1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

Selected long noncoding RNAs (lncRNAs) have been shown to play important roles in carcinogenesis. Although the cellular functions of these transcripts can be diverse, many lncRNAs regulate gene expression. In contrast, factors that control the expression of lncRNAs remain largely unknown. Here we investigated the impact of RNA binding proteins on the expression of the liver cancer-associated lncRNA HULC (highly up-regulated in liver cancer). First, we validated the strong up-regulation of HULC in human hepatocellular carcinoma. To elucidate posttranscriptional regulatory mechanisms governing HULC expression, we applied an RNA affinity purification approach to identify specific protein interaction partners and potential regulators. This method identified the family of IGF2BPs (IGF2 mRNA-binding proteins) as specific binding partners of HULC. Depletion of IGF2BP1, also known as IMP1, but not of IGF2BP2 or IGF2BP3, led to an increased HULC half-life and higher steady-state expression levels, indicating a posttranscriptional regulatory mechanism. Importantly, HULC represents the first IGF2BP substrate that is destabilized. To elucidate the mechanism by which IGF2BP1 destabilizes HULC, the CNOT1 protein was identified as a novel interaction partner of IGF2BP1. CNOT1 is the scaffold of the human CCR4-NOT deadenylase complex, a major component of the cytoplasmic RNA decay machinery. Indeed, depletion of CNOT1 increased HULC half-life and expression. Thus, IGF2BP1 acts as an adaptor protein that recruits the CCR4-NOT complex and thereby initiates the degradation of the lncRNA HULC. Conclusion: Our findings provide important insights into the regulation of lncRNA expression and identify a novel function for IGF2BP1 in RNA metabolism. (Hepatology 2013;58:1703–1712)


cyclic adenosine monophosphate


cAMP responsive element binding protein


green fluorescent protein


hepatocellular carcinoma


highly up-regulated in liver cancer


IGF2 mRNA-binding protein




noncoding RNA


quantitative reverse transcription, polymerase chain reaction


small interfering RNA.

Hepatocellular carcinoma (HCC) is the fifth most common cancer in the world and the third most common cause of cancer mortality.[1] Representing a multifactorial genetic and epigenetic disease with a complex etiology, the major cause of HCC is long-term liver injury caused by, e.g., infection with hepatitis B or C virus, alcoholic liver disease, aflatoxin exposure, or a variety of inherited metabolic diseases.[2] Although numerous small and high-dimensional profiling analyses have been performed in human HCC (reviewed[3]), the molecular mechanisms and factors involved in liver carcinogenesis are still not fully understood.

Recently, it has been uncovered that the human genome encodes much more information than previously anticipated. The vast majority (70%-90%) of the human genome sequence is pervasively transcribed into RNA, while only a small fraction (∼2%) contains information for protein-coding genes.[4-7] Thus, the largest fraction of the human genome encodes ncRNAs (noncoding RNAs). Some of these transcripts are highly conserved, show regulated and tissue-specific expression, and exert critical functions in the cell.[8-13] Mechanistically, some ncRNAs were shown to have a strong impact on the regulation of gene expression[14-18] or posttranscriptional processing.[19, 20] Moreover, several ncRNAs are deregulated in human diseases including cancer, influence disease onset as well as progression, or can be of prognostic value.[21-23] Thus, studying long ncRNA expression, regulation, and function in human liver cancer is essential to fully understand the underlying molecular mechanisms.

The long ncRNA HULC (highly up-regulated in liver cancer) is one of the first strongly overexpressed noncoding transcripts to be identified in human HCC.[24] The HULC gene is located on chromosome 6p24.3 and is conserved in primates. Transcription of HULC yields an ∼500 nt long, spliced and polyadenylated ncRNA that localizes to the cytoplasm where it has been reported to associate with ribosomes.[24] HULC expression has been described to be regulated by the transcription factor CREB (cyclic adenosine monophosphate [cAMP] responsive element binding protein) in Hep3B cells.[25] In addition, the HBx protein has been linked to the activation of HULC expression in HepG2 cells by way of interaction with CREB.[26] Elevated HULC levels in HBx expressing HepG2 cells induce a higher proliferation rate and tumor growth and lead to a down-regulation of the tumor suppressor p18. Moreover, HULC might function as a microRNA (miRNA) sponge for miRNA-372 and thereby could regulate gene expression at a posttranscriptional level.[25] While it is clear that HULC plays an important role in liver carcinogenesis and acts as an oncogenic ncRNA, the regulatory mechanisms controlling HULC expression are largely unknown.

Our aim was to determine the regulatory mechanisms that control this ncRNA, highly expressed in HCC. We hypothesized that RNA binding proteins could have an impact on HULC expression and set up an RNA affinity purification assay to identify specific protein interaction partners of HULC. In this study, we identified the IGF2BP (IGF2 messenger RNA [mRNA]-binding protein) family of RNA binding proteins as specific interaction partners of HULC in human liver cancer cells. HULC was discovered as the first IGF2BP substrate that is not stabilized or translationally regulated, but destabilized by way of CNOT1-mediated deadenylation recruited by IGF2BP1.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information
Tumor Material and Patient Characteristics

Sixty human HCCs were analyzed for HULC expression using microarray analysis. Median age at surgery was 57 years (range 16-78), and the male/female ratio was 4:1. All diagnoses were confirmed by histological reevaluation, and use of the samples was approved by the local Ethics Committee. The cohort contained a balanced repertoire of relevant underlying etiologies: hepatitis B virus (HBV) (n = 15), HCV (n = 12), alcohol (n = 10), cryptogenic (presumably mostly nonalcoholic fatty liver disease [NAFLD]; n = 19) or genetic hemochromatosis (n = 3). The patients' characteristics are shown in Supporting Table 1.

