the phosphatase and tensin homolog (PTEN), which acts as tumor suppressor. The loss of PTEN in mice results in formation of different cancer types. Active mTOR further phosphorylates its downstream targets S6K and 4E-BPT. Due to phosphorylation, 4E-BP1 dissociates from eIF4E and cap depended translation initiation is performed. Inhibition of mTOR expression by knock down experiments results in considerably decreased in vitro and in vivo cell growth in CRC.

(177) eIF5A, an indispensable member of the translation initiation process, is found to be aberrantly expressed in different malignancies including HCC, ovarian cancer, and lung cancer. One of its isoforms, eIF5A2, is overexpressed in HCC tissues, and this up-regulation may be a result of chromosome 3q amplification where the eIF5A2 gene resides. Clinical studies have demonstrated a correlation between up-regulation of eIF5A2 level with tumor metastasis and venous infiltration. Therefore, eIF5A2 has been proposed as an indicator of tumor invasiveness in HCC. In addition, targeting eIF5A2 by siRNA and combined treatment with GC7 effectively reduces the migration ability of tumor cells, suggesting that targeting eIF5A2 and hypusination could be a potential treatment for HCC.

(178) eIF4E is involved in the regulation of the mRNA translation process. It can enhance the translation of some important growth factors and cell growth regulators and affect protein synthesis, the cell cycle, cancer gene activation, and apoptosis; it also play an important role in malignant transformation and metastasis. eIF4E regulates the translation of cancer-related mRNAs that are involved in tumor occurrence and development.

(179) Previous studies showed an overexpression in malignant tumors including head and neck squamous cell carcinoma, laryngeal cancer, lung cancer, breast cancer, thyroid cancer and other cancer tissues. However, studies of eIF4E in liver cancer are rare. Studies showed that the protein expression three liver cancer cell lines were higher than in normal liver tissue. The HepG2 cell line had an especially high level of eIF4E protein expression. Based on these studies, eIF4E protein expression may be closely associated with the occurrence of human liver cancer development and prognosis. It has been confirmed in vivo and in vitro that sorafenib treatment can inhibit the RAF/MEK/ERK signaling transduction pathway, reduce the eIF4E phosphorylation level, reduce Mcl-1 protein, and induce hepatoma cell apoptosis. Accordingly, that lower levels of eIF4E gene expression may inhibit liver cancer. Targeting and adjusting the eIF4E level and activity may inhibit cancer cell growth, which may become a new paradigm in the field of biological treatment of liver cancer.

(180) Collectively, our data show that there are different eIF expressions in HCC and that some eIF subunits are overexpressed in HCC compared to normal liver tissue and therefore we aimed at elucidative the involvement of eIFs in the development of HCC formation and progression.

(181) The involvements of eIFs in cancer formation has been suggested and already, at least in part, have been proven for many eIF subunits and various tumor entities. eIFs can play a role, depending on the particular subunit and the respectively evaluated tissue types, in tumor development. The network of eIFs seems to display all elements of an entire oncogenic as well as tumor suppressive cascade. This thereby implicates enhanced eIF activation in HCC progression and suggests that eIFs may be an attractive target for HCC therapy.

Example 5: Colorectal Carcinoma

(182) Colorectal cancer (CRC) is the third most common cause of cancer related death and with more than one million cases annually the third most frequently diagnosed cancer entity worldwide. The risk factors for CRC comprise high fat intake, alcohol, red meal, obesity, smoking, increasing age and physical inactivity and is more prevalent in developed than developing countries. Current clinical management strategies include surgery, chemotherapy, radiation and palliative care, but they are not as effective as previously expected. Various numbers of drugs were shown to have antitumor activity against CRC or metastatic CRC, but improvements in the efficacy of current medications are necessary.

(183) Therefore it is important to better understand the pathogenesis of CRC in detail. Deregulation of protein synthesis has a major impact on cancer formation and progression.

