Introduction

Cancer cells reprogram their proteome in order to proliferate and adapt to the stressful tumor microenvironment. It is therefore not surprising that deregulations in expression and activity of translational regulators are often found in tumors. Frequently, overexpression of translation factors has been described in cancers to uncouple protein synthesis from inhibition by tumor cell stresses,1 which is consistent with the findings that elevated protein synthesis is a common feature in the vast majority of malignancies.2 In addition to these quantitative changes in overall protein synthesis, qualitative changes by selective translation of a subset of mRNAs modifies the proteome and favors neoplastic growth.2

Mammalian target of rapamycin (mTOR), functionally assembling to mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2) with a number of other proteins, is a central mediator of translational control, whose activation is influenced by several intra- and extracellular stimuli (see 4E-BP1 in cellular signaling pathways). mTORC1 essentially adapts mRNA translation rates by regulating the activity of its main downstream effectors: eukaryotic initiation factor 4E-binding protein 1 (EIF4EBP1, better known as 4E-BP1) and ribosomal protein S6 kinase 1 (S6K1).3 In addition, ribosomal protein S6 kinase 2 (S6K2), IGF2 mRNA-binding protein 2 (IMP2) and La-related protein 1 (LARP1) were also shown to be direct mTORC1 targets and to be synergistically implicated in the regulation of mRNA translation.4, 5, 6 In cancer, oncogenic activation of mTORC1 signaling because of alterations in signaling pathways upstream of mTORC1 leads to an increase of overall and selective mRNA translation.3 However, tumors often grow in hostile environments due to defective tumor vasculature or genotoxic or oxidative stress induced by rapid cell division or therapy.7 As mRNA translation is one of the most energy consuming cellular processes, mTORC1 activity has to be blocked to preserve energy and generate an adaptive response, which combines to maintain cell survival.8 Two of the most important cancer-promoting translational regulators, mTORC1 and eukaryotic initiation factor 4E (eIF4E), have been thoroughly investigated. However, the role of 4E-BP1 that is directly regulated by mTORC1 and directly regulates eIF4E in cancer is not well understood and is therefore the focus of this review.

4E-BP1 belongs to a family of eIF4E-binding proteins (4E-BP1, −2 and −3), each of which is encoded by a distinct gene.4 Nonetheless, they show a high degree of homology to each other.4, 9, 10 Among these family members, however, deregulations in 4E-BP1 expression or activation are reported far more frequently in a wide range of cancer entities compared with deregulations of 4E-BP2/3 (Table 1). The 4E-BP1 protein structure is given in Figure 1.9, 10 Non-phosphorylated and thus active 4E-BP1 inhibits cap-dependent translation initiation by binding eukaryotic translation initiation factor 4E (eIF4E) and prohibiting the formation of the 48S pre-initiation complex.11, 12

Table 1 Selected publications on 4E-BP1 dysregulations in most reported cancer entities
Full size table
Figure 1
figure 1

Protein structure and mTORC1-dependent phosphorylation sites of 4E-BP1. 4E-BP1 contains three functional domains: the eIF4E-binding domain,14 a C-terminal TOR signaling motif (TOS)173 and an N-terminal RAIP (Arg13, Ala14, Ile15 and Pro16) motif.174 The TOS and RAIP motif both contribute to the binding of Raptor,175, 176 a scaffold protein of mTOR building a ‘bridge’ between mTOR and 4E-BP1, necessary for efficient phosphorylation of 4E-BP1.29, 30 mTORC1-dependent phosphorylation of 4E-BP1 proceeds in a hierarchical way: initial phosphorylation of Thr-37 and Thr-46 is followed by Thr-70 and Ser-65,177, 178 whereby phosphorylation of Thr-37 and Thr-46 are thought to be the priming event for subsequent phosphorylation of the C-terminal phosphorylation sites (Thr-70, Ser-65).177, 178