In Vitro Transcription

For in vitro transcription, the Megascript T7 kit (Life Technologies, Carlsbad, CA) was used according to the manufacturer's recommendations. Briefly, 1 μg linearized plasmid template was used and reactions were incubated for 16 hours in the presence or absence of Biotin-16-UTP (Epicentre, Madison, WI). The ratio between UTP and Biotin-16-UTP was 20:1. The reaction was stopped by addition of 1 μL Turbo-DNase. RNA was precipitated with LiCl. RNA integrity and size were controlled using agarose gel electrophoresis.

RNA Affinity Purification

Beads were preblocked with 1 mg/mL BSA (Roche), 0.2 mg/mL yeast tRNA (Roche), and 0.2 mg/mL Glycogen (Carl Roth, Karlsruhe, Germany) in low salt wash buffer (20 mM Hepes, pH 7.9; 100 mM KCl; 10 mM MgCl2; 0.01% NP40; 1 mM DTT) before addition of RNA. RNA was incubated with 50 μL Streptavidine-Sepharose beads (GE Healthcare, Little Chalfont, UK) in 500 μL HS-WB300 (20 mM Hepes, pH 7.9; 300 mM KCl; 10 mM MgCl2; 0.01% NP40; 1 mM DTT) for 4 hours. Unbound RNA was washed away with 3× 1 mL HS-WB400 (20 mM Hepes, pH 7.9; 400 mM KCl; 10 mM MgCl2; 0.01% NP40; 1 mM DTT). Cytoplasmic cell extract (2-3 mg) was added and incubated overnight at 4°C. The next day the extract was removed and beads were washed 6 times with 1 mL HS-WB400. Beads were resuspended in 50 μL 6 M urea; 1 mM DTT; 0.01% NP-40 and incubated at room temperature for 30 minutes in a shaking block at 900 rpm. Then the supernatant was transferred into a new tube and proteins were precipitated with 5 volumes of prechilled acetone for 1 hour at −20°C. Proteins were pelleted by way of centrifugation at 13,000g at room temperature. Pellets were washed twice with 1 mL 80% ethanol, dried for 5 minutes, and resuspended in 20 μL protein sample buffer.

RNA Stability Analysis

HepG2 cells were transfected with small interfering RNAs (siRNAs) as stated above. After 48 hours, alpha-amanitine (AppliChem, Darmstadt, Germany) was added (10 μg/mL f.c.) and cells were harvested at the indicated timepoints. All experiments were done in biological triplicates.

RNA-Protein-Interaction Analysis

Detailed methodological information on RNA affinity purification and coimmunoprecipitation experiments including a protocol for cytoplasmic extract preparation, RNA in vitro synthesis and biotin labeling, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and Coomassie staining as well as IGF2BP cloning can be found in the Supporting Experimental Procedures.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information
HULC Is Overexpressed in HCC and Correlates With Staging and Grading

HULC was shown to be overexpressed in human HCCs using an HCC-specific cDNA microarray.[25] To validate these previous findings in an independent, larger patient cohort, we performed unbiased microarray analysis of 60 HCC and 7 normal liver samples using the Agilent SurePrint G3 Human Gene Expression array. We identified HULC as the second most highly up-regulated nonprotein-coding gene in HCC (fold change = 6.51, P = 3.3 × 10−5, t test) (Fig. 1A). Only the ERBB2 pseudogene showed a stronger up-regulation in human HCCs (fold change = 8.23, P = 4.6 × 10−7; data not shown). We confirmed the overexpression of HULC in HCC by qRT-PCR (quantitative reverse transcription-polymerase chain reaction) analysis in a subset of 34 tumor samples and 6 normal livers (Fig. 1B) significantly correlating with the microarray data (R = 0.452, Spearman). The respective patient data of this subset are in Table 1. The relative expression level of HULC, as determined by qRT-PCR, was about 8-fold higher in HCC samples than in normal liver tissue (Fig. 1C). Interestingly, we detected a significantly higher expression level of HULC in low-grade and low-stage tumors (Fig. 1D). However, HULC expression did not correlate with age, sex, tumor size, or hemangiosis (Table 1). HULC expression was previously shown to be induced by the viral HBx protein[26] and increased in HBV-producing cells.[27] Thus, we tested whether HULC levels correlated with different tumor etiologies (Table 1). However, there was no significant correlation between HULC expression and HBV or HCV infection (Mann-Whitney U, P = 0.078 (HBV versus non-HBV); P = 0.220 (HCV versus non-HCV)), the average HULC level was even lower in HBV-infected patients than in other HCC samples (Fig. 1E).


Figure 1. Differential expression of HULC in HCC and normal liver patient samples. (A) Microarray analysis of 60 HCC and 7 normal liver samples showing significantly increased HULC expression in HCC. The horizontal line represents the mean of nontumor samples. (B) Validation of differential HULC expression using qRT-PCR. PPIA was used as the reference gene. (C) Boxplot analysis of differential HULC expression in HCC and normal liver showing a significant higher expression in tumor samples (P < 0.001; Mann-Whitney U test). (D) Correlation analysis of HULC expression with tumor grade and stage. (E) Correlation analysis of HULC expression with tumor etiology.

Download figure to PowerPoint

Table 1. Correlation Analysis of HULC Expression and Patient Characteristics
AgeMedian (range)55 (16-78)0.568a
  1. a

    Pearson correlation;

  2. b

    Mann-Whitney U test;

  3. c

    Kruskal-Wallis test.

