Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Aug 11.
Published in final edited form as: Cancer Cell. 2014 Aug 11;26(2):248–261. doi: 10.1016/j.ccr.2014.06.018

Lin28b is sufficient to drive liver cancer and necessary for its maintenance in murine models

Liem H Nguyen 1,*, Daisy A Robinton 2,*, Marc Seligson 2,*, Linwei Wu 1,3, Lin Li 1, Dinesh Rakheja 4, Sarah Comerford 5, Saleh Ramezani 6, Xiankai Sun 6, Monisha Parikh 1, Erin Yang 1, John T Powers 2, Gen Shinoda 2, Samar Shah 2, Robert Hammer 5, George Q Daley 2, Hao Zhu 1
PMCID: PMC4145706  NIHMSID: NIHMS610967  PMID: 25117712

SUMMARY

Lin28a/b are RNA-binding proteins that influence stem cell maintenance, metabolism, and oncogenesis. Poorly differentiated, aggressive cancers often overexpress Lin28, but its role in tumor initiation or maintenance has not been definitively addressed. We report that LIN28B overexpression is sufficient to initiate hepatoblastoma and hepatocellular carcinoma in murine models. We also detected Lin28b overexpression in MYC-driven hepatoblastomas, and liver-specific deletion of Lin28a/b reduced tumor burden, extended latency, and prolonged survival. Both intravenous siRNA against Lin28b and conditional Lin28b deletion reduced tumor burden and prolonged survival. Igf2bp proteins are upregulated and Igf2bp3 is required in the context of LIN28B overexpression to promote growth. Thus, multiple murine models demonstrate that Lin28b is both sufficient to initiate liver cancer and necessary for its maintenance.

INTRODUCTION

MicroRNAs (miRNAs) and their protein regulators play important roles in cancer. One of the most ancient miRNAs is let-7, a 13-member tumor suppressor miRNA family downregulated in a large fraction of tumor types (Boyerinas et al., 2010; Roush and Slack, 2008). Let-7 negatively regulates the translation of oncogenes like Myc, Kras and Hmga2 (Johnson et al., 2005; Mayr et al., 2007; Sampson et al., 2007). The oncofetal RNA-binding proteins Lin28a and Lin28b (collectively referred to as Lin28) inhibit the biogenesis of all let-7s, thereby promoting embryo, stem cell, and tumor growth (Ambros and Horvitz, 1984; Chang et al., 2009; Shyh-Chang and Daley, 2013; Vadla et al., 2012; Viswanathan et al., 2008). Though some of the effects of Lin28 are due to let-7 suppression, Lin28 also binds and influences the translation of many messenger RNAs (Balzer and Moss, 2007; Cho et al., 2012; Hafner et al., 2013; Jin et al., 2011; Li et al., 2012; Peng et al., 2011; Wilbert et al., 2012; Xu et al., 2009). In this way, Lin28 and let-7 are at the center of a large RNA processing network that includes other RNA-binding proteins, non-coding RNAs, and their targets. Given the complexity and sheer number of interactions within these networks, a critical problem in the field is to define the phenotypic and therapeutic implications of disrupting individual components, of which Lin28 and let-7 are two of the most intensely studied (Hafner et al., 2013).

In cancer, Lin28a and Lin28b are aberrantly expressed in a wide spectrum of tumor types and are associated with advanced disease and poor prognosis (Viswanathan et al., 2009). Gain and loss of function studies in cancer cell lines have established that Lin28 can promote growth, invasion, and metastasis (Guo et al., 2006; Liang et al., 2010; Wang et al., 2010; You et al., 2013). Transgenic Lin28b overexpression leads to lymphoma and neuroblastoma development in mice, suggesting that in particular contexts Lin28b is alone sufficient to drive cancer (Beachy et al., 2012; Molenaar et al., 2012). However, it remains unclear whether Lin28a and Lin28b are relevant to tumor maintenance in endogenous tumors, and therefore compelling therapeutic targets.

In particular, Lin28 and let-7 are frequently and profoundly dysregulated in liver cancer (Cheng et al., 2013; Guo et al., 2006; Viswanathan et al., 2009; Wang et al., 2010), which represents the ideal setting to define the impact of disabling this RNA network. Hepatocellular carcinoma (HCC), the most common primary liver tumor, is the leading cause of cancer death worldwide and is becoming more prevalent in the U.S. as a consequence of chronic hepatitis infection, metabolic syndrome, and cirrhosis (Canberk et al., 2013; McMahon, 2009; Page and Harrison, 2009). Pediatric hepatoblastoma is a rare malignant embryonal tumor that shares molecular characteristics with HCC, though it usually arises in the absence of tissue injury (Canberk et al., 2013; Ishak and Glunz, 1967). In both adult and pediatric settings, treatments are inadequate and better therapeutic targets are needed (European Association for the Study of the Liver, 2012). Here, using genetically engineered murine models, we demonstrate that Lin28b overexpression is both sufficient to induce liver cancer and necessary for tumor maintenance. We further demonstrate the involvement of let-7 dependent and independent mechanisms, thereby implicating the Lin28/let-7 network as an important pathway in liver carcinogenesis.