Full size image

At least nine initiation factors are required in this process. Among them is eIF4F, which itself is composed of three initiation factors: eIF4G, eIF4A and eIF4E. eIF4E is the cap-binding protein, associating with 5'cap, while eIF4A functions as an RNA-helicase.13 eIF4G has the role of a scaffold protein and associates with eIF4A, eIF4E, poly-A-binding protein (PABP) and eIF3, thereby linking mRNA, eIF4F and the ribosomal 40S subunit to each other.13 As 4E-BP1 shares a common eIF4E-binding motif with eIF4G, the eIF4E-binding motifs of either protein compete for the same phylogenetically invariant dorsal surface of eIF4E.11, 12, 14 By competing with each other, 4E-BP1 can sterically block the binding of eIF4G to eIF4E and thereby prevent the assembly of the 48S pre-initiation complex and restrict translational activity.11, 12, 14 As tumor cells must adapt to different metabolic conditions in their microenvironment, the association rate of the eIF4F complex is tightly regulated. This is essentially mediated by mTORC1, which thus couples the rate of mRNA translation initiation to various extracellular stimuli.3, 8, 15, 16, 17

Although mTORC1-dependent phosphorylation sites are conserved in all three proteins, 4E-BP1 and 4E-BP2 are mostly regulated in a similar manner by mTORC1, whereas considerably less is known about the regulation of 4E-BP3.4 4E-BP2 additionally undergoes post-translational modification. For instance, asparagine deamidation leads to enhanced 4E-BP2 phosphorylation and inactivation by mTORC1.4, 18, 19 As every 4E-BP contains the eIF4E-binding motif, their physiological roles largely overlap, which further underlines their importance in translational control.10

As 4E-BP1 inhibits the pro-oncogenic eIF4E and thereby influences overall and selective mRNA translation (see 4E-BP1: regulation of overall and selective mRNA translation), it is mainly described as a tumor suppressor.20, 21 This is consistent with many studies reporting inactivation by hyperphosphorylation of 4E-BP1 in numerous tumor entities (Table 1). However, because tumor cells must adapt to different metabolic conditions in their microenvironment, mRNA translation must also be adapted. In this context 4E-BP1 can support tumor adaptation to stress and therefore may facilitate tumor progression by selectively modulating the translation of specific key transcripts, apart from overall translation rates. This complex dual role of 4E-BP1 in tumor progression, as well as its molecular and cellular functions will be the topic of this review.

4E-BP1 in cellular signaling pathways

The activity of 4E-BP1 is directly dependent on upstream signals controlling mTORC1 activity. This implies that 4E-BP1 activity is modulated by various stimuli and stress conditions, such as growth factors, nutrients and oxygen levels.8, 15, 17, 22 By responding to these signals, 4E-BP1 contributes to adaption of the rate of mRNA translation initiation to the intracellular metabolic status and extracellular stimuli.3, 8, 15, 16, 17 Here, we only briefly summarize the main signaling pathways influencing mTORC1 activity and 4E-BP1 phosphorylation, as they were extensively reviewed elsewhere.3, 15, 16, 17 Figure 2 gives an overview on the mTORC1 signaling pathway and the main modulating stimuli.

Figure 2
figure 2

mTORC1 signaling pathway and its upstream effectors regulating the activation status of 4E-BP1. 4E-BP1 activation is directly dependent on mTORC1 kinase activity, which is adjusted by several upstream regulators, mostly important growth factors, amino acids, energy and oxygen levels. Integrating these signals, mTORC1 is a key integrator of microenvironmental signals and regulates global and specific translation rates, cellular proliferation and tumorigenesis via 4E-BP1.

Full size image

Growth factors

Growth factors are potent stimulators of cellular proliferation. Consistently, signaling pathways associated with growth factor receptors are frequently deregulated in cancer, such as the phosphatidyl-inositol-3-kinase (PI3K) pathway and the MAPK/ERK pathway.23, 24, 25 As both of these pathways signal upstream of mTORC1, their oncogenic activation leads to an overactivation of mTORC13, 24 (Figure 2). Therefore, growth factor stimulation increases phosphorylation of 4E-BP1 via mTORC1 and enhances translational activity.11, 12, 14 As the role of these pathways in cancer has been already reviewed extensively,23, 24, 25 we here refer to these authors (Table 2).