 Genetic hemochromatosis2
 HBV vs. non-HBV 0.078b
 HCV vs. non-HCV 0.220b
 Alcohol vs. nonalcohol 0.190b
Tumor sizeMedian (range)5 (2-29)0.310a
GradingWell-differentiated HCC50.012c
 Moderately differentiated HCC23
 Poorly differentiated HCC5
 Well vs. poorly differentiated 0.016b
 Moderately vs. poorly differentiated 0.016b
 T1 vs. T2 0.015b
 T1 vs. T3 0.027b
 T1 vs. T4 0.034b
Vascular invasionPresent90.007b
HULC Interacts With IGF2 mRNA Binding Proteins

After transcription, an ncRNA will likely associate with proteins to form a ribonucleoprotein complex that will govern ncRNA stability, degradation, and function. Thus, posttranscriptional regulators could interact with HULC and contribute to its regulation and consequently its functional impact. Therefore, we aimed at the identification of interacting proteins as potential regulators using an RNA affinity purification approach. An overview of the method is given in Fig. 2A. We used cytoplasmic extracts prepared from Huh7 HCC cells and incubated these with a 500 nt long, in vitro transcribed and biotinylated HULC RNA. An RNA molecule of the same length but unrelated in sequence was used as a negative control. Proteins associated with HULC or the control RNA were eluted, separated on a polyacrylamide gel, and visualized with sensitive Coomassie blue staining (Fig. 2B). Multiple proteins with an observed molecular weight of ∼70 kDa were specifically pulled down with HULC (Fig. 2B, box). Subsequent mass spectrometry analyses identified several protein interaction partners. Among the top five candidate proteins, we reproducibly identified three members of the IGF2 mRNA-binding protein family, namely IGF2BP1, IGF2BP2, and IGF2BP3, also known as IMP1, IMP2, and IMP3 (Supporting Table 5). Specific binding of IGF2BP1, IGF2BP2, and IGF2BP3 was confirmed by western blot analysis (Fig. 2C). We validated the interaction between IGF2BPs and HULC also in HepG2 cells (Fig. 2D). HnRNP A1, an unrelated RNA binding protein, and Vinculin, a protein associated with the cytoskeleton, were included as controls for specificity.


Figure 2. Identification of IGF2 mRNA-binding proteins as interaction partners of HULC. (A) Protein interaction partners were detected using RNA affinity purification, incubating cell lysates with in vitro transcribed, biotinylated RNA, and analyzing differential bands with mass spectrometry. (B) Representative SDS-PAGE gel after Coomassie blue staining shows a prominent band at about 70 kDa, representing mainly IGF2 mRNA-binding proteins (IGF2BPs, see also Supporting Table 4). (C) Validation of specific binding of IGF2BPs to biotinylated HULC in Huh7 by western blotting. (D) Validation of specific binding of IGF2BPs to biotinylated HULC in HepG2 cells by western blotting. (E) Representative western blot analysis after FLAG-immunoprecipitation. HepG2 cells were transiently transfected for 72 hours with FLAG-tagged GFP (negative control), IGF2BP1, IGF2BP2, or IGFBP3. (F) Analysis of copurified RNA and respective enrichment as determined by qRT-PCR after anti-FLAG immunoprecipitation validating specific binding of HULC RNA to IGF2BPs. All immunoprecipitation experiments were done in biological replicates (n = 3). Given is the mean and the respective standard error of the mean (±SEM).

Download figure to PowerPoint

As an independent approach to verify the interaction between HULC and IGF2BPs in vivo, we performed RNA immunoprecipitation assays. FLAG-tagged IGF2BP1, IGF2BP2, IGF2BP3, or GFP (green fluorescent protein; negative control) were transiently overexpressed in HepG2 cells and immunoprecipitated with an anti-FLAG antibody (Fig. 2E). After isolation of the copurifying RNA, the enrichment of selected transcripts was measured by way of qRT-PCR. Thereby, we confirmed the specific enrichment of both HULC and a bona fide target of IGF2BPs, IGF2 mRNA (Fig. 2F). No enrichment of HULC was seen in GFP pull downs. The highly abundant 5.8S rRNA (negative control) was not enriched in any of the purifications. Thus, we identified the IGF2 mRNA binding proteins as specific interaction partners of HULC. Furthermore, we characterized the interaction between HULC and IGF2BP1 in more detail and could show that also endogenous, nontagged IGF2BP1 specifically bound to HULC (Supporting Fig. 1A). To identify the site of interaction, we performed an in vitro binding assay using recombinant human IGF2BP1 and in vitro transcribed HULC full-length or fragmented RNA (Supporting Fig. 1B,C). The assay revealed a direct and specific binding of IGF2BP1 to multiple sites across the noncoding transcript (Supporting Fig. 1D).

IGF2BP1 Specifically Regulates HULC Expression and Stability

IGF2BPs are well-known RNA binding proteins that were shown to regulate translation, localization, or stability of their target RNAs.[28-34] Specifically, IGF2BP1 stabilizes MYC, MDR1, and PTEN mRNAs.[35-37] To determine whether HULC expression was controlled by IGF2BPs, we specifically depleted IGF2BP1, IGF2BP2, or IGF2BP3 from HepG2 cells using siRNAs (small interfering RNAs) (Fig. 3A,B). The knockdowns were efficient as analyzed by qRT-PCR (Fig. 3A), and specific to each of the IGF2BP family members as shown by western blot analysis (Fig. 3B). Interestingly, the knockdown of each IGF2BP alone led to an enhanced HULC expression. The strongest increase was observed after IGF2BP1 depletion, which was highly significant compared to control siRNA or IGF2BP2 and IGF2BP3 siRNA transfections (Fig. 3C). To distinguish between a transcriptional and a posttranscriptional mechanism, we specifically blocked RNA Polymerase II transcription with alpha-amanitin. This experiment revealed a strong impact of IGF2BP1 on HULC RNA stability (Fig. 3D). While the half-life of HULC was between 7.0 hours and 7.7 hours after control siRNA and IGF2BP2 or IGF2BP3 siRNA transfection, the half-life was almost twice as long when IGF2BP1 was depleted (13.3 hours ± 1.5 hours). The stabilizing effect was also seen after IGF2BP1 depletion with a second independent siRNA (Supporting Fig. 2A). Moreover, transient overexpression of IGF2BP1 in HepG2 significantly decreased HULC expression levels (Supporting Fig. 2B,C).