RESULTS

LIN28B alone is sufficient to drive liver tumorigenesis in vivo

To assess whether human LIN28B would drive tumorigenesis in a murine model of liver cancer, we engineered drug inducible transgene overexpression using the reverse tetracycline-transactivating protein (rtTA) under the control of an ApoE promoter with hepatocyte-specific regulatory elements (Figure S1A). In this system, ApoE-rtTA mice were crossed to TRE-LIN28B mice (reported previously (Zhu et al., 2011), Figure 1A), and double transgenic mice (referred to as ApoE-LIN28B) were treated with 1mg/mL doxycycline (dox) throughout development, resulting in high liver-specific LIN28B expression (Figure S1B). Neonates were jaundiced with slight hepatomegaly, and one week after birth LIN28B overexpressing mice showed reduced growth relative to TRE-LIN28B single transgenic controls (Figure S1C). In p4 pups overexpressing LIN28B, histology revealed features of acute liver injury with cytoplasmic ballooning of hepatocytes and bile stasis in canaliculi (Figure S1D). Phospho-H3 staining showed increased proliferation in the liver (Figures S1E and S1F). Despite early indications of liver injury, surviving adult ApoE-LIN28B mice induced with dox did not exhibit significant signs of acute transaminitis, cholestasis, or chronic hepatic synthetic dysfunction (Figures S1G and S1H).

Figure 1. LIN28B is sufficient to drive liver tumorigenesis in vivo.

Figure 1

Open in a new tab

(A) Schematic of ApoE promoter-driven rtTA transgenic mouse crossed to TRE-LIN28B transgenic mouse. Double transgenic mice given dox (referred to as ApoE-LIN28B) overexpress human LIN28B in hepatocytes.

(B) Gross pathology of a 14-week old control single-transgenic TRE-LIN28B liver and multifocal ApoE-LIN28B liver tumors. Tumors occupy 5 to 40% of the liver mass.

(C) H&E stain of a 14-week old control single-transgenic TRE-LIN28B liver.

(D) H&E stain of an ApoE-LIN28B liver tumor. Yellow arrowhead points to features of HCC including large, pleiomorphic cells with prominent nucleoli and coarse chromatin.

(E) H&E stain of an ApoE-LIN28B tumor showing features of large HCC cells including prominent eosinophilic globules. Yellow arrowhead points to features of HCC including large, pleiomorphic cells with prominent nucleoli and coarse chromatin. Regions of embryonic hematopoiesis are identified with a green arrowhead.

(F) HCC component of the LIN28B-driven liver tumor showing widespread fatty change and vacuolization. The green arrowhead points to a region of embryonic hematopoiesis.

(G) H&E stain showing regions that resemble fetal hepatoblastoma, characterized by cells smaller than normal hepatocytes with higher nuclear/cytoplasmic ratios, characteristic oval nuclei with finely clumped chromatin, and small nucleoli.

(H) H&E stain showing regions of cholangioblastic and ductular differentiation (red arrowhead), along with abundant fatty vacuoles.

(I) Immunostaining of normal TRE-LIN28B control liver tissue, ApoE-LIN28B tumor adjacent normal liver tissue, and ApoE-LIN28B liver tumor tissue with Phospho-H3 antibody, demonstrating mitotically active cells.

(J) FLAG immunostaining of normal TRE-LIN28B control liver tissue (left), ApoE-LIN28B tumor adjacent normal tissue (middle), and ApoE-LIN28B tumor tissue (right).

(K) Transgenic LIN28B mRNA levels in control TRE-LIN28B livers (n = 4), ApoE-LIN28B tumor-adjacent normal liver tissues (n = 3) and ApoE-LIN28B liver tumors (n = 6), as determined by qRT-PCR.

(L) Western blot depicting levels of β-Actin and transgenic human LIN28B protein in control TRE-LIN28B livers and ApoE-LIN28B liver tumors.

(M) Levels of mature let-7 family members in control TRE-LIN28B livers (n = 3), ApoE-LIN28B tumor adjacent normal livers (n = 3), and ApoE-LIN28B liver tumor tissue (n = 6) as determined by qRT-PCR.

All data in this figure are represented as mean +/− SEM, * p < 0.05, ** p < 0.01. Normal control liver in these panels refers to single transgenic mice with TRE-LIN28B but lacking ApoE-rtTA. See also Figure S1.

