Insulin-like growth factor 2 mRNA-binding protein 1 (IGF2BP1) in cancer

The insulin-like growth factor-2 mRNA-binding protein 1 (IGF2BP1), a member of a conserved family of single-stranded RNA-binding proteins (IGF2BP1-3), expresses in a broad range of fetal tissues and more than 16 cancers but only in a limited number of normal adult tissues. This gene is required for the transport of certain mRNAs that play essential roles in embryogenesis, carcinogenesis, and chemo-resistance [1, 2], by affecting their stability, translatability, or localization [1, 3, 4]. IGF2BP1 consists of six canonical RNA-binding domains, including four K homology (KH) domains and two RNA recognition motifs (RRMs) (Fig. 1a) [5]. Even though the RRM domains of IGF2BP1 can potentially contribute to the stabilization of IGF2BP-RNA complexes in a target-dependent manner [6], in vitro studies indicated that RNA binding was majorly facilitated by the KH domains [7]. The KH1/2 domain is significant for the stabilization of IGF2BP-RNA complexes. For example, the KH1/2 domain could regulate binding of IGF2BP1 to cis-determinants in the ACTB 3′-UTR as well as, more strikingly, the MYC-CRD (coding region stability determinant) RNA in vitro [8]. However, recent structural analyses of KH3/4 domain of human IGF2BP1 showed the formation of an antiparallel pseudo-dimer conformation where KH3 and KH4 each contacts the targeted RNA [9]. As far as we know, the KH domains, particularly the KH3–4 di-domain, are essential for the binding of IGF2BP1 to targeted mRNAs in a N6-methyladenosine (m⁶A)-dependent manner in which KH domains recognize the consensus GG (m⁶A) C sequence of mRNAs [10]. IGF2BP1 is considered as a m⁶A-binding protein, with > 3000 mRNA transcript targets [11]. Importantly, as shown in Fig. 1b, the m⁶A alterations of those mRNAs are required for the targeting of IGF2BP1 with mRNAs such as MYC, as well as for IGF2BP1-mediated control of mRNA expression. In addition, some co-factors of IGF2BP1, such as ELAV-like RNA binding protein 1 (ELAVL1), which may be recruited by IGF2BP1, protect m⁶A-containing mRNAs from degradation and subsequently promote their translation [10]. However, other m⁶A-binding proteins may be also the readers of the m⁶A-containing mRNAs that are also recognized by IGF2BP1, whereas result in different effects on those mRNAs from that of IGF2BP1 does. A set of YT521-B homology (YTH) domain-containing proteins (YTHDFs), for instance, can read and control the fate of m⁶A-containing mRNAs via modulating pre-mRNA splicing, promoting translation, or facilitating mRNA decay (Fig. 1c) [12‐15].

In numerous studies in vivo and in vitro, the emerging cancer-related mRNAs have been found, including PTEN, ACTB, MAPK4, MKI67, c-MYC, and CD44. By regulating those mRNAs, IGF2BP1 has been identified to play important roles in cell proliferation and growth of normal tissues and tumor tissues, as well as tumor cell adhesion, apoptosis, migration, and invasion [8]. Thus, IGF2BP1 is considered to be one of the most promising therapeutic targets for treating cancers, as well as the use of inhibitors of IGF2BP1-mediated cell signaling would emerge as a potential strategy for cancer treatment. However, a few recent studies have found its suppressive role in tumor growth and metastasis [16, 17]. Therefore, mechanisms resulting in the paradoxical findings need to be elucidated. Additionally, the emerging non-coding RNAs (ncRNAs) including microRNAs (miRNAs) and long non-coding RNAs (lncRNAs) have been demonstrated to be involved in the mediation of cancer onset and progression by targeting IGFBP1 and thus could become novel therapeutic targets of cancers. In this review, we provided a new overview of the roles of IGF2BP1 in embryo development and in multiple cancers. We focused on the interconnectivity between IGF2BP1 and its targeted mRNAs or ncRNAs involved in the biological processes of embryogenesis and tumorigenesis, as well as aid in the identification of potential targets for cancer therapy and contribute to the cancer drug-discovery research.