Figure 3. Impact of IGF2BP1 depletion on HULC expression and stability. (A) IGF2BP RNA expression in HepG2 cells 72 hours after siRNA transfection as determined by qRT-PCR. Shown is the remaining expression level relative to the respective IGF2BP level in cells transfected with control siRNA. Shown is the mean of at least three independent experiments (±SEM). (B) Validation of the specificity of individual siRNAs against IGF2BP1, IGF2BP2, and IGF2BP3 at the protein level as determined by western blot. Shown is a representative western blot of three independent experiments. Vinculin expression was detected to verify equal loading. (C) HULC expression after IGF2BP knockdown. Depletion of IGF2BP1 had the strongest effect and significantly increased steady-state HULC expression as determined by qRT-PCR. Shown is the mean of at least three independent experiments (±SEM). Expression was normalized to PPIA and the siControl sample was set to 1.0. (D) HULC stability analysis in HepG2 cells after alpha-amanitin treatment. Cells were transfected with siRNAs against IGF2BPs and 48 hours later, a time course for RNA stability was started by adding the RNA-Polymerase II inhibitor. Cells were harvested at the indicated timepoints. Expression levels were normalized to “0 h” and RNU6 was used as the reference gene. HULC half-life nearly doubled upon IGF2BP1 knockdown compared to knockdown of IGF2BP2, IGF2BP3, or control siRNA. Shown is the mean of at least three independent experiments (±SEM).

Download figure to PowerPoint

This suggested that IGF2BP1, but neither IGF2BP2 nor IGF2BP3, regulated HULC posttranscriptionally. To our knowledge, HULC was the first IGF2BP1 target RNA that was destabilized by this protein. Hence, we wanted to elucidate the mechanism of HULC destabilization by IGF2BP1.

IGF2BP1 Interacts With CNOT1, a Component of the CCR4-NOT Deadenylase Complex

Intracellular RNA degradation occurs by way of two major pathways starting from the 5′ end or the 3′ end of the RNA, respectively, and could involve miRNAs.[38] HULC was previously shown to be part of a negative feedback loop acting as a sponge for miRNA-372.[25] Thus, we tested whether IGF2BP1 depletion influences mature miR-372 expression in HepG2 cells, but we could not detect a significant down-regulation of miR-372 (Supporting Fig. 3A). In addition, we could not observe a down-regulation of HULC upon miR-372 overexpression in three different liver cancer cell lines (Supporting Fig. 3B). These findings implicate an alternative, miR-372-independent regulatory mechanism.

Hence, we hypothesized that IGF2BP1 might associate with components of the RNA decay machinery to mediate RNA degradation. To pursue this hypothesis, we transfected HepG2 cells with FLAG-tagged IGF2BP1 or GFP as a control. After anti-FLAG immunoprecipitation, we tested whether IGF2BP1 interacted with XRN1 or CNOT1 by western blot analysis (Fig. 4A). XRN1, the major cytoplasmic 5′-3′-exonuclease, did not copurify with IGF2BP1. In contrast, CNOT1 showed specific binding to IGF2BP1, but not to GFP (Fig. 4A). CNOT1 is the scaffold protein of the CCR4-NOT complex, an important deadenylase responsible for poly(A) tail shortening and inducing 3′-5′-decay of numerous RNAs in the cytoplasm.[39] Thus, IGF2BP1 interacted with a central component of the RNA decay machinery.


Figure 4. IGF2BP1 interacts with components of the RNA decay machinery. (A) Overexpression of FLAG-IGF2BP1 or FLAG-GFP for 72 hours in HepG2 cells. Subsequent immunoprecipitation revealed a specific interaction of IGF2BP1 with CNOT1 but not XRN1. Shown is a representative western blot of three independent experiments. (B) Efficient knockdown of CNOT1 at RNA (upper panel) and protein level (lower panel) was achieved in HepG2 cells using two independent siRNAs. Expression of GAPDH is used to verify equal loading of the western blots. Shown is a representative western blot of three independent experiments. (C) Depletion of CNOT1 increased HULC expression as measured by way of qRT-PCR. PPIA was used as the reference gene in qRT-PCR. Shown is the mean of at least three independent experiments (±SEM). (D) Analysis of HULC RNA stability using alpha-amanitin treatment after CNOT1 knockdown in HepG2 cells. HULC RNA half-life was drastically increased by CNOT1 knockdown compared to control siRNA transfection. Shown is the mean of at least three independent experiments (±SEM). RNU6 was used as a stable reference gene. Expression is shown relative to control siRNA transfection and “0 h.” (E) Proposed model of HULC expression and stability regulation. IGF2BP1 binds HULC and recruits it to the CCR4-NOT1 complex, initiating deadenylation and degradation by way of the 3′-5′ degradation pathway.