By 6 months of age, 9/9 double transgenic mice, but no ApoE-rtTA or TRE-LIN28B single transgenic controls, had multifocal tumors that occupied up to 40% of the liver mass (Figure 1B). Compared to control TRE-LIN28B liver (Figure 1C), tumor histology revealed distinct HCC-like and hepatoblastoma-like morphologic patterns (Figures 1D-1H). HCC-like areas were composed of large cells with large nuclei, coarse chromatin, prominent nucleoli, and granular eosinophilic cytoplasm (Figures 1D and 1E, yellow arrowheads). In some areas, the HCC cells showed prominent and abundant fat vacuoles (Figure 1F). Hepatoblastoma-like areas were consistent with a fetal hepatoblastoma pattern composed of smaller cells with characteristic oval nuclei with finely clumped chromatin, small nucleoli, and eosinophilic cytoplasm (Figures 1G and 1H). Focally, the fetal hepatoblastoma pattern was contiguous with a cholangioblastic pattern composed of irregular ductal structures (Figure 1H, red arrowhead). Both HCC and hepatoblastoma regions exhibited embryonic hematopoiesis (green arrowheads in Figures 1E and 1F), commonly observed in human hepatocellular neoplasms. Tumors had a high mitotic rate as measured by Phospho-H3 and Ki-67 staining (Figures 1I and S1I).

Tumor tissue showed greater LIN28B transgene expression than LIN28B-driven tumor adjacent normal tissue, indicating enhanced expression of the human transgene within the tumor (Figures 1J-1L). Not surprisingly, mature let-7 biogenesis was potently suppressed (Figure 1M). Lin28a was not expressed in these tumors (data not shown). Hmga2, a well-known let-7 target, was not de-repressed after LIN28B induction in tumor adjacent normal tissue or in tumors themselves (Figure S1J). Likewise, c-Myc was not increased after LIN28B induction (Figure S1J). Expression of fetal markers, including Alpha-fetoprotein (Afp) and Insulin-like growth factor 2 (Igf2), were increased while Albumin mRNA was potently suppressed (Figures S1J–S1L), indicating either a block in hepatocyte differentiation or dedifferentiation in tumors. In addition, there was widespread overexpression of imprinted genes, a common feature of poorly differentiated tumors (Figure S1M). These data show that LIN28B overexpression is sufficient to drive liver cancers with morphologic and genetic features of HCC and hepatoblastoma.

Since ApoE-rtTA and ApoE-LIN28B mice treated with 1g/L of dox throughout embryogenesis suffer from liver injury and compromised survival rates, we reduced the dox dose ten-fold to 0.1g/L, which resulted in the survival of a higher percentage of double transgenic mice. To encourage tumor development with shorter latency, we injected a single 25mg/kg dose of HCC-initiating mutagen diethylnitrosamine (DEN) at 2 weeks of age. We then increased the dox to 1g/L at 3 weeks of age. After 100 days, 1/17 (6%) TRE-LIN28B, 0/15 (0%) ApoE-rtTA mice, and 17/33 (52%) ApoE-LIN28B mice treated with DEN harbor grossly visible tumors (Figures S1N and S1O; green arrowheads point to tumors). This shows that LIN28B overexpression is not only sufficient to initiate cancer, but also promotes the transformation and growth of premalignant cells established by DEN.

As Lin28 is known to activate the Insulin-Pi3k-mTOR pathway (Zhu et al., 2011), we asked if 2-deoxy-2-(18F) fluoro-D-glucose (FDG) PET/CT imaging, which exploits the fact that cancers are more glucose avid than normal tissues, could identify these tumors. In clinical practice, well-differentiated HCCs are not readily detected by PET, whereas high grade, poorly differentiated tumors are FDG-avid (Pant et al., 2013). While ApoE-LIN28B-induced tumors were difficult to detect by non-contrast CT alone, they were indeed FDG-avid (Figure 2). In our murine model, PET/CT imaging provides confirmation that these ApoE-LIN28B tumors are both glucose avid and poorly differentiated, akin to aggressive disease in humans.

Figure 2. LIN28B-induced liver tumors are detectable by FDG-PET/CT.

Figure 2

Open in a new tab

The left-most panels depict an FDG-PET image (above) and CT scan (below) of a control single transgenic TRE-LIN28B mouse with an absence of signal in the liver. The middle and right panels illustrate FDG-PET scans (above) and CT scans (below) of two induced ApoE-LIN28B mice. Signals in the liver are highlighted with yellow boxes. Signals are also evident in other metabolically active FDG avid zones, e.g. marrow and heart, as well as in the bladder.