Table 2 Proto-oncogenes and tumor suppressors affecting the regulation of translation rates by 4E-BP1
Full size table

Energy status

Depletion of nutrients or oxygen results in lower cellular energy levels. In order to proliferate, tumor cells have to strictly regulate their energy consumption, as protein synthesis is a highly energy consuming process.2 An important regulating kinase of nutrient signaling is 5'AMP-regulated kinase (AMPK), which acts as an intracellular energy sensor in response to changes in the cellular AMP:ATP and ADP:ATP ratio.26 During ATP-depletion, AMPK can block mTORC1 activity by two mechanisms: AMPK is able to activate TSC1/2, a tumor suppressing negative regulator of mTORC1, leading to lower levels of phosphorylated 4E-BP1 and a prohibition of protein synthesis under ATP-lacking conditions.27 AMPK can also reduce mTORC1 activity by phosphorylating Raptor,28 a scaffold protein of mTOR building a ‘bridge’ between mTOR and 4E-BP1, necessary for efficient phosphorylation of 4E-BP1.29, 30 Additionally, liver kinase B1 (LKB1) was shown to be an upstream regulator of AMPK, required for the activation of AMPK and thereby inhibiting mTORC1 pathway in response to nutrient starvation.31, 32, 33 Conversely, increased mTORC1 activity mediates enhanced mitochondrial activity and biogenesis by selectively promoting translation of nucleus-encoded mitochondria-related mRNAs by inactivation of 4E-BPs, which engenders an increase in ATP production capacity.34 Thereby mTORC1 is able to couple the increase of ATP consumption due to increased protein synthesis to an increase in ATP production.34

Oxygen levels

mTORC1 activity is inhibited by hypoxia through multiple independent mechanisms.35, 36, 37, 38, 39, 40 Hypoxia decreases the ADP:ATP ratio, in turn inducing AMPK to block mTORC1 signaling. In addition, by enhancing regulated in development and DNA damage response 1 (REDD1) transcription,35, 36, 37 hypoxia leads to lower mTORC1 activity by releasing TSC2 from its growth factor induced binding to 14-3-3 proteins.38 Finally, promyelocytic leukemia (PML) tumor suppressor and BCL2/adenovirus E1B 19-kDa protein-interacting protein 3 (BNIP3) were shown to inhibit the association of Rheb and mTORC1 under low oxygen conditions.39, 40 All these mechanisms converge in decreasing the level of phosphorylated (that is, inactive) 4E-BP1 in response to hypoxia.

Amino acids

Amino acids are a prerequisite for protein synthesis during cellular proliferation and therefore are strong stimuli of mTORC1 activity. Amino acids accumulate in the lysosomal lumen and initiate signaling in a vacuolar H+-adenosine triphosphate ATPase (v-ATPase)-dependent mechanism41 by which mTORC1 is recruited to the lysosomal membrane, where Rheb as a positive regulator of mTORC1 is located.42, 43, 44 Therefore, the stimulation of nutrients, especially of amino acids, can be seen as a prerequisite for mTORC1 activation by growth factors.

Other signals

Several other stimuli, such as genotoxic stress, Wnt stimulation, inflammation, phosphatidic acid (PA) and glucose have an influence on mTORC1 activity as well.45, 46, 47, 48, 49, 50, 51 Genotoxic stress and the resulting DNA damage increase p53 activity, which leads to an inhibition of mTORC1 signaling via activation of AMPK and enhanced expression of PTEN.45, 46 Stimulation of the Wnt pathway increases mTORC1 activity via inhibition of glycogen synthase kinase 3 (GSK3).47 Inflammation can enhance mTORC1 signaling via pro-inflammatory cytokines, such as TNFα or IκB kinase-β (IKKβ). Both inactivate TSC2 by direct interaction.48 Finally, PA increases the activity of mTOR by facilitating the assembly and stabilizing the mTORC1 and mTORC2 complexes.49, 50 Efeyan et al. also showed that glucose can recruit mTORC1 to the lysosmal membrane via Rag GTPases in order to activate mTORC1.51

4E-BP1: regulation of overall and selective mRNA translation

4E-BP1 was discovered as an interactor and inhibitor of the translation initiation factor eIF4E and therefore characterized as a repressor of overall mRNA translation.52 However, during past years it became increasingly clear that 4E-BP1 exerts selectivity in its ability to restrain translation through preferential blockage of the translation of specific transcripts.53 More surprisingly, 4E-BP1 may induce the translation of few specific transcripts through alternative mechanisms of translation initiation54 (Figure 3).