Download figure to PowerPoint

Depletion of CNOT1 Stabilizes HULC

Interaction with CNOT1 might be crucial for the destabilizing effect of IGF2BP1 on HULC. Consequently, depletion of CNOT1 should increase the half-life and steady-state expression level of HULC. To test this hypothesis, we depleted CNOT1 in HepG2 cells with two independent siRNAs and analyzed the CNOT1 expression both at the RNA and protein level. The knockdown was highly effective with both siRNAs (Fig. 4B). In both cases, the steady-state levels of HULC were strongly elevated (>2-4-fold) after CNOT1 depletion (Fig. 4C). SiRNA 1, which was slightly more effective in reducing CNOT1 levels (Fig. 4B), also had a greater effect on HULC expression (Fig. 4C). Furthermore, blocking transcription after depletion of CNOT1 revealed a strong impact of CNOT1 on HULC RNA stability (Fig. 4D). The half-life of HULC was significantly prolonged with both siRNAs targeting CNOT1 (up to 25.9 hours ± 7.1 hours) compared to control siRNA (5.6 hours ± 0.9 hours). Interestingly, the depletion of IGF2BP1 and CNOT1 simultaneously had no additive effect on HULC up-regulation, indicating that both proteins are mechanistically linked to each other (Supporting Fig. 4). Thus, we propose a model in which HULC expression is negatively regulated by way of binding to IGF2BP1. By associating directly or indirectly with CNOT1, IGF2BP1 recruits the CCR4-NOT deadenylase complex onto its RNA substrate (Fig. 4E).


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

To understand the mechanisms underlying hepatocarcinogenesis, a large number of genetic and epigenetic profiling studies had been conducted.[3] These studies mainly focused on the role of protein-coding genes and rarely included long, nonprotein-coding transcripts. Our study validated the significant up-regulation of HULC, a liver-enriched lncRNA in human HCC samples. Moreover, we showed for the first time that the expression of HULC is significantly higher in low-stage and low-grade tumors, which points towards a functional role of HULC in the early steps of tumor development. Chronic inflammation, caused, e.g., by viral infections or alcohol abuse, is a critical factor that triggers liver carcinogenesis. In our analysis, we could not detect a positive correlation with HBV or HCV infections. This is surprising in light of recent reports that established a link between HULC expression and HBV status, and showed that the HBx protein up-regulates HULC by way of CREB.[26, 27] Based on the necessarily limited size of every patient cohort, we cannot formally exclude the possibility of a correlation with viral infections, but we do not see any trend towards HULC induction in primary patient samples infected with HBV. Previously, no direct association with HBV or HCV infection in patient samples was shown, but only cell culture models were used to establish this connection. Future studies with larger patient cohorts may further detail the correlations with different etiologies.

After confirming the high up-regulation of HULC in liver cancer, we wanted to explore the regulation of this transcript in human liver cancer cells. First, we could not verify the previously described regulation of HULC by miR-372 in three different liver cancer cell lines. Thus, we performed RNA affinity purification experiments to identify RNA-binding proteins that bind and potentially regulate HULC posttranscriptionally. Through this approach, we identified a novel and unexpected function of the well-known RNA-binding protein IGF2BP1. IGF2BP1 acts as a trans-acting factor that represses HULC stability and expression. Moreover, IGF2BP1 associates with CNOT1 and thereby brings HULC into close proximity to the CCR4-NOT deadenylase complex, which initiates RNA degradation from the 3′ end.[39] After the initial deadenylation of HULC, its final degradation may occur by way of the 3′-5′- or the 5′-3′-exonucleolytic pathways.

Surprisingly little is known about the general role of RNA decay in the context of cancer. While factors such as miRNAs and AU-rich element binding proteins are known to specifically target mRNAs for degradation, we are still far from a comprehensive understanding of the network that controls the stability of individual RNAs. Here, we discovered that IGF2BP1 might act as an adaptor protein that helps to destabilize HULC in human liver cancer cells. However, the regulatory mechanisms governing the expression, activity, localization, and RNA binding capacity of IGF2BP1 are mostly unknown. Derived from PAR-CLIP data to identify RNA substrates of the IGF2BP family, a potential RNA recognition consensus element has been proposed.[40] This short CAUH (H = A, U, or C) motif is present in HULC RNA 10 times, distributed over the whole transcript and might represent a part of the binding site for the IGF2BPs that can associate as homo- or heterodimers (see Supporting Fig. 1). However, this very short element lacks specificity—stochastically, it should be found every 85 nucleotides—so that additional, so far undiscovered bindings motifs are likely.[41, 42] It will be of future interest to elucidate the underlying control mechanisms that define whether an RNA is bound, stabilized, or destabilized by IGF2BP1 and which signaling pathways induce, control, and limit the interaction and subsequent RNA degradation of its targets, notably of HULC. This is especially important since we did not find any negative correlation between IGF2BP1 and HULC expression at the mRNA level (data not shown). Hence, the regulation of HULC in primary liver cancer might be independent of IGF2BP1-mediated posttranscriptional regulation and mainly controlled at the transcriptional level—or so far undetermined inhibitory mechanisms (e.g., posttranslational modifications) might affect the activity, localization, or binding of IGF2BP1 proteins to HULC transcripts in primary human HCC.