Lin28b is aberrantly activated in mouse models of MYC-driven hepatoblastoma

Next, we asked if Lin28a or Lin28b are overexpressed in two MYC-driven mouse models, both of which harbor a dox-inducible human c-MYC transgene (Shachaf et al., 2004). In the first model (LAP-MYC), MYC is under the control of a tetracycline-dependent repressor expressed from the liver-activated promoter (LAP) (Figure 3A) (Beer et al., 2004; Shachaf et al., 2004), such that removal of dox results in MYC expression and hepatoblastoma development. This convenient two-transgene model has a predictable onset of tumorigenesis in the FVB strain background (Beer et al., 2004). We also employed a second model (Alb-MYC) in which TRE-MYC is crossed to a Rosa26-Lox-STOP-Lox-rtTA (RLSL-rtTA); Albumin (Alb)-Cre mouse. In this strain, MYC is expressed in response to dox after Cre excises the STOP cassette in Albumin-positive cells (Figure 3B). While these triple transgenic mice are more challenging to breed, they enable complex genetic intercrossing, as tumors arise predictably in mixed genetic strain backgrounds.

Figure 3. Lin28b is aberrantly activated in mouse models of MYC-driven hepatoblastoma.

Figure 3

Open in a new tab

(A) Schematic showing LAP-tTa; TRE-MYC (or “LAP-MYC”) transgenic mouse model. TRE-MYC is driven by the liver-specific LAP promoter when present with the LAP-tTA transgene in the absence of dox.

(B) Schematic showing Alb-Cre; RLSL-rtTA; TRE-MYC (or “Alb-MYC”) triple transgenic mouse model. MYC is expressed after hepatocyte-specific Cre-mediated excision of the stop cassette in Albumin-positive cells, followed by subsequent introduction of dox

(C) Gross pathology of an Alb-MYC induced liver tumor.

(D) Representative H&E staining of an Alb-MYC driven liver tumor illustrating a mixed embryonal (left) and fetal (right) hepatoblastoma.

(E) Left panel: Transgenic human c-MYC mRNA levels in wild-type normal liver (n = 8) and Alb-MYC liver tumors (n = 12) as determined by qRT-PCR. Middle panel: Lin28a mRNA levels in wild-type normal liver (n = 8) and Alb-MYC liver tumors (n = 12) as determined by qRT-PCR. Right panel: Lin28b mRNA levels in wild-type normal liver (n = 8) and Alb-MYC liver tumors (n = 12) as determined by qRT-PCR.

(F) Representative immunostaining of c-MYC (above) and Lin28b (below) in wild-type normal liver tissue (left) and Alb-MYC driven liver tumors (right).

(G) Mature let-7 expression levels in wild-type normal liver (n = 8) and Rosa-LSL-rtTA; Alb-MYC induced tumors (n = 12) as determined by qRT-PCR.

(H) Mouse Lin28a and Lin28b mRNA levels as a function of MYC induction time in LAP-tTA; TRE-MYC liver tumors, as determined by qRT-PCR. Inset panel shows a western blot depicting protein levels of c-MYC, Lin28b and β-Actin in liver tumors induced at p-3, p11 or p16. All data in this figure are represented as mean +/− SEM, * p < 0.05, ** p < 0.01.

Both the “tet-off” LAP-MYC model and the “tet-on” Alb-MYC model overexpressed c-MYC under the control of dox, and led to robust hepatoblastoma formation in the liver (Figure 3C shows a representative Alb-MYC mouse with tumors). The tumor type from both models was a mixed embryonal and fetal hepatoblastoma, with a higher fraction of embryonal hepatoblastoma cells that are smaller with higher nuclear to cytoplasmic ratio and oval to angulated hyperchromatic nuclei (Figure 3D). Expression analysis of normal versus tumor tissue showed high expression of MYC, accompanied by modestly increased levels of Lin28a (∼5 fold) and marked elevation of Lin28b transcript (∼1,000–10,000 fold; Figure 3E). Immunohistochemistry revealed strong staining for MYC and Lin28b in tumors relative to control wild-type (WT) liver (Figure 3F), and was accompanied by significant suppression of all measured let-7 family members except for let-7i (Figure 3G).

MYC-driven liver tumors have been reported to be more aggressive when MYC is activated at earlier developmental time points (Beer et al., 2004), suggesting that genetic or epigenetic factors associated with the embryonic state collaborate to promote high-grade malignancy. Lin28a and b are heterochronic factors that regulate embryonic growth and metabolism, and we observed that earlier MYC induction correlated with high levels of Lin28b mRNA and protein overexpression (Figure 3H). These results indicated that Lin28b was associated with a more aggressive, embryonic tumor type characteristic of an oncofetal gene expression program.