Figure 3
figure 3

Schematic illustration of 4E-BP1 regulating eIF4F assembly and its impact on global and selective translation rates. (a) Phosphorylated 4E-BP1 is not able to bind eIF4E and to prohibit the eIF4F formation, resulting in an increase of overall translation rate and specific translation of pro-oncogenic transcripts. (b) Non-phosphorylated 4E-BP1 binds eIF4E, prohibits eIF4F formation and thus leads to a decrease in overall translation rate and a decrease in specific translation of pro-oncogenic transcripts. On the other hand, the eIF4E-binding also has pro-oncogenic properties by leading to selective translation of mitochondrial proteins and a switch from cap-dependent to cap-independent translation supporting selective mRNA translation of the pro-angiogenic factors HIF-1α and VEGF under hypoxia.

Full size image

Control of overall mRNA translation by 4E-BP1

Upon activation 4E-BP1 interacts with eIF4E (see Introduction), thereby inhibiting mRNA translation initiation and reducing overall mRNA translation rates.52 Earlier studies reported that recombinant 4E-BP1 induces inhibition of cap-dependent translation of capped luciferase or chloramphenicol acetyltransferase reporter transcripts.9, 55 Such effects are prevented by addition of recombinant eIF4E or when a 4E-BP1 mutant protein lacking the eIF4E-binding site is assayed.9, 55

Selective inhibition of mRNA translation by 4E-BP1

Several studies revealed that, even though 4E-BP1-eIF4E impacts on overall mRNA translation activity,52 the translation of a subset of transcripts is preferentially regulated by 4E-BP1-eIF4E at the initiation step.53 A model for eIF4E-mediated selective translation emerged recently: because of its low expression rate, eIF4E is the least abundant translation initiation factor in most types of cells, resulting in a competitive situation for mRNAs to become efficiently translated.53 As a consequence, mRNAs with short and unstructured 5'UTR (as found in housekeeping genes), which are less dependent on the unwinding activity of the eIF4F complex, are efficiently translated but are largely unaffected by changes in eIF4E activity.53 In contrast, the long, G/C rich and highly structured 5'UTR-containing mRNAs, which are typically found in mRNAs of proto-oncogenes, growth factors and angiogenesis factors (that is, weak mRNAs), are inefficiently translated when the active eIF4F complex is limiting.53, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66 However, when eIF4E levels, activity or availability are increased, the translation of such ‘weak’ transcripts is preferentially enhanced; these include mRNAs of c-Myc, Cyclin D1, ODC, FGF2 and vascular endothelial growth factor (VEGF).53, 56, 67, 68, 69, 70, 71 Thus, 4E-BP1 will preferentially inhibit the translation of such pro-growth and pro-oncogenic transcripts by restraining eIF4E availability, which underlies its ability to block cell cycle progression72, 73, 74 (see Physiological role of 4E-BP1 in cell cycle progression, cell growth and proliferation).

Recent studies employing a transcriptome-wide approach revealed a broader picture of the mRNA landscape translationally regulated by 4E-BP1-eIF4E. Using ribosome profiling, Thoreen et al.75 highlighted that upon mTORC1 inhibition 4E-BP1 mediates selective translational repression of ribosomal protein and translation elongation factor mRNAs. Strikingly, these transcripts are characterized by the presence of a 5' terminal oligopyrimidine tract (5’TOP), a structural motif previously known to mediate selective translational repression following growth arrest.76 The authors proposed that active 4E-BP1, by interacting with eIF4E, selectively prevents the binding of eIF4E to 5’TOP-containing transcripts.75 Another study demonstrated that, in addition to transcripts of the translational apparatus, 4E-BP1 also represses translation of cell invasion and metastasis mRNAs.77 Such regulation is critical for the impact of 4E-BP1 on tumor cell invasion (see 4E-BP1 as an inhibitor of tumorigenesis). Finally, very recent data indicate that the translation of transcripts encoding antioxidant proteins is selectively enhanced by eIF4E in transformed mouse embryonic fibroblasts (MEFs).78 While the 5’UTRs of this subset of transcripts do not exhibit higher G/C content nor stronger secondary structure, they contain a cytosine-rich 15-nucleotide motif which may mediate the selective activity of eIF4E.78