(184) Materials and Methods

(185) Human Colorectal Cancer Samples

(186) Formalin fixed paraffin embedded CRC samples and liver metastases from CRC patients and respective healthy control tissue from a total of 44 patients were collected and were used to generate 2 tissue microarrays (TMAs). The histological diagnosis, differentiation, and stage were classified according to the WHO classification (S. Hamilton, et al. World Health Organization Classification of Tumours International Agency for Research on Canccer (IARC) 2000; IARC Press).

(187) The CRC TMA was composed of 346 tissue spots including carcinoma and healthy tissue of 16 patients suffering from CC (50% female; 50% male) and 11 patients with RC (27% female; 73% male). In addition a Liver-Metastases TMA (LM TMA) was generated. This included liver metastasis tissue from 11 CC (27% female; 73% male) and 6 RC patients (100% male) with respective healthy liver control tissue. Multiple metastases of these patients were used to generate the LM TMA with a total of 185 spots.

(188) 10 primary CC, 10 healthy colon controls and 10 primary RC, 10 healthy rectum controls served as protein- and mRNA controls.

(189) Chemosensitivity Testings

(190) TABLE-US-00010 TABLE 10 Chemotherapeutic drugs used for chemosensitivity testing. Drug Subclass Oxaliplatin* Alkylating agent Irinotecan* Topoismerase I Inhibitor 5-FU* Antimetabolite (Pyrimidinantagonist) Cetuximab* Epidermal growth factor receptor inhibitor AZD8931 Reversible inhibitor of signaling by epidermal growth factor receptor AZD6244 Mitogen-activated protein kinase kinase (MEK or MAPK/ERK kinases) 1 and 2 inhibitor Afatinib Tyrosine Kinase inhibitor Avastin* Angiogenesis inhibitor Regorafenib* Multi-kinase inhibitor Nintedanib Angiokinase inhibitor for VEGFR1/2/3, FGFR1/2/3 and PDG-FRα/β mTOR FR mTOR inhibitor IGF 1/2 mAB ** IGF-1/IGF-2 co-neutralizing monoclonal antibody AZ1 ** Aziridinylbenzoquinone Volitinib ** c-Met inhibitor Standard drugs for CRC treatment*; Novel drugs in preclinical testing **.

(191) Cell Culture HCT116

(192) HCT116 cell lines obtained from the American Type Culture Collection (ATCC) and were maintained in McCoy 5A medium supplemented with 10% fetal bovine serum (FBS) and penicillin/streptomycin (100 μg/ml), and incubated in a humidified atmosphere of 5% CO.sub.2 at 37° C.

(193) Results

(194) Overexpression of Members of the mTOR Pathway in Rectum Carcinoma Compared to Colon Carcinoma

(195) Protein expression of mTOR, PTEN, 4E-BP1 and AKT was significantly upregulated in RC samples compared to healthy control tissue (p<0.05). In CC mTOR, PTEN and 4E-BP1 revealed no significant changes in comparison to healthy control tissue besides AKT (p<0.05) showing a significant upregulation. Compared to CC statistical analysis revealed a significant upregulation of PTEN and 4E-BP1 in RC (p<0.05). Compared to healthy control tissue, CC and RC tissues revealed no obvious changes in their phosphorylation status concerning phospho mTOR Ser2448, phospho PTEN Ser380, phospho 4EBP1 Ser65 and phospho AKT Ser478.

(196) The results of the quantitative RT-PCR, with actin as internal reference, showed mTOR and PTEN as abundantly expressed in RCs, but not in CCs.

(197) Expression of eIF3, eIF4 and their Subunits in CRC

(198) IHC was performed on the CRC TMA representing different eIF subunits. Density of the IHC staining was predominantly evaluated as 100%.

(199) Immunohistochemistry of the analyzed eIF subunits did not show significant differences comparing CRC tissue and respective healthy control tissue. The observed staining intensities displayed an irregular expression pattern. IHC staining for eIF3H revealed strong staining in 100% of CRC tissue samples and healthy control samples. The staining for eIF3A and eIF3B displayed strong to weak staining intensities with irregular expression pattern. eIF3M revealed no staining in CRC tissue and respectively healthy control tissue.