The essential role of IGF2BP1 in embryo development has been established. IGF2BP1 is highly expressed during the stages between zygote and embryo phases, and nearly abolished in the normal adult organism [8]. In embryo studies of Xenopus, zebrafish, and mice, this gene was found to be expressed in various developmental cell types, including the migrating neural crest and branchial arches, and cranial neural crest (CNC) [18]. Notably, its modest expression was observed in the lung, spleen, and brain of 16-week-old male mice [8]. However, one report found that IGF2BP1 is not only ubiquitously expressed in organs during human embryonic development, but also much less presented in adult human prostate, testis, kidneys, and ovaries [19]. Therefore, the expression pattern observed for IGF2BP1 could indeed be characterized as “oncofetal,” since it is largely absent from normal adult tissues.

Mice with IGF2BP1 deficiency exhibit dwarfism, severely decreased viability, and impaired gut development. Similarly, knockdown of this gene may lead to 60% of perinatal death and significantly smaller body with hypoplastic tissues among almost all organs in mice [20]. The SNP rs9674544 in IGF2BP1 was identified to be significantly associated with primary tooth development in infancy [21]. Furthermore, in the neuronal development, IGF2BP1 regulates the neurite outgrowth, neuronal cell migration, and axonal guidance partially by controlling the spatiotemporal activation of protein synthesis such as ACTB mRNA [22, 23]. IGF2BP1 controls the subcellular sorting of the ACTB in primary fibroblasts and neurons by binding to the cis-acting zipcode in the ACTB’s 3-UTR [24]. Moreover, IGF2BP1 was also involved in determining cell fate in testis stem cells and controlling neuronal differentiation and matured neuronal system during regeneration [25‐27]. One study showed that downregulation of IGF2BP1 expression in the dorsal neural tube was both necessary and sufficient for the delamination and emigration of CNC, whereas inhibition of its expression enhanced CNC delamination and induced epithelial dissociation. Furthermore, IGF2BP1 expression is negatively associated with epithelial-to-mesenchymal transition and plays an important role in sustaining epithelial integrity. Those regulation processes might involve partly the mechanism of which IGF2BP1 interacts with ITGA6 mRNA, either directly or indirectly, to control its expression [28]. In another study on the role of IGF2BP1 in amphibian neural crest migration, the reduced IGF2BP1 expression by antisense morpholino oligonucleotides (AMO), throughout the entire embryo, was showed to increase CNC migration, suggesting the reduction in CNC migration originally observed in the AMO-injected Xenopus embryos is a result of a global, non-cell autonomous reduction in IGF2BP1 expression [18]. Those findings reveal the essential roles of IGF2BP1 in regulating cell growth and differentiation during development of organisms and suggest that aberrant expression of this gene could cause dysplasia of tissues and organs by dysregulating the expression levels of its targets such as ACTB mRNA and ITGA6 mRNA.

So far, the regulatory networks that IGF2BP1 participates in embryonic development are little known. Interestingly, a novel ultra-conserved lncRNA, THOR (ENSG00000226856), which is testis-specific in adult tissues of human, zebrafish, and mouse, has been demonstrated to be broadly expressed during the early development of both zebrafish and mouse. By IGF2BP1 binding to exons 2 and 3 of THOR, it regulates IGF2BP1’s target mRNA levels to promote tumorigenesis [29]. Therefore, it may be deduced that THOR involves the development of organs and tissues and tumorigenesis by modulating IGF2BP1 expression levels, but the specifically regulated target mRNAs remain to be explored.

IGF2BP1 and 3 have an amino acid sequence identity of 73% with each other and many the same or similar functions in cytosol. The main IGF2BP family member described in the context of cancers is IGF2BP3 [30, 31]. The comprehensive description in the regulative mechanisms of IGF2BP1 in human cancers is little, although this gene has been demonstrated to play important roles in tumorigenesis and drug-resistance of cancer therapy in vitro and in vivo studies. Even though most of the presently found cancer-related mRNA targets of IGF2BP1 have been identified to promote tumor proliferation and growth, migration, and invasion, some mRNAs have been indicated to at least indirectly suppress tumor growth and metastasis. Additionally, some ncRNAs were found to involve the regulation of tumor events by targeting IGF2BP1. Thus, bearing in mind the described limitation of post-transcriptional modulation of gene expression mediated by RNA-binding proteins (RBPs), miRNAs, and lncRNAs in determining fate of tumorigenesis, we in the following review recent findings on the expression of IGF2BP1 in cancers and focus on the interconnectivity of IGF2BP1 with its targeted mRNAs or ncRNAs.