Selective promotion of mRNA translation by 4E-BP1

Unexpectedly, it was reported that 4E-BP1 can positively regulate the translation of a subset of mRNAs.54, 79 In Drosophila, d4E-BP (the Drosophila 4E-BP1 ortholog) was shown to stimulate the translation of some transcripts encoding respiratory chain complex proteins and mitochondrial ribosomal proteins in response to dietary restriction.79 Indeed, overexpression of an active d4E-BP mutant leads to increased translation of reporter constructs containing the 5’UTR of some of these mRNAs.79 The underlying mechanism has yet to be characterized, but interestingly these 5’UTRs are characterized by their short size and weak secondary structure.79 Such d4E-BP selective translational control is critical to increase mitochondrial activity in response to dietary restriction, which supports lifespan extension of Drosophila.79 In addition, another study demonstrated that 4E-BP1 may increase the translation of the pro-angiogenesis factors HIF-1α and VEGF under hypoxia through a cap-independent mechanism.54 Overexpression of 4E-BP1, while restraining cap-dependent translational activity under hypoxia, promoted the translation of reporters containing the HIF-1α and VEGF internal ribosome entry sites (IRES), which are sequences known to mediate cap-independent translation.54 Such translational regulation has profound effects on tumor angiogenesis (see 4E-BP1 as a promoter of tumorigenesis).

Physiological role of 4E-BP1 in cell cycle progression, cell growth and proliferation

Cellular proliferation requires a proper coordination between cell growth (that is, an increase in cell size and mass) and cell cycle progression, even though these are both two separable processes.72, 74 This is critical to maintain the proper size of individual cells and organs while cells divide.72, 73, 74 Consequently, the relative rates of cell growth and cell cycle progression define the size of the resulting cells.72

A large body of evidence supports that the mTORC1 pathway regulates both cell growth and cell cycle progression. As mTORC1 senses growth factor and nutrient levels, it is critically involved in coupling the rate of cell growth to cell cycle progression.72 Indeed it was shown that rapamycin inhibits cell cycle progression in all cell cycle phases, wherein cells in G1 phase are affected the most.74, 80 Braun-Dullaeus et al. report that hyperphosphorylated 4E-BP1 is associated with increased levels of cyclin B1, D1, E and Cyclin dependent kinases 1, 2 (CDK1 and CDK2) among others and that this overexpression can be blocked by rapamycin, whereas mRNA levels stay unaltered.81 Consistently, Cyclin E levels, which are critical for G1/S transition, are lower in dTOR null mutant Drosophila leading to cell cycle arrest, which can be prevented by ectopic expression of cyclin E.82 Enhancement of S6K1 activity or eIF4E overexpression increases cell size and accelerates S-phase entry,73, 74 while overexpression of a phosphorylation site-defective 4E-BP1 leads to a reduction of cell size74 and of G1/S-phase transition rates.73, 74, 82 Similarly, ectopic expression of overactive d4E-BP1 mutant in Drosophila causes smaller wing size by reduction of cell size and cell number.83 Altogether these results indicate that mTORC1 signaling and 4E-BP1 play a central role in both cell growth and cell cycle progression.

However, this model proposing that both 4E-BP1 and S6Ks can indistinctly control both cell size and cell cycle progression downstream of mTORC1 has been challenged.84 Indeed, Dowling et al. establish that cell cycle progression and cell size are separately regulated in higher eukaryotes by 4E-BP1 and S6Ks, respectively.84 This study showed that depletion of Raptor by short-hairpin RNA (shRNA) impairs G1/S progression in wild-type MEFs, whereas this effect was not seen in 4E-BP1/2 double-knockout mouse embryonic fibroblasts (MEFs). In contrast, depletion of Raptor, serum deprivation or addition of an active-site TOR inhibitor (asTORi) decreases cell size to the same extent in both wild-type and 4E-BP1/2 double-knockout cells, indicating a role for 4E-BP1 in cell cycle regulation, but not in cell growth regulation. However, S6Ks were shown to have an impact on cell growth, but not on cell cycle progression.84 Additionally, Lynch et al. reported that an overactive 4E-BP1 mutant can block cell proliferation, without influencing cell size.85 Mechanistically, it was proposed that 4E-BP1 selectively inhibits the translation of transcripts required for cell proliferation such as those encoding ODC, cyclin D3 and VEGF84 (see Selective inhibition of mRNA translation by 4E-BP1). Based on these results it is tempting to speculate that 4E-BPs regulate cell cycle progression but not cell growth in higher eukaryotes, whereas S6Ks play a crucial role in cell growth, but not in cell cycle progression.

The role of 4E-BP1 in pathophysiology of cancer

The mTORC1 pathway and its downstream effectors have a pivotal role in cell growth and cell cycle progression.3, 72, 73, 74 It is therefore expected, that the mTORC1 pathway, including 4E-BP1, is highly relevant in the pathophysiology of cancer, which is consistent with its frequent deregulation observed in various malignancies86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137 (Table 1).