(200) Compared to the healthy colon and rectum tissues, eIF3A, eIF3B, eIF3B, eIF3D and eIF3M were significantly increased on protein level in CC and RC samples (p<0.05) (FIG. 13A). On mRNA level eIF3A and eIF3B revealed a significant upregulation in RC samples. eIF3B displayed the highest upregulation by factor 40 (p<0.05), followed by eIF3A being about 28 fold increased (p<0.05). In comparison to CC eIF3H and eIF3M were increased on mRNA level, but no significant difference was observed on protein level (FIG. 13B).

(201) The protein expression of eIF3C, eIF3j and eIF3K was significantly upregulated in RC compared to CC and healthy tissue (p<0.05) (FIG. 13A). The mRNA expression of eIF3C and eIF3j showed an overexpression in RC compared to CC samples (FIG. 13B) compared to RC samples and healthy tissue.

(202) Immunohistochemistry of the analyzed eIF subunits did not show significant differences comparing CRC tissue and respectively healthy control tissue. The observed staining intensities displayed an irregular expression pattern. The staining for eIF4E and eIF4G displayed strong to weak staining intensities with irregular expression pattern.

(203) Compared to healthy control tissue, protein expression of Phosphor eIF4B, eIF4B and eIF4G was significantly increased in CC and RC samples. In CC tissue eIF4B showed the highest upregulation by factor 20 (p<0.05), followed by eIF4G being about 13 fold increased (p<0.05). The same could be detected in RC tissue. In case of CC, eIF4G showed the highest increase by 16 fold overexpression, eIF4B with a 6 fold overexpression. Statistical analysis of CC and RC samples revealed a significant upregulation of eIF4B in CC compared to RC (p<0.05) (FIG. 14A).

(204) Compared to healthy rectum control tissue, protein expression of eIF4E was significantly upregulated in RC samples. Protein expression of eIF4E was 5.5 times upregulated in RC (p<0.05) whereas in CC no significant changes were detectable. Statistical analysis revealed a significant upregulation of eIF4E in RC compared to CC (p<0.05) (FIG. 14A).

(205) The results of mRNA expression analysis of eIF4B and eIF4E demonstrated no changes of expression in CC and RC samples compared to normal controls. However, eIF4G displayed a higher mRNA expression in RC compared to CC samples (FIG. 14B).

(206) Expression of eIF2α, eIF5 and eIF6 in CRC

(207) IHC of the analyzed eIF subunits did not show significant differences comparing CRC tissue and respectively healthy control tissue. The observed staining intensities displayed an irregular expression pattern.

(208) IHC staining for eIF2α revealed strong staining in 100% of CRC tissue samples and healthy control samples. The staining for eIF6 displayed strong to weak staining intensities with irregular expression pattern.

(209) Protein expression of eIF2α was significantly upregulated in CC samples (p<0.05). eIF2α was 4.5 times upregulated in CC (p<0.05) whereas in RC no significant changes were detectable compared to healthy tissue. eIF5 and eIF6 were significantly increased in CC and RC samples compared to control tissue. ImageJ analysis of eIF5 and eIF6 protein pattern revealed a significant 4 fold upregulation in CC and RC tissue (p<0.05) (FIG. 15A).

(210) Protein Expression of Eukaryotic Initiation Factors in Colon Cancer Patients Derived Xenograft Models

(211) To display protein expression of different eIF subunits during chemotherapeutic treatment, 4 colon primary carcinoma PDX models and 1 colon metastasis PDX model were generated. In comparison to untreated control, mTOR showed a trend to be upregulated under treatment with Oxaliplatin and Cetuximab in colon primary carcinoma PDX models. A partly downregulation was visible in the AZ1 treated CPC PDX model. Protein expression of mTOR in the colon metastasis (CM) PDX model revealed a tendency to be upregulated under Afatinib treatment. According to untreated control, mTOR, eIF2α, eIF3J, eIF4B and eIF5 showed a tendency to be increased in the Afatinib treated CM PDX model. In addition PTEN seemed to be decreased in the Avastin treated CM PDX model. Protein expression of PTEN, eIF2α, eIF3A, eIF3J, eIF3B, eIF4B, eIF4G and eIF5 was heterogeneous and displayed no visible changes comparing untreated and treated colon cancer tissue. Statistical analysis using the Kruskal-Wallis test revealed that the observed tendencies were not statistically significant.