IGF2BP1 is a known oncofetal protein linked to several malignant human diseases: Its expression is induced in human malignant melanomas or colorectal carcinomas with activated WNT/β-catenin/TCF signaling.[43, 44] High IGF2BP1 expression is a poor prognostic marker in high-stage and high-grade ovarian carcinomas and lung cancers.[45-47] This study has unraveled that IGF2BP1 can also destabilize client transcripts. Hence, it opened up a new field of potential IGF2BP targets and IGF2BP-mediated silencing effects. Future studies may determine whether other IGF2BP1-bound transcripts, both coding and noncoding, are destabilized and degraded by way of the CNOT1 pathway in HCC or other tumor entities. Our study has revealed a novel mechanism that will help to fully establish the function of IGF2BP1 as a gene regulator in human cancer.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

The authors thank Drs. Dirk Ostareck-Lederer and Peter Angel for helpful discussions and Dr. Markus Landthaler for providing IGF2BP1 plasmids. We thank Marion Stentrup and Andreas Heim for experimental support. We thank the DKFZ Genomics and Proteomics Core facility, especially Drs. Martina Schnölzer and Tore Kempf, for excellent performance of mass spectrometry analysis.

Author Contributions

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

M.H., T.G., and S.D. designed and executed the experiments, analyzed the data, and wrote the article. H.U., M.G., and E.F. performed experiments and analyzed the data. S.O. and G.S. contributed to design of experimentation and helped with interpretation of results. B.S. and R.G. performed microarray expression analysis. T.L., K.B., and P.S. provided tissue samples and helped with data interpretation.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information
  • 1
    El-Serag HB, Rudolph KL. Hepatocellular carcinoma: epidemiology and molecular carcinogenesis. Gastroenterology 2007;132:2557-2576.
  • 2
    Thorgeirsson SS, Grisham JW. Molecular pathogenesis of human hepatocellular carcinoma. Nat Genet 2002;31:339-346.
  • 3
    Breuhahn K, Gores G, Schirmacher P. Strategies for hepatocellular carcinoma therapy and diagnostics: lessons learned from high throughput and profiling approaches. Hepatology 2011;53:2112-2121.
  • 4
    Bertone P, Stolc V, Royce TE, Rozowsky JS, Urban AE, Zhu X, et al. Global identification of human transcribed sequences with genome tiling arrays. Science 2004;306:2242-2246.
  • 5
    Birney E, Stamatoyannopoulos JA, Dutta A, Guigo R, Gingeras TR, Margulies EH, et al. Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project. Nature 2007;447:799-816.
  • 6
    Carninci P, Kasukawa T, Katayama S, Gough J, Frith MC, Maeda N, et al. The transcriptional landscape of the mammalian genome. Science 2005;309:1559-1563.
  • 7
    Kapranov P, Cheng J, Dike S, Nix DA, Duttagupta R, Willingham AT, et al. RNA maps reveal new RNA classes and a possible function for pervasive transcription. Science 2007;316:1484-1488.
  • 8
    Calin GA, Liu CG, Ferracin M, Hyslop T, Spizzo R, Sevignani C, et al. Ultraconserved regions encoding ncRNAs are altered in human leukemias and carcinomas. Cancer Cell 2007;12:215-229.
  • 9
    Chen LL, Carmichael GG. Long noncoding RNAs in mammalian cells: what, where, and why? Wiley Interdiscip Rev RNA 2010;1:2-21.
  • 10
    Eissmann M, Gutschner T, Hammerle M, Gunther S, Caudron-Herger M, Gross M, et al. Loss of the abundant nuclear non-coding RNA MALAT1 is compatible with life and development. RNA Biol 2012;9:1076-1087.
  • 11
    Guttman M, Amit I, Garber M, French C, Lin MF, Feldser D, et al. Chromatin signature reveals over a thousand highly conserved large non-coding RNAs in mammals. Nature 2009;458:223-227.
  • 12
    Lipovich L, Johnson R, Lin CY. MacroRNA underdogs in a microRNA world: evolutionary, regulatory, and biomedical significance of mammalian long non-protein-coding RNA. Biochim Biophys Acta 2010;1799:597-615.
  • 13
    Ponting CP, Oliver PL, Reik W. Evolution and functions of long noncoding RNAs. Cell 2009;136:629-641.
  • 14
    Feng J, Bi C, Clark BS, Mady R, Shah P, Kohtz JD. The Evf-2 noncoding RNA is transcribed from the Dlx-5/6 ultraconserved region and functions as a Dlx-2 transcriptional coactivator. Genes Dev 2006;20:1470-1484.
  • 15
    Gutschner T, Hammerle M, Eissmann M, Hsu J, Kim Y, Hung G, et al. The non-coding RNA MALAT1 is a critical regulator of the metastasis phenotype of lung cancer cells. Cancer Res 201;73:1180-1189.
  • 16
    Martianov I, Ramadass A, Serra Barros A, Chow N, Akoulitchev A. Repression of the human dihydrofolate reductase gene by a non-coding interfering transcript. Nature 2007;445:666-670.
  • 17
    Rinn JL, Kertesz M, Wang JK, Squazzo SL, Xu X, Brugmann SA, et al. Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 2007;129:1311-1323.
  • 18
    Khalil AM, Guttman M, Huarte M, Garber M, Raj A, Rivea Morales D, et al. Many human large intergenic noncoding RNAs associate with chromatin-modifying complexes and affect gene expression. Proc Natl Acad Sci U S A 2009;106:11667-11672.
  • 19
    Beltran M, Puig I, Pena C, Garcia JM, Alvarez AB, Pena R, et al. A natural antisense transcript regulates Zeb2/Sip1 gene expression during Snail1-induced epithelial-mesenchymal transition. Genes Dev 2008;22:756-769.
  • 20
    Tripathi V, Ellis JD, Shen Z, Song DY, Pan Q, Watt AT, et al. The nuclear-retained noncoding RNA MALAT1 regulates alternative splicing by modulating SR splicing factor phosphorylation. Mol Cell 2010;39:925-938.