Role of eIF4E in cancer

Most oncogenic and tumor suppressing upstream effectors of 4E-BP1 were reviewed extensively23, 24, 25, 53, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147 (Table 2). Thus, in order to understand the role of 4E-BP1 in cancer, it is necessary to characterize its downstream target eIF4E here, especially as eIF4E may act as an oncogene.53 Indeed, overexpression of eIF4E stimulates cell proliferation and is sufficient to transform embryonic fibroblasts.148, 149, 150 Conversely, genetic targeting of eIF4E abolishes Ras-mediated transformation of embryonic fibroblasts in vitro and in vivo,151, 152 whereas the overexpression of eIF4E in CREF fibroblasts induces the formation of both spontaneous and experimental metastases and leads to quicker metastatic spread when re-injecting eIF4E high expressing cells into nude mice.65, 66 Furthermore, eIF4E is able to enhance the selective synthesis of known oncogenes and cancer-promoting factors, such as c-Myc, Cyclin D1, ODC, FGF2 and VEGF, MMP9 and Heparanase.53, 145 Finally, eIF4E can prevent abnormal accumulation of reactive oxygen species by mediating the selective synthesis of antioxidant proteins to support cellular transformation.78 These results indicate a crucial role for eIF4E in oncogenesis, angiogenesis and metastatic spread.

Consistently, eIF4E was shown to be overexpressed in many solid tumors and cancer cell lines (Table 1), such as colorectal, breast, bladder, lung, prostate, gastric, head and neck carcinomas, lymphomas and neuroblastomas.53, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162 A role for eIF4E as a prognostic marker has been suggested for certain cancers.163 Increased expression of eIF4E is for example correlated with poorer clinical outcome and decreased survival in breast, head and neck, colorectal, lung, prostate, bladder, skin and cervical carcinomas, as well as in lymphomas.1, 53, 164, 165 Furthermore, eIF4E is associated with increased malignancy in meningiomas, glioblastomas, astrocytomas,1, 166 as well as decreased survival rates in advanced prostate cancer1, 128 and locally advanced esophageal cancer.1, 106, 128

4E-BP1: inhibitor or promoter of tumorigenesis?

In numerous cancer entities an overactivation of mTORC1 is reported, leading to enhanced 4E-BP1 phosphorylation/inactivation (Table 1). Such inactivation is expected to prevent 4E-BP1-mediated inhibition of the oncogenic eIF4E (see Role of eIF4E in cancer). However, several authors also report a role for non-phosphorylated and hence active 4E-BP1 in facilitating tumorigenesis, especially under conditions of cellular stress in the tumor's microenvironment (see 4E-BP1 as a promoter of tumorigenesis). It is therefore tempting to speculate that high levels of phosphorylated, inactive 4E-BP1 in a highly vascularized tumor promote tumor progression due to less translational inhibition of pro-oncogenic eIF4E-sensitive transcripts. In contrast, in the poorly vascularized center of a fast growing tumor, the cells adapt to starvation by not inactivating 4E-BP1, which may selectively promote the translation of mRNAs that support survival of tumor cells under starvation (Figure 3). Figure 4 gives an overview on the tumor suppressing and pro-oncogenic functions of 4E-BP1.

Figure 4
figure 4

Overview on tumor suppressing and pro-oncogenic functions of 4E-BP1.

Full size image

4E-BP1 as an inhibitor of tumorigenesis

Overexpression of a constitutively active 4E-BP1 mutant is able to decrease cell size, to inhibit cell cycle progression (decrease of G1-progression rates), to suppress tumorigenicity and to mimick rapamycin treatment.73, 74, 85, 167 Furthermore, 4E-BP1 was shown to inhibit oncogene-mediated transformation of rat embryonic fibroblasts in vitro and in vivo by inhibiting cell proliferation and inducing apoptosis.21, 69, 85 In prostate cancer cells, overexpression of a constitutively active 4E-BP1 mutant significantly decreases prostate cancer cell invasion without affecting cell cycle progression.77 Conversely, silencing 4E-BP1 causes colon cancer epithelial cells to undergo epithelial-mesenchymal transition (EMT) and promotes cell migratory, invasive capabilities and metastasis.168 In addition, Petroulakis et al.169 demonstrated that 4E-BP1/2 double-knockout mice crossed with p53 knockout mice exhibit significantly shorter tumor-free survival than p53 knockout mice, indicating a tumor suppressor-like role for 4E-BP1. However, 4E-BP1 cannot be considered a genuine tumor suppressor on its own, as it was reported that 4E-BP1 knockout mice show no evidence of tumor development.170