(212) Protein Expression of Eukaryotic Initiation Factors in Colon Cancer Patients Derived Xenograft Models

(213) To analyze protein expression of different eIF subunits under chemotherapeutic treatment, 4 rectum primary carcinoma (RPC) PDX models and 2 rectum metastasis (RM) PDX model were generated. Protein expression of eIF3J showed a tendency to be downregulated in RM PDX models under treatment with Nintedanib, mTOR FR and IGF 1/2 mAB. In addition these models displayed a trend in downregulation of eIF3A under treatment with Irinotecan, 5-FU, Avastin, Regorafenib, Nintedanib and mTOR FR. Protein expression of eIF4G was increased in Oxaliplatin and Cetuximab treated RPC PDX.

(214) Protein expression of PTEN, eIF2α, eIF3J, eIF3B, eIF4B and eIF5 was heterogeneous and displayed no visible changes comparing untreated and treated rectum cancer tissue.

(215) Statistical analysis using the Kruskal-Wallis test revealed that the observed tendencies were not statistically significant.

(216) IHC Staining of Liver Metastases from CRC

(217) Density of the IHC staining was evaluated as 100%. In comparison to healthy liver tissue, several eIFs were highly upregulated in liver metastases which were derived from CRC.

(218) IHC staining for eIF1 revealed no staining in healthy liver tissue samples, whereas liver metastases from CC and RC displayed a moderate to strong staining intensity (FIGS. 5A and 5B). The same was observed for IHC staining for eIF2α (FIGS. 5A and 5B), eIF3H (FIGS. 6A and 6B), eIF3B (FIGS. 5A and 5B) and eIF4G (FIGS. 6A and 6B). IHC staining for eIF4E (FIGS. 6A and 6B) revealed a moderate to high staining intensity in 63% of liver metastases from CC, whereas the intensity in metastases from RC was 90%. The same tendency was observed for eIF6 (FIGS. 6A and 6B) and eIF3A (FIGS. 5A and 5B).

(219) mRNA Expression of eIF Subunits in Liver Metastases from CRC

(220) The mRNA expression results verified the IHC staining for all eIFs, they showed a strong upregulation in liver metastases compared to healthy liver tissue. Compared to respective healthy control tissue, mRNA expression levels of eIF1, eIF3B (FIG. 5C), eIF4E and eIF4G (FIG. 6C) were significantly higher in liver metastases from CC and RC. In comparison to healthy liver tissue, mRNA expression of eIF2α, eIF3A (FIG. 5C) and eIF6 (FIG. 6C) was significantly upregulated in RC liver metastases, therefore, eIF3A was upregulated with a 28 fold increase (p<0.05) (FIG. 5C). eIF2α revealed a significant 20 fold upregulation (FIG. 5C). A 12 fold increase was observed for eIF6 (p<0.05) (FIG. 6C). Real-time analysis of eIF3C revealed a high upregulation 0.05) in CC liver metastases compared to healthy liver control tissue).

(221) Silencing of eIF1, eIF5 and eIF6 in CRC Cell Lines

(222) Based on the results of the eIF characterization in CC patients, eIF1, eIF5 and eIF6 turned out to be mainly altered and are thus promising candidates for future therapeutic approaches. Thus, in order to investigate the effect of silencing eIF1, eIF5 and eIF6, HCT116 cells were transfected with a siRNA and the subsequently knockdown effect was assessed for three time points. An inhibition of protein levels close to 90% was achieved for eIF1, eIF5 and eIF6 at all three time points. The transfection strongly reduced the proliferation of HCT116 cells which expressed eIF1, eIF5 and eIF6 specific siRNAs, but had no effect on MOK control.

