The EIF4EBP1 gene encoding 4EBP1 is transcriptionally upregulated by MYC and linked to shorter survival in medulloblastoma

Abstract

Medulloblastoma (MB) is the most common malignant brain tumor in childhood and is stratified into four molecular groups ‒ Wingless and Int-1 (WNT), Sonic hedgehog (SHH), Group 3 and Group 4. Group 3 MB patients exhibit the poorest prognosis, with a 5-year overall survival of <60%, followed by Group 4 MB patients. Apart from MYC amplification in a subset of Group 3 MBs, the molecular pathomechanisms driving aggressiveness of these tumors remain incompletely characterized. The gene encoding the mTOR substrate and mRNA translation inhibitor eukaryotic translation initiation factor 4E-binding protein 1 (EIF4EBP1) represents a possible MYC target gene whose corresponding protein, 4EBP1, was shown to be more active in Group 3 versus Group 4 MBs. However, the prognostic role of 4EBP1 in MB and the mechanisms supporting 4EBP1 overexpression in Group 3 MB are still elusive. We analyzed EIF4EBP1 mRNA expression in publicly available data sets and found an upregulation in MB as compared to non-neoblastic brain. EIF4EBP1 mRNA expression levels were higher in Group 3 compared to Group 4 MBs. EIF4EBP1 mRNA expression was correlated with MYC expression, most prominently in Group 3 MBs. Survival analyses highlighted that high EIF4EBP1 mRNA expression was associated with reduced overall and event-free survival across all MB patients and in Group 3/Group 4 MB patients. Immunohistochemical evaluation of 4EBP1 protein expression in MB tissues confirmed that high levels of 4EBP1 are associated with poor outcome. Functional analyses revealed that MYC directly regulates EIF4EBP1 promoter activity, providing a mechanism for increased EIF4EBP1 mRNA levels in Group 3 MBs. Finally, we observed that 4EBP1 may support colony formation of in vitro cultured MB cells. Our data highlight that transcriptional upregulation of EIF4EBP1 by MYC promotes in vitro tumorigenicity of MB cells and associates with shorter survival of MB patients.

Fig. 1. EIF4EBP1 mRNA is upregulated and is co-expressed with MYC in MBs.

A Expression levels of EIF4EBP1 mRNA in a pool of non-neoplastic brain tissues (NNBT) (Roth et al. (n = 9) [63] and Pomeroy et al. (n = 11) [40] cohorts) compared to a pool of MB tissues (denBoer (n = 27) [64], Delattre (n = 54), Gilbertson (n = 73) [39], Hsieh (n = 22) [65], Kool et al. (n = 62) [41], Pfister (n = 223) [7] and Pomeroy (n = 188) [40] cohorts). B, C Expression levels of EIF4EBP1 mRNA according to the four MB groups SHH, WNT, Group 3 and Group 4 using the Cavalli et al. cohort [8] or a pool of the Kool et al., Gilbertson, Pfister and Pomeroy cohorts [, –41] compared to a pool of non-neoplastic brain tissue (NNBT) (Roth et al. [63] and Pomeroy et al. [40] cohorts) (see Table S2 for the number of patient samples per group and Table S7 for the results of pair-wise statistic tests between different groups). D Expression levels of EIF4EBP1 mRNA according to subgroups of Group 3 MBs from the Cavalli et al. cohort [8]. E Expression levels of EIF4EBP1 mRNA according to the Heidelberg subtypes from the Pfister cohort [7]. Significance in (A–E) was calculated using an unpaired and two-tailed parametric t test (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). F Expression levels of 4EBP1 protein in MB tissues clustered to the four groups, including two subsets of SHH (A and B) and Group 3 (A and B), from Archer et al. [43]. p value was calculated using an unpaired and two-tailed parametric t test (**p < 0.01). G–L Expression levels of EIF4EBP1 mRNA in MB patient samples plotted against the mRNA expression levels of MYC in all MB patients (G and H), in Group 3 MBs (I and J) or WNT MBs (K and L) using the Cavalli et al. [8] and the Pfister [7] cohorts as indicated (see Table S2 for the numbers of patient samples per group). Co-expression levels were quantified by calculating the Pearson correlation coefficient.

Fig. 2. EIF4EBP1 mRNA expression correlates with overall survival in all MB patients and in Group 3/Group 4 MB patients.

A–F Kaplan–Meier survival estimates of overall survival of MB patients stratified by their EIF4EBP1 mRNA expression levels across all MB patients (A, B), in Group 3 and 4 MB patients combined (C, D), in Group 3 MB patients (E) or in WNT MB patients (F) using data sets from the Cavalli et al. [8] and Pomeroy [40] cohorts as indicated. The data were obtained from R2 Genomics and visualization platform and the first versus last quartile was used as cut-off. Significance was calculated with the log-rank test.