The activation status of 4E-BP1 has been analyzed in numerous human tumor samples. In most cases, this was performed by assessing the levels of phospho-4E-BP1 (p-4E-BP1), whereas total 4E-BP1 levels were mostly not assessed. However, the mere anaylsis of p-4E-BP1 is not sufficient to determine whether 4E-BP1 is inactive. This requires determining the ratio between p-4E-BP1 and total 4E-BP1 levels. Given that data for total 4E-BP1 levels are lacking in most tumor samples analyzed for p-4E-BP1 levels, we argue that no definitive conclusion can be drawn on the activation status of 4E-BP1 in these samples. In addition, the ratio between eIF4E and 4E-BP1 levels is a determinant for 4E-BP1 activity as previously reported.171 Therefore, further studies are warranted to determine p-4E-BP1:4E-BP1 ratio and eIF4E:4E-BP1 ratio in human tumor samples. It remains that overexpression of p-4E-BP1 is reported in a number of tumor entities, such as colorectal,102 breast,96, 97 lung115, 116, 117 and prostate carcinoma,128 as well as leukemia86, 87 (Table 1). The level of p-4E-BP1 positively correlates with different outcome variables, such as poor prognosis, relapse, poor differentiation, tumor size, metastasis and tumor progression (Table 1). An increased eIF4E:4E-BP1 ratio is for example associated with higher risk for relapse in head and neck squamous cell carcinoma.113

4E-BP1 as a promoter of tumorigenesis

Paradoxically, Petroulakis et al. showed that 4E-BP1/2 double-knockout embryonic fibroblasts expressing p53 are resistant to Ras-mediated transformation and undergo cellular senescence instead. The authors propose a mechanism by which loss of 4E-BP1/2 enhances translation of Gas2 mRNA, encoding a p53 stabilizing protein, thereby leading to p53 accumulation and induction of cellular senescence.169 In glioblastoma cells, the absence of 4E-BP1 renders tumors more sensitive to metabolic and genotoxic stresses. Indeed, 4E-BP1 knocked down tumors exhibit increased sensitivity to hypoxia-induced cell death in vitro and to radiation in vivo due to a decrease in the viable fraction of radioresistant hypoxic cells.172 4E-BP1 can further contribute to development of aggressive breast carcinoma by supporting tumor angiogenesis.54 Advanced breast carcinomas were shown to overexpress 4E-BP1, leading to a hypoxia-induced switch from cap-dependent to cap-independent translation supporting the selective mRNA translation of pro-angiogenic factors HIF-1α and VEGF under hypoxia.54 Consequently, knockdown of 4E-BP1 hampers growth of breast cancer xenografts as a result of reduced tumor vascular density.54 Several studies reporting on 4E-BP1 levels in cancer show a correlation between 4E-BP1 levels and different variables indicating poor outcome.93, 103, 107, 109, 131 High 4E-BP1 mRNA expression levels are reported to be an independent prognostic factor for poor outcome in breast cancer.99, 98

Conclusion and perspectives

4E-BP1 is a central integration node of several signaling pathways sensing growth factor stimulation, nutrients or oxygen level. Thereby, 4E-BP1 mediates translational adaptation of tumor cells growing in variable microenvironments by regulating overall and selective mRNA translation. As an inhibitor of the pro-oncogenic eIF4E, 4E-BP1 is mainly described as a tumor suppressor. In contrast, 4E-BP1 may also promote tumor progression by selective mRNA translation, in particular under conditions of cellular stress, concluding that 4E-BP1 contains a dual role in tumor progression. Consistently, dysregulations in 4E-BP1's phosphorylation status or expression rate were described for many tumor entities and contain a prognostic value, although in future studies the levels of 4E-BP1, p-4E-BP1 and eIF4E have to be determined altogether, as the biological function of either protein depends on the expression and activation of the others. Nevertheless, 4E-BP1 not only has clinical relevance as a prognostic and/or predictive biomarker but also as a potential pharmacological target for a more specific and less toxic future anti-cancer therapy.