Deletion of Lin28a/b increases latency and prolongs survival in hepatoblastoma

Prior studies in HCC cell lines have indicated that shRNA suppression of Lin28 reduces cell proliferation in vitro and xenograft formation in vivo (Molenaar et al., 2012; Viswanathan et al., 2009). However, cell lines and human xenografts do not faithfully model endogenous tumors, and because of off-target effects, RNA interference can be an unreliable method of target validation (Gu et al., 2012; Qiu et al., 2005). Therefore, we sought to determine whether the increased Lin28 expression is necessary for tumor progression by genetically deleting Lin28a and b in the context of the MYC-driven tumor models. Tumor-bearing TRE-MYC; RLSL-rtTA; Lin28a+/+; Lin28b+/+; Alb-Cre mice induced on the day of birth showed a median survival of 40 days, whereas TRE-MYC; RLSL-rtTA; Lin28aFl/Fl or Fl/−; Lin28bFl/Fl or Fl/−; Alb-Cre mice (double knockout or DKO mice) showed markedly prolonged survival, with the majority remaining alive at 150 days when the experiment was stopped (Figure 4A, p < 0.0001). Upon gross observation, DKO mice showed reduced tumor burden with intact regions of normal liver tissue (Figure 4B). Both Lin28a/b WT and DKO tumors showed embryonal and fetal hepatoblastoma morphology, indicating that Lin28 loss did not alter tumor histology (Figure 4C). Curiously, we observed no significant differences in cell proliferation, as measured by phospho-H3 (Figure S2A), but did detect 3-fold more cells undergoing apoptosis, as measured by Cleaved Caspase-3 in DKO tumors (Figures S2A and S2B).

Figure 4. Genetic deletion of Lin28a/b increases latency and prolongs survival in MYC-driven liver tumor models.

Figure 4

Open in a new tab

(A) Kaplan–Meier survival curve of Lin28a/b WT; Alb-MYC mice (yellow; n = 29) or Lin28a/b DKO; Alb-MYC mice (blue; n = 13) dox-induced on the day of birth (p < 0.0001). Median survival of the Lin28a/b WT; Alb-MYC mice is 40 days, whereas the Lin28a/b DKO; Alb-MYC mice had not achieved median survival at 150 days.

(B) Gross anatomy of Alb-MYC liver tumors with intact Lin28a/b (WT; upper panel) and in the setting of Lin28a/b liver specific deletion (DKO; lower panel). DKO mice were euthanized at a more advanced age, near their time of death.

(C) H&E staining of Lin28a/b WT (upper) and Lin28a/b DKO (lower) Alb-MYC tumors, showing embryonal (left) and fetal (right) hepatoblastoma histology.

(D) Human c-MYC, mouse Lin28b, Lin28a, Igf2, Albumin, and Afp mRNA levels in Lin28a/b WT; Alb-MYC liver tumors (n = 8) and Lin28a/b DKO; Alb-MYC liver tumors (n = 8) as determined by qRT-PCR.

(E) Western blot illustrating protein levels of Lin28a, Lin28b, c-MYC, and β-Tubulin in e9.5 mouse embryos (Lin28a/b positive control), WT testis (Lin28a/b and MYC positive control), Lin28a/b WT normal liver, Lin28a/b WT; LAP-MYC tumors, Lin28a/b DKO normal liver and Lin28a/b DKO; LAP-MYC liver tumors.

(F) Expression of mature let-7 family members in Lin28a/b WT (n = 10) and Lin28a/b DKO (n = 14) Alb-MYC tumors.

All data in this figure are represented as mean +/− SEM, * p < 0.05, ** p < 0.01. See also Figure S2.

Importantly, MYC mRNA and protein levels remained intact in DKO tumors, showing that the primary oncogenic driver of the tumor model was not suppressed transcriptionally or translationally in the absence of Lin28a/b (Figures 4D and 4E). Expression analysis confirmed that Alb-Cre markedly reduced Lin28b expression on the mRNA and protein levels in DKO mice (Figures 4D and 4E), whereas we detected no significant change in the low levels of Lin28a mRNA or protein expression (Figures 4D and 4E). To explore mechanisms responsible for growth impairment, we examined markers of differentiation. Lin28a/b loss resulted in a significant reduction in Igf2 expression (Figure 4D), consistent with previous reports linking Igf2 expression and translation to Lin28 (Polesskaya et al., 2007; Zhu et al., 2010). In contrast, levels of Alb and Afp mRNA did not differ between WT and DKO tumors (Figure 4D). Analyzing a previously established 16-gene signature for hepatoblastoma classification showed no difference in 15/16 genes, suggesting that tumor differentiation is unchanged (Cairo et al., 2008) (Figure S2C). We did observe broad-spectrum increases in mature let-7 levels, though only let-7g and let-7i were significantly elevated (Figure 4F). Notably, let-7s were significantly repressed in DKO hepatoblastoma when compared to DKO normal liver, but levels of repression were not as great as was in the WT tumors (data not shown). This suggests that in tumors formed in the absence of Lin28a/b, let-7 repression was in part caused by mechanisms independent of Lin28a and Lin28b, and that growth impairment is not exclusively due to let-7 increases.