(223) After transfection of HCT 116 cells with the respective siRNAs for 24h, 48h and 72h, mRNA expression of eIF1, eIF5 (24h p<0.01, p<0.05, p<0.01; 48h p<0.001; 72h p<0.001, p<0.01, p<0.01) and eIF6 (48h p<0.001; 72h p<0.05) in HCT116 cells were reduced for all three subunits compared to a negative control group. Moreover, the Immunoblotting results suggested a reduction of the protein levels by the silenced eIF1, eIF5 and eIF6 genes. The siRNA of eIF1, eIF5 and eIF6 gene constructs effectively inhibited the expression of eIF1, eIF5 and eIF6 gene. The effect of eIF1, eIF5 and eIF6 gene knockdown on apoptosis was analyzed by comparing the apoptosis levels upon eIF1, eIF5 and eIF6 knockdown with negative control cells by a YO-PRO®-1 staining. The apoptosis rate of the transfected cells with eIF1 (72h p<0.05), eIF5 (72h p<0.05, p<0.001) and eIF6 knockdown constructs (48h p<0.001; 72h p<0.001) was significantly decreased compared to negative control cells 72h after transfection with siRNA.

(224) After eIF1, eIF5 and eIF6 silencing, in vitro cell viability was significantly reduced. eIF1 (24h p<0.001; 48h p<0.001, p<0.01; 72h p<0.001), eIF5 (24h p<0.001, p<0.01; 48h p<0.001; 72h p<0.001) and eIF6 (24h p<0.001; 48h p<0.001; 72h p<0.001) silencing lead to a reduction of cell viability at all 3 time points (24h, 48, 72h).

(225) Furthermore, clonogenicity as evaluated by Giemsa staining. Colony formation was reduced 14 days after seeding in all transfected cells. The effect of eIF1, eIF5A and eIF6 knockdown on CRC cell motility was investigated by identifying the transmigration competence of cells through filters coated with an extracellular matrix. The cells exhibited a reduced capability to transmigrate upon eIF1, eIF5 (p<0.05, p<0.01) and eIF6 knockdown.

(226) Silencing of eIF1, eIF5 and eIF6 Leads to Reduced Translation

(227) The effects of eIF1, eIF5 and eIF6 knockdown on translation initiation were investigated by polysome profiling. After sucrose density gradient centrifugation of cell lysates, polysomes, 80S ribosomes and free 40S and 60S subunits were detected by monitoring their A254 nm as described in the methods section.

(228) Non-transfected HTC116 cells showed some free, unjoined 40S and 60S subunits, a large 80S peak and low amounts of polysomes. After eIF1 knockdown, increased levels of free 60S subunits, and a marked decrease of the 80S peak were observed, suggesting a defect in translation initiation. Furthermore, less polysomes were recorded in the eIF1 knockdown profile, indicating reduced translation rates. eIF5 knockdown also led to decreased levels of polysomes. In addition, the levels of free 40S and 60S ribosomal subunits relative to 80S ribosomes were increased, suggesting less efficient translation initiation. Similarly, also eIF6 knockdown resulted in a decrease in polysomes and an increase of the levels of free ribosomal subunits relative to 80S ribosomes.

(229) To conclude, knockdown of all three initiation factors caused defects in translation initiation, resulting in a reduction of polysomes indicating reduced overall translation.