IGF2BP1 was found to be commonly and significantly upregulated in almost all tumor cell lines and tumor tissues with a range from 17 to 72.4% of incidence (Table 1) and act as an essential oncogene that promotes the stability, localization, and translation of cancer-related mRNA targets by the regulation of some lncRNAs and miRNAs (Table 2). Almost all of the reported miRNAs including miR-491-5p, miR-625, miR-98-5p, miR-9, miR-196b, miR-1275, miR-708, let-7 miRNA family, miR-150, miR-506, miR-873, miR-140-5p, and miR-124-3p are downregulated in corresponding cancer samples, whereas all of the three lncRNAs including HCG11, GHET1, and THOR are highly expressed in corresponding solid tumors. As shown in Fig. 2, upregulation of the three lncRNAs can elevate IGF2BP1 level, while upregulation of the miRNAs could repress IGF2BP1 expression. HCG11 can upregulate the activity of IGF2BP1 and thus leads to activation of MAPK signaling [40]. GHET1 may enhance he stability of c-Myc mRNA and expression by physically associating with IGF2BP1 and elevating the physical interaction between c-Myc mRNA and IGF2BP1 [75]. However, THOR indirectly controls the levels of IGF2BP1-target mRNAs including KRAS, IGF2, CD44, PABPC1, ACTB, GLI1, MYC, CTNNB1, MAPT, PPP1R9B, PTEN, BTRC, and H19. Nearly the levels of all those targets are reduced in both MM603 and H1299 cells with THOR knock down similar to that with IGF2BP1 knockdown, whereas elevates with overexpression of THOR in SKMEL5 and H1437 cells [29]. Taken together, those findings show that lncRNAs HCG11, GHET1, and THOR dysregulate the expression of IGF2BP1-downstream targets by modulating IGF2BP1. Similarly, the miRNAs in tumors target the 3′-UTR of IGF2BP1, by which they regulate the expression of IGF2BP1-downstream targets and affect tumor cell proliferation, migration and invasion, and growth.

In the context of modulation by miRNAs/lncRNAs (Fig. 2), aberrant upregulation of IGF2BP1 promotes the expression of c-MYC and MKI67, as well as CD44 and PTEN [41, 44, 84, 101]. By stabilizing c-MYC and MKI67 transcripts, IGF2BP1 enhances tumor cell proliferation and growth. Interestingly, c-Myc and IGF2BP1 each constitutes a potential feedback mechanism to reciprocally regulate expression of the other. Additionally, IGF2BP1 prevents CD44 and PTEN mRNA turnover, consequently enhances CD44 expression, and induces the formation of invadopodia and therefore may promote tumor cell migration and invasiveness [84]. The elevation of PTEN inhibits PIP3/PIP2 ratios and then interferes with the activation of RAC1, which enhances cell polarization, and thus, this contributes to directed tumor cell migration as well as tumor invasion. However, the increased RGS4 expression enhanced by IGF2BP1 depresses tumor cell migration and invasion [17]. Although the expression levels of PTGS2, ACTB, and MAPK4 are reduced by IGF2BP1 interfering with their mRNA translation, there are different results on tumor events. Reduced PTGS2 decreases the suppression of tumor cell apoptosis and the promotion of tumor cell invasion [17]. The inhibition of MAPK4 antagonizes MK5-directed phosphorylation of HSP27. PHSP27 at both residues induces the degradation of oligomers and increases the sequestering of actin monomers by the phosphorylated protein [101]. The reduced ACTB also decreases G-actin levels. This shift in the cellular G-/F-actin equilibrium contributes to cell adhesion and actin dynamics, and finally promotes cell migration velocity [102]. In addition, promoter methylation of IGF2BP1 may silence its expression and subsequently modulates the downstream biological pathways [81, 82]. Taken together, we can deduce a conclusion that IGF2BP1 plays an important role in the occurrence of tumor events and shows a degree of tumor suppressor effect. However, determining whether the occurrence of tumor events or not partly depends on IGF2BP1 targeting different cancer-related mRNAs, although the driving factors that contribute to this gene to choose different mRNA targets are unclear. It is worth noting that the methylation events of the promoter in IGF2BP1 at least involve the selected process because high promoter methylation of IGF2BP1 was demonstrated to block β-catenin binding to the IGF2BP promoter, leading to inactivation of the gene and enhancing proliferation and migration of metastatic breast cancer cells (Fig. 1d) [54, 103]. Additionally, the selective manner may partially result from the competitive combination on m⁶A-containing mRNAs between IGF2BP1 and other m⁶A-binding proteins such as YTHDFs, since they could read the same m⁶A regions of mRNAs while determining a totally different fate of m⁶A-containing mRNAs (Fig. 1c).