In mice, both Lin28a and Lin28b are highly expressed in embryonic tissues but largely silent in adults (Moss and Tang, 2003; Shinoda et al., 2013; Yang and Moss, 2003). We confirmed that Lin28a and b are expressed in embryonic and neonatal but not adult liver (Figures 5A–5C). In prior studies we found that constitutional Lin28a knockout mice were 10% smaller than WT littermates, and displayed perinatal lethality but no other obvious phenotypes (Shinoda et al., 2013). Liver-specific Lin28a and Lin28b DKO mice are also small but viable, and have normal liver function (Figures S2DE and S2EF). Thus, our observation of reduced tumorigenicity in Lin28a/b DKO mice does not appear to be due to developmental defects in the liver, and indicates that liver-specific or systemic anti-Lin28 therapy would be well tolerated.

Figure 5. Lin28a/b expression in the liver throughout development.

Figure 5

Open in a new tab

(A) Lin28b mRNA expression during a liver development time-course as determined by qRT-PCR. Heat map shown below illustrates high (red) and low (green) relative expression in late embryos, neonates and young adult livers.

(B) Lin28a and Lin28b protein expression throughout development in the liver relative to β-Actin protein, as determined by western blot analysis.

(C) Immunostaining of Lin28b in e14.5 (left), p2 (middle) and p30 (right) liver tissue. All data in this figure are represented as mean +/− SEM, * p < 0.05, ** p < 0.01.

Igf2bp3 acts as an oncogene downstream of Lin28b

In pursuit of specific Lin28 and let-7 targets that might be involved in hepatoblastoma and HCC biology, we examined the Igf2 mRNA binding proteins (Igf2bp1-3), members of a candidate oncofetal gene family regulated by let-7 (Boyerinas et al., 2008). IGF2BP3 in particular is highly expressed in human liver tumors but not in normal liver (Figure S3A) (Wurmbach et al., 2007), and thus has been proposed as a sensitive HCC biomarker (Chen et al., 2013a; Chen et al., 2013b; Jeng et al., 2008; Nischalke et al., 2012; Wachter et al., 2012). Relatively little is known about the functional activities, biological regulation, and therapeutic relevance of these genes in cancer. Overexpression of both mouse Lin28b and human LIN28B in SV40-immortalized H2.35 hepatocytes led to let-7 suppression and enhanced cell proliferation in vitro (Figures S3B–S3D) and tumor growth in xenografts (Figures 6A and 6B). In vitro, H2.35 cells that overexpress Lin28b showed elevated mRNA for all three Igf2bp orthologs (Figure S3E). In vivo, the corresponding xenografts exhibited robust increases in Igf2bp1 and Igf2bp3 protein levels (Figure 6C). Tumor tissue from the ApoE-LIN28B hepatoblastoma/HCC model likewise showed elevated Igf2bp1 and Igf2bp3, but not Igf2bp2, mRNA and protein levels relative to ApoE-LIN28B-driven tumor adjacent normal tissue and normal liver, also confirmed by immunostaining (Figures 6D-6F). In LAP-MYC hepatoblastomas, Igf2bp1 and Igf2bp3, but not Igf2bp2, were both highly overexpressed (Figure S3F). Notably, LIN28B and IGF2BP3 are also co-expressed in advanced, but not early stage, human HCCs (Figure S3G) (Wurmbach et al., 2007). Taken together, these data indicate that LIN28B overexpression is tightly associated with upregulation of Igf2bp1 and Igf2bp3 in mouse and human liver cancers.

Figure 6. Igf2bp3 acts as an oncogene downstream of Lin28b.

Figure 6

Open in a new tab

(A) Growth curve showing volume of xenografts composed of SV40-immortalized H2.35 hepatocytes overexpressing pBABE empty vector (n = 5) or FLAG-LIN28B (n = 5) over time.

(B) Gross anatomy of xenografts 50 days post-injection.

(C) Western blot showing LIN28B, Igf2bp1, Igf2bp3 and β-Tubulin protein levels in H2.35-xenografts with and without LIN28B overexpression.

(D) Left panel: Igf2bp1 mRNA levels in normal TRE-LIN28B control liver (n = 4), LIN28B-induced tumor adjacent normal liver (n = 3) and LIN28B-induced liver tumor (n = 6) tissue, as determined by qRT-PCR. Right panel: Igf2bp3 mRNA levels in normal TRE-LIN28B control liver (n = 4), LIN28B-induced tumor adjacent normal liver (n = 3) and LIN28B-induced liver tumor (n = 6) tissue, as determined by qRT-PCR.

(E) Western blot depicting levels of Igf2bp1, Igf2bp3 and β-Actin protein in normal TRE-LIN28B control liver and ApoE-LIN28B liver tumor tissues.

(F) Immunostaining of Igf2bp1 and Igf2bp3 in normal TRE-LIN28B control liver and ApoE-LIN28B liver tumors. Dashed lines indicate the border between LIN28B-induced tumor tissue (upper right) and tumor-adjacent normal liver tissue (lower left).