  • 21
    Faghihi MA, Modarresi F, Khalil AM, Wood DE, Sahagan BG, Morgan TE, et al. Expression of a noncoding RNA is elevated in Alzheimer's disease and drives rapid feed-forward regulation of beta-secretase. Nat Med 2008;14:723-730.
  • 22
    Gutschner T, Diederichs S. The hallmarks of cancer: a long non-coding RNA point of view. RNA Biol 2012;9:703-719.
  • 23
    Ji P, Diederichs S, Wang W, Boing S, Metzger R, Schneider PM, et al. MALAT-1, a novel noncoding RNA, and thymosin beta4 predict metastasis and survival in early-stage non-small cell lung cancer. Oncogene 2003;22:8031-8041.
  • 24
    Panzitt K, Tschernatsch MM, Guelly C, Moustafa T, Stradner M, Strohmaier HM, et al. Characterization of HULC, a novel gene with striking up-regulation in hepatocellular carcinoma, as noncoding RNA. Gastroenterology 2007;132:330-342.
  • 25
    Wang J, Liu X, Wu H, Ni P, Gu Z, Qiao Y, et al. CREB up-regulates long non-coding RNA, HULC expression through interaction with microRNA-372 in liver cancer. Nucleic Acids Res 2010;38:5366-5383.
  • 26
    Du Y, Kong G, You X, Zhang S, Zhang T, Gao Y, et al. Elevation of highly up-regulated in liver cancer (HULC) by hepatitis B virus X protein promotes hepatoma cell proliferation via down-regulating p18. J Biol Chem 2012;287:26302-26311.
  • 27
    Matouk IJ, Abbasi I, Hochberg A, Galun E, Dweik H, Akkawi M. Highly upregulated in liver cancer noncoding RNA is overexpressed in hepatic colorectal metastasis. Eur J Gastroenterol Hepatol 2009;21:688-692.
  • 28
    Atlas R, Behar L, Elliott E, Ginzburg I. The insulin-like growth factor mRNA binding-protein IMP-1 and the Ras-regulatory protein G3BP associate with tau mRNA and HuD protein in differentiated P19 neuronal cells. J Neurochem 2004;89:613-626.
  • 29
    Bell JL, Wachter K, Muhleck B, Pazaitis N, Kohn M, Lederer M, et al. Insulin-like growth factor 2 mRNA-binding proteins (IGF2BPs): post-transcriptional drivers of cancer progression? Cell Mol Life Sci 2012.
  • 30
    Doyle GA, Betz NA, Leeds PF, Fleisig AJ, Prokipcak RD, Ross J. The c-myc coding region determinant-binding protein: a member of a family of KH domain RNA-binding proteins. Nucleic Acids Res 1998;26:5036-5044.
  • 31
    Elcheva I, Goswami S, Noubissi FK, Spiegelman VS. CRD-BP protects the coding region of betaTrCP1 mRNA from miR-183-mediated degradation. Mol Cell 2009;35:240-246.
  • 32
    Huttelmaier S, Zenklusen D, Lederer M, Dictenberg J, Lorenz M, Meng X, et al. Spatial regulation of beta-actin translation by Src-dependent phosphorylation of ZBP1. Nature 2005;438:512-515.
  • 33
    Nielsen J, Christiansen J, Lykke-Andersen J, Johnsen AH, Wewer UM, Nielsen FC. A family of insulin-like growth factor II mRNA-binding proteins represses translation in late development. Mol Cell Biol 1999;19:1262-1270.
  • 34
    Noubissi FK, Goswami S, Sanek NA, Kawakami K, Minamoto T, Moser A, et al. Wnt signaling stimulates transcriptional outcome of the Hedgehog pathway by stabilizing GLI1 mRNA. Cancer Res 2009;69:8572-8578.
  • 35
    Lemm I, Ross J. Regulation of c-myc mRNA decay by translational pausing in a coding region instability determinant. Mol Cell Biol 2002;22:3959-3969.
  • 36
    Sparanese D, Lee CH. CRD-BP shields c-myc and MDR-1 RNA from endonucleolytic attack by a mammalian endoribonuclease. Nucleic Acids Res 2007;35:1209-1221.
  • 37
    Stohr N, Kohn M, Lederer M, Glass M, Reinke C, Singer RH, et al. IGF2BP1 promotes cell migration by regulating MK5 and PTEN signaling. Genes Dev 2012;26:176-189.
  • 38
    Houseley J, Tollervey D. The many pathways of RNA degradation. Cell 2009;136:763-776.
  • 39
    Doidge R, Mittal S, Aslam A, Winkler GS. Deadenylation of cytoplasmic mRNA by the mammalian Ccr4-Not complex. Biochem Soc Trans 2012;40:896-901.
  • 40
    Hafner M, Landthaler M, Burger L, Khorshid M, Hausser J, Berninger P, et al. Transcriptome-wide identification of RNA-binding protein and microRNA target sites by PAR-CLIP. Cell 2010;141:129-141.
  • 41
    Git A, Standart N. The KH domains of Xenopus Vg1RBP mediate RNA binding and self-association. RNA 2002;8:1319-1333.
  • 42
    Nielsen J, Kristensen MA, Willemoes M, Nielsen FC, Christiansen J. Sequential dimerization of human zipcode-binding protein IMP1 on RNA: a cooperative mechanism providing RNP stability. Nucleic Acids Res 2004;32:4368-4376.
  • 43
    Elcheva I, Tarapore RS, Bhatia N, Spiegelman VS. Overexpression of mRNA-binding protein CRD-BP in malignant melanomas. Oncogene 2008;27:5069-5074.
  • 44
    Noubissi FK, Elcheva I, Bhatia N, Shakoori A, Ougolkov A, Liu J, et al. CRD-BP mediates stabilization of betaTrCP1 and c-myc mRNA in response to beta-catenin signalling. Nature 2006;441:898-901.
  • 45
    Gu L, Shigemasa K, Ohama K. Increased expression of IGF II mRNA-binding protein 1 mRNA is associated with an advanced clinical stage and poor prognosis in patients with ovarian cancer. Int J Oncol 2004;24:671-678.
  • 46
    Kato T, Hayama S, Yamabuki T, Ishikawa N, Miyamoto M, Ito T, et al. Increased expression of insulin-like growth factor-II messenger RNA-binding protein 1 is associated with tumor progression in patients with lung cancer. Clin Cancer Res 2007;13:434-442.
  • 47
    Kobel M, Weidensdorfer D, Reinke C, Lederer M, Schmitt WD, Zeng K, et al. Expression of the RNA-binding protein IMP1 correlates with poor prognosis in ovarian carcinoma. Oncogene 2007;26:7584-7589.