Fig. 3. High 4EBP1 protein expression is associated with unfavorable prognosis of MB patients.

A, B Representative images of negative (A) and positive (B) 4EBP1 immunohistochemical staining of selected MB samples represented on the MBs TMA. C–F Kaplan–Meier survival estimates of overall survival (C, E) or progression free-survival (D, F) of MB patients stratified by their 4EBP1 staining score in all patients (C, D) or in Group 3 and Group 4 combined (E, F).

Fig. 4. MYC activates EIF4EBP1 promoter activity and transcription in MBs.

A ChIP peak locations within the human EIF4EBP1 promoter, exon 1 and part of intron 1 (hg38; Chr8: 38,030,342 - 38,031,906) from ChIP-sequencing data for MYC (Encode consortium, Encyclopedia of DNA Elements at UCSC [57, 58]) and an illustration of the luciferase reporter construct containing the EIF4EBP1 promoter, exon 1 and part of intron 1 (−192; +1372) coupled to Firefly luciferase, with the indicated binding sites of the transcription factor MYC. The three E boxes present in the promoter, and corresponding introduced mutations, are indicated. B, C HEK293-T cells were transfected with the (−192; +1372) EIF4EBP1 promoter reporter construct, together with 25 ng, 50 ng and 100 ng MYC (B) or the (−192; +1372) EIF4EBP1 promoter reporter constructs containing a mutation of each of the E boxes (as indicated in A), together with 25 ng MYC (C). For (B) and (C), a Renilla Luciferase vector was used as an internal control and luciferase activities were detected using the Dual-Luciferase Reporter Assay. Firefly luciferase activity was normalized to Renilla luciferase activity and the ratio was normalized to the corresponding 0 ng (B) or control (C) condition. Data represent the mean of four (B) or three (C) independent replicates ± standard deviation (SD). Significance was calculated using an unpaired and two-tailed parametric t test (*p < 0.05, ****p < 0.0001). A representative immunoblot analyzing expression of MYC is presented in (B). D, E Med8A (D) and HD-MB03 (E) MB cells were transiently transfected with negative control siRNAs (siCtrl), or two different siRNAs each targeting MYC (siMYC#1 and siMYC#2). Cells were re-transfected after 96 h with their corresponding siRNA and incubated for a total of 168 h. MRNA was harvested to determine the expression levels of EIF4EBP1 and MYC by qRT-PCR. Data obtained by qRT-PCR represent the mean of three independent replicates ± SD and the fold change in expression was normalized to the negative control (siCtrl). Significance was calculated using an unpaired and two-tailed parametric t test (*p < 0.05, **p < 0.01, ****p < 0.0001). F Med8A MB cells were transiently transfected with negative control (siCtrl) or a pool of four different siRNAs (see Table S5) targeting MYC (siMYC). Cells were incubated 72 h and protein was harvested for immunoblotting using the indicated antibodies. G, H Control and MYC overexpressing (MYC OE) ONS76 (G) or UW228.3 (H) cells were lysed. Levels of EIF4EBP1 mRNA were determined by qRT-PCR. Levels of 4EBP1 and MYC proteins were determined by immunoblots using the indicated antibodies. Data obtained by qRT-PCR represent the mean of three independent replicates ± SD and the fold change in expression was normalized to the control. Significance was calculated using an unpaired and two-tailed parametric t test (***p < 0.001, ****p < 0.0001).

Fig. 5. 4EBP1 contributes to the tumorigenic potential of MYC-amplified MB cell lines.

A, B Control (ishCtrl) or stable inducible 4EBP1 knockdown (ish4EBP1#1 and #2) Med8A (A) or HD-MB03 (B) cells were treated with 1 μg/ml doxycycline and grown in soft agar for 21 days. Colonies and single cells were counted, and colony formation efficiency was calculated and normalized to control. Data are reported as means ± SD of three individual replicates with indicated significance. Protein expression of MYC and 4EBP1 was analyzed by immunoblotting. Significance was calculated using an unpaired and two-tailed parametric t test (***p < 0.001, ****p < 0.0001). C, D Control (ishCtrl) or stable inducible 4EBP1 knock down (ish4EBP1#1 and #2) Med8A (C) or HD-MB03 (D) were treated with 1 μg/ml doxycycline and cell proliferation was measured with an EdU assay. Data are reported as means ± SD of three individual replicates. Statistics were calculated using an unpaired and two-tailed parametric t test (n.s. = not significant).