(G) Cell proliferation of pBABE control and LIN28B-overexpressing H2.35 cells treated with siScramble or siIgf2bp3 for 3 days (n = 3 for each of the 4 groups).

(H) Western blot depicting levels of Igf2bp3, LIN28B, and β-Actin in H2.35 cells infected with retroviruses carrying pBABE empty vector or pBABE-LIN28B overexpression vector and treated with siRNA against Igf2bp3.

All data in this figure are represented as mean +/− SEM, * p < 0.05, ** p < 0.01. See also Figure S3.

Multiple lines of evidence support interactions between Lin28s and Igf2bps, dependent and independent of let-7 (Boyerinas et al., 2008; Mayr et al., 2007; Polesskaya et al., 2007). Igf2bp1-3 are known to be directly suppressed by let-7 (Boyerinas et al., 2008). Hafner et al. showed that LIN28A and B directly bind IGF2BP1-3 mRNAs using Photoactivatable-Ribonucleoside-Enhanced Crosslinking and Immunoprecipitation (PAR-CLIP) in HEK293 cells (Hafner et al., 2013). Their proteomic analysis also showed that IGF2BP1-3 protein levels increased when LIN28B was overexpressed and decreased when LIN28B was knocked-down. We found that antibody precipitation of FLAG-LIN28B from both H2.35 cells and BE(2)C neuroblastoma cells co-precipitated IGF2BP1-3 proteins, supporting a direct physical interaction (Figures S3H and S3I). Furthermore, this interaction was abrogated upon treatment with RNase, indicating that binding was dependent on RNA. To determine whether Igf2pb3 is a downstream effector of Lin28b, we retrovirally overexpressed LIN28B in H2.35 cells and then performed siRNA knockdown of Igf2bp3. In the absence of LIN28B overexpression, Igf2bp3 knockdown reduced cell proliferation, but not significantly (Figure 6G). However, in the context of LIN28B-overexpression, the reduction in cell proliferation was significant (Figures 6G and 6H), indicating that LIN28B positive cells were more dependent on Igf2bp3. These data suggest that in this cell line, LIN28B preferentially requires Igf2bp3 to carry out growth-promoting functions and that these two RNA-binding protein families can work in concert to promote liver cancer development.

The growth of MYC-induced hepatoblastoma is impaired after anti-Lin28b therapy

We asked whether RNA-based therapy against Lin28b would antagonize tumor growth, thereby implicating Lin28b as essential for tumor initiation and maintenance. MYC was induced shortly before birth via dox withdrawal such that 24 day-old mice had multiple established tumors (Figure S4A, green arrowheads). These mice were then randomized to no treatment, scrambled control siRNA, or two different anti-Lin28b siRNAs. Both siLin28b 1 and 2, which target distinct coding sequences within Lin28b, were effective at knocking down Lin28b overexpression in H2.35 cells, though siLin28b 1 was more efficient than siLin28b 2 (Figure S4B). These siRNAs were then injected into LAP-MYC cohorts intraperitoneally at p14 and intravenously at p24 and p31. Activity levels and weights of untreated mice were identical to those of siRNA treated mice, indicating that the therapies were not obviously toxic (Figure S4C). Treatment with siLin28b 1 extended median survival from 47 to 57 days (p = 0.0006; Figure 7A) and 4/15 (26%) treated mice remained alive at 80 days, well beyond the 47-day median survival of untreated mice. Treatment with siLin28b 2 also extended median survival, but this did not reach statistical significance (p = 0.054; Figure 7A). Both experimental cohorts, as compared to control cohorts, had reduced abdominal girth, which is well correlated with tumor growth and progression (p = 0.0040 for siLin28b 1 and p = 0.199 for siLin28b 2; Figures 7B and 7C). Gross and histological examination showed that mice treated with siLin28b 1 and 2 had significantly less tumor burden compared to controls at 35 days of age (Figure 7D). qRT-PCR confirmed reduced Lin28b mRNA after siLin28b 1 but not siLin28b 2 treatment, and western blot showed substantially reduced protein for siLin28b 1 but not for siLin28b 2 (Figures 7E and 7F). Mature let-7 expression did not increase (Figure S4D), while MYC, Igf2bp1, and Igf2bp3 did not decrease in the experimental groups (Figures 7E, 7F and S4E). The fact that Igf2bp1 and Igf2bp3 are not altered suggests that their expression is not entirely regulated by Lin28b and that the phenotype of Lin28b loss is not always dependent on reduced Igf2bp1/3 expression. While siLin28b 1 treated tumors had a similar number of proliferating cells compared to scrambled siRNA treated tumors, siLin28b 1 tumors had 6-fold more cells undergoing apoptosis than siScramble tumors (Figures 7G and 7H). This is consistent with the higher levels of apoptosis seen in the Lin28a and Lin28b DKO tumors (Figures S2B and S2C), and accounts for reductions in tumor growth.