As per the tumor-promotion role of IGF2BP1 in most of cancers, it seems that IGF2BP1 and its targeted transcripts could be attractive anticancer drug targets; however, small molecule inhibitors of IGF2BP1 and other cancer-related mRNA stabilizing proteins, as well as the upstream ncRNAs, are little known [8]. Note worthily, a small molecule, BTYNB, might function as a potential therapeutics by inhibiting cell proliferation of IGF2BP1-positive cancer cells without effect in IGF2BP1-negative cells. BTYNB may restrain binding of IGF2BP1 to the coding region stability determinant of c-Myc mRNA and downregulate several mRNA transcripts including c-Myc, β-TrCP1, and eEF2 both in IGROV-1 and SK-MEL2 cancer cells, as well as decrease activation of nuclear transcriptional factors-kappa B (NF-κB). Moreover, it also selectively reduces the levels of other cancer-related IGF2BP1 mRNA targets including CALM1, CDC34, COL5A, and BTRC, similar to the effect by RNAi knockdown of IGF2BP1 both in IGROV-1 and SK-MEL2 cancer cells [104].

IGF2BP1 is broadly and highly expressed in embryonic and tumor tissues. The bulk of correlative studies describing enhanced expression or de novo synthesis of IGF2BP1 in human cancer and animal model provide strong evidence that IGF2BP1 serves important roles in controlling embryonic development, as well as functions as an oncogenic factor in most of cancers. In most of the cancers, IGF2BP1 enhances tumor cell proliferation, survival, adhesion-independent growth and invasion, and chemo-resistance. The upregulated expression of IGF2BP1 is associated with poor overall survival and metastasis in multiple cancers. However, the tumor-suppressive role of IGF2BP1 has been observed in breast cancer and colon stromal cells. At least in breast cancer, it is confirmed that IGF2BP1 inhibits tumor cell growth and invasion. IGF2BP1 has both tumor-driving and tumor-suppressive roles in cancers in a context-based manner.

Although the origin of difference between oncogenic and tumor-suppressive roles is unclear, the apparently contradictory functions of IGF2BP1 may result from tumor cells arrived from different origins or conditions with different responses to IGF2BP1 expression. In this context, it seems that some undefined driver factors contribute to IGF2BP1 selectively binding and regulating its mRNA targets and lead to either development or suppression for tumors. Though the mechanism of IGF2BP1 selectively binding to its mRNA targets is unclear, epigenetic modifications of IGF2BP1 at least are involved in the process. In addition, the selective manner may partially result from the competitive combination on m⁶A-containing mRNAs between IGF2BP1 and other m⁶A-binding proteins such as YTHDFs.

The reasonable inferences about the functions of IGF2BP1 in cancer biology can be confirmed in the future by in vivo and in vitro studies, according to specific origins or conditions of the tumor cells. Additionally, the emerging miRNAs and lncRNAs that target and regulate IGF2BP1 have been ascertained to involve IGF2BP1-mediated mRNA modulation. Therefore, more interconnectivity between IGF2BP1 and its targeted mRNAs or ncRNAs can be explored in contexts related to specific cancer conditions. NcRNA-IGF2BP1-mRNA target-axes may be worth of more studies as their target potential for cancer therapy. Though currently available inhibitors of ncRNA-IGF2BP1-mRNA target-axes are limited, an inhibitor of IGF2BP1 binding to targeted mRNAs, BTYNB, has showed therapeutic potential by inhibiting cell proliferation of IGF2BP1-positive cancer cells.