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Author Contributions
  8. References
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

hep26537-sup-0001-suppfig1.eps1643KSupporting Figure 1: IGF2BP1 directly interacts with HULC in vivo and in vitro. (A) The interaction of IGF2BP1 with HULC ncRNA in HepG2 cells was verified by RNA-immunoprecipitation (RIP) after purification of endogenous IGF2BP1. (B) Potential IGF2BP1 binding sites (CAU) are distributed across HULC RNA. For further binding studies, HULC was split into two fragments (HULC_A (210 nt) and HULC_B (274 nt)) or left complete (HULC_FL, 484 nt). Arrows indicate PCR primer binding sites for later analysis. (C) Overview of the in vitro interaction assay applied to study IGF2BP1 binding to HULC. The same antibodies and beads as in (A) were used. An IGF2BP1 antibody or IgG control was coupled to Protein A/G beads. Beads were incubated with human recombinant IGF2BP1 and in vitro transcribed HULC fragments. Binding of IGF2BP1 towards different parts of HULC was analyzed via RT-PCR. (D) The assay efficiently and specifically purified rhIGF2BP1 as detected by western blotting (upper panel). No differential binding of IGF2BP1 to HULC could be observed using RT-PCR analysis (lower panel, expected PCR products: full-length = 110 bp, part A = 110 bp, part B = 138 bp). Each fragment bound directly and specifically to IGF2BP1. All experiments were executed at least three times. Shown are representative figures and the mean (±SEM).
hep26537-sup-0002-suppfig2.eps1608KSupporting Figure 2: Knockdown of IGF2BP1 induces while overexpression of IGF2BP1 reduces HULC expression. (A) Depletion of IGF2BP1 from HepG2 cells with two independent siRNAs resulted in a prolonged half-life of HULC. 48 h after siRNA transfection, cells were treated with alpha-amanitin (10 μg / mL f.c.) and samples were harvested at the indicated time points. The experiment was carried out in triplicates and given is the mean expression (±SEM). Gene expression was measured via qRT-PCR and HULC expression was normalized to the “0 h” time point. RNU6 expression was used as reference gene in each sample. (B) HepG2 cells were transiently overexpressed with pDEST-IGF2BP1. IGF2BP1 levels were measured using quantitative RT-PCR (left) and western blotting (right) showing efficient IGF2BP1 overexpression in HepG2 cells. PPIA was used as reference gene for PCR and GAPDH as loading control for western blotting. (C) After transient overexpression of IGF2BP1 for 72 h in HepG2 cells, HULC levels were measured showing a small but significant downregulation. The experiment was carried out in triplicates and given is the mean expression (±SEM).
hep26537-sup-0003-suppfig3.eps1108KSupporting Figure 3: Regulation of HULC expression after IGF2BP1 depletion is not mediated by miR-372. (A) HepG2 cells were transfected with control siRNA or two different IGF2BP1 siRNAs. After 72h, levels of mature miR-372 were measured using quantitative RT-PCR. No significant reduction of miR-372 could be detected despite a significant reduction of IGF2BP1. (B) HepG2, Hep3B and Huh7 cells were transfected with a miR-372 mimic or an appropriate negative control siRNA. After 24h, HULC expression was analyzed using quantitative RT-PCR. No significant downregulation of HULC could be detected despite a more than 2.000-fold overexpression of miR-372 as detected by qPCR (data not shown). All experiments were carried out in triplicates and given is the mean expression (±SEM). Gene expression levels were normalized to PPIA in each sample.
hep26537-sup-0004-suppfig4.eps1171KSupporting Figure 4: Combined depletion of IGF2BP1 and CNOT1 has no additive effect on HULC up-regulation. (A) HepG2 cells were transfected with either single IGF2BP1 or CNOT1 siRNAs or with a combination of both siRNAs (100 pmol per siRNA). After 72h, the levels of IGF2BP1 and CNOT1 were measured using quantitative RT-PCR showing significant and comparable downregulation of IGF2BP1 (upper panel) and CNOT1 (lower panel), irrespective of single or double knockdown. (B) HULC expression levels after individual or combined IGF2BP1 and CNOT1 knockdown for 72 h in HepG2 cells were significantly increased. Importantly, the combined knockdown had no additive effect on HULC expression which implicates that both proteins might be mechanistically linked. The experiment was carried out in triplicates and given is the mean expression (±SEM). Gene expression was normalized to PPIA in each sample.

Supporting Table 1: Patient's characteristics of expression profiling cohort

Supporting Table 2: qRT-PCR primer sequences

Supporting Table 3: siRNA sequences

Supporting Table 4: PCR primer with T7-overhang for in vitro transcription

Supporting Table 5: HULC interacting proteins

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.