(230) Discussion

(231) In this example it could be shown that protein expression pattern of pmTOR/mTOR, pPTEN/PTEN, p4EBP1/4EBP1 was significantly upregulated in RC samples compared to healthy rectum tissue. Increased levels of activated mTOR and 4E-BP1 might indicate the promotion of cap dependent translation and the involvement of the mTOR pathway to CRC carcinogenesis. Increased protein expression of mTOR and 4E-BP1 might also indicate the observed PTEN upregulation not sufficiently blocking their phosphorylation. Protein expression of pAKT/AKT was significantly upregulated in RC indicating increased activation of mTOR signaling and its downstream located eIF pathway components. Comparing CC and RC, PTEN was significantly increased in RC samples, which might indicate an increased inhibition of the mTOR pathway in rectum cancer. Also the mRNA data showed an abundant expression in RC compared to CC. Differences in the expression between CC and RC for eIF3C, elf3j and eIF3K were also observed, the protein level in RC tissue was increased. eIF3K was shown to interact with other eIF3 members including eIF3C and eIF3j and is suggested to be implicated in other processes, beyond translation initiation. eIF3C was found to be an oncogene and was shown to be significantly increased in cancer cells. An increase of eIF3H and eIF3M in CC and RC samples was also observed. Previous studies indicated eIF3H as being associated with CRC risk and suggested eIF3H to act as CRC susceptibility gene. eIF3 subunits are targeted in therapies for muscle atrophy and viral infection, but no eIF3 targeting agent is available for cancer therapy yet.

(232) Obtained protein data revealed a significant upregulation of eIF4B, eIF4B and eIF4G in CC and RC samples and demonstrated an increase for eIF4E only in RC samples compared to healthy control tissue. The mRNA expression analysis of eIF4B and eIF4E demonstrated no changes of expression in CC and RC samples compared to normal controls. In comparison a higher mRNA expression of eIF4G was found in RC compared to CC samples.

(233) eIF4B has been reported to modulate, in addition to eIF4G, the helicase activity of eIF4A and to establish bridges between mRNA and the 40S ribosomal subunit. Phosphorylation of eIF4B strengthens its interaction with eIF3A. Upon knockdown of eIF4B in HCT116 cells attenuated proliferation and increased stress-driven apoptosis was observed.

(234) The present protein data revealed a significant upregulation of various eIF subunits in CRC tissue. Protein expression of eIF2α was significantly upregulated in CC compared to healthy colon controls. This may indicate an increased level of tumor initiation and progression of CC. It is known that eIF5 is one of the essential core proteins of translation initiation. eIF6 operates as ribosomal anti-associated factor that binds to 60S ribosomal subunits in the nucleus and releases them into the cytoplasm after phosphorylation by growth factors. Interaction of the 60S with the 40S subunit is held and therefore translation initiation is blocked. eIF6 expression also limits cell growth and transformation. It is known that eIF6 is part of a multi-protein complex connected with the RNA-induced silencing complex (RISC), which is the major complex regulating miRNA activity. Previous studies showed that eIF6 is overexpressed in ovarian serous carcinoma, leukemia, head and neck carcinoma, as well as CC. We also saw a significant increase of eIF6 but only on mRNA level in high grade CC and RC samples. These findings suggest that eIF6 may play a central role in the translation initiation in low and high grade CC and RC.

(235) It was previously also shown that eIF5 is one of the essential core proteins of translation initiation. eIF5 is associated with eIF1, which is able to directly assemble with eIF3 and the (eIF2)-GTPMet-tRNAiMet. eIF5 over-expression was shown in different cancer types and is considered as predictive tumor marker. eIF1 was demonstrated to bind eIF5 and thereby potentially interferes with its GAP function. Previous studies in yeast showed that the interaction between eIF1 shift the 43S complex into a scanning-incompetent state, stalling it at the AUG. eIF5 overexpression involved in ovarian cancer and is regarded as a predictive tumor marker.

(236) As the eIF subunits 1, 5 and 6 seemed to be the most promising candidates in targeting CRC we also investigated them more detailed in knockdown experiments. eIF1, eIF5 and eIF6 were confirmed to be involved in cell proliferation by silencing in human CRC cell line HCT 116. After successful knockdown of eIF1, eIF5 and eIF6 on mRNA and protein level, proliferation rate and clonability of HCT 116 cells were significantly inhibited. Apoptosis significantly increased at late stage (72h). Changes in translation control and protein synthesis are key roles in tumor formation. Regarding our results, they include changes in protein synthesis and selective translational control of mRNA, and inhibition of tumor cell apoptosis. The knockdown of eIF1, eIF5 and eIF6 result in a reduction of polysomes indicating reduced overall translation.