Figure 7. The growth of MYC-induced hepatoblastoma is impaired after anti-Lin28b therapy.

Figure 7

Open in a new tab

(A) Kaplan–Meier survival curve of MYC-induced mice randomized to no treatment (n = 29), siScramble (n = 6), siLin28b 1 (n = 15) or siLin28b 2 (n = 6). MYC was induced 2 days prior to birth by withdrawing dox. Median survival for the no treatment group is 46 days, for the siScramble group is 47 days, and for both the siLin28b 1 and 2 groups is 57 days (p = 0.0006 for siLin28b 1 and p = 0.054 for siLin28b 2).

(B) Abdominal circumference (in cm) of MYC-induced mice at 35 days after treatment with siScramble (n = 6), siLin28b 1 (n = 6), or siLin28b 2 (n = 5).

(C) Pictures of the abdominal region of MYC-induced mice at p35 after treatment with siScramble, siLin28b 1 or siLin28b 2.

(D) Representative H&E staining of MYC-induced mice at p35 after treatment with siScramble, siLin28b 1 or siLin28b 2.

(E) Expression levels of Lin28b and c-MYC mRNAs in the tumors of MYC-induced mice treated with siScramble, siLin28b 1 or siLin28b 2. Gene expression was normalized to β-Actin.

(F) Western blot depicting protein levels of c-MYC, Lin28b, and β-Actin in the tumors of MYC-induced mice at p35 treated with siScramble, siLin28b 1 or siLin28b 2.

(G) Phospho-H3 and Cleaved Caspase-3 immunostaining of MYC-induced tumors treated with siScramble or siLin28b 1.

(H) Number of phospho-H3 and Cleaved Caspase-3 positive cells per 20x field in the tumors of -MYC-induced mice at p35 following treatment with siRNAs.

(I) Kaplan-Meier survival curve of MYC-induced mice with intact Lin28b or conditional deletion of Lin28b. The control group retained at least one wild-type Lin28b allele after tamoxifen exposure: Lin28b+/+, Lin28b+/Fl and Lin28b+/−. It was possible for the experimental group mice to have no wild-type Lin28b alleles after tamoxifen exposure: Lin28bFl/− and Lin28bFl/Fl. The median survival time of the Lin28b+/+, Lin28b+/Fl,and Lin28b+/− control group was 67 days, whereas that of Lin28bFl/−, Lin28bFl/Fl experimental group was 99 days.

All data in this figure are represented as mean +/− SEM, * p < 0.05, ** p < 0.01. See also Figure S4.

We then validated the siRNA results with conditional genetic deletion of Lin28b in LAP-MYC mice. To this end, we generated LAP-tTA; TRE-MYC; Ubiquitin-Cre/ERT2; Lin28b+/+, Lin28bFl/+, Lin28bFl/Fl, and Lin28bFl/− mice. In the experimental groups (Lin28bFl/Fl and Lin28bFl/−), Lin28b floxed alleles could be completely excised via tamoxifen exposure, while in the control groups (Lin28b+/+, Lin28b+/−, and Lin28bFl/+), at least one Lin28b WT allele remains. As in the siRNA experiment, dox was withdrawn at birth to generate tumors by 1 month of age, at which point all mice were injected with tamoxifen to conditionally delete Lin28b throughout the animal and within tumors. We first interrogated our tamoxifen regimen to demonstrate Cre excision occurred in 90–100% of hepatocytes in Rosa-Lox-STOP-Lox-tdTomato reporter mice (Figure S4F). Genotyping of mice with floxed-Lin28b alleles showed >75% excision of floxed alleles in 6/10 (60%) of examined mice, and <50% excision in 4/10 (40%) mice examined (representative genotyping shown in Figure S4G). Conditional deletion of Lin28b in Lin28bFl/− and Lin28bFl/Fl mice significantly prolonged survival relative to control mice (Figure 7I, p = 0.019). 8/17 (47%) of the tumors from conditional Lin28b knockout mice demonstrated complete excision of Lin28b (Figure S4G), demonstrating that it is possible for MYC-driven tumors to compensate for the loss of Lin28b. Nevertheless, control mice had much greater tumor burden with a median survival of 67 days, while conditionally Lin28b-deleted mice showed reduced tumor burden with a median survival of 99 days (Figure 7I). Together with the data from our siLin28b experiments, these findings support Lin28b as a required maintenance factor in MYC-driven tumors, suggesting that targeting Lin28b might be a promising therapeutic strategy for hepatoblastoma and HCC.

DISCUSSION

We report that Lin28b is sufficient to drive malignant liver neoplasms and is necessary for tumor growth in multiple murine models. These data demonstrate the functional importance of LIN28B overexpression, which has been observed extensively in human liver cancers (Cairo et al., 2008; Viswanathan et al., 2009; Wang et al., 2010). Moreover, in aggressive MYC-driven tumors with Lin28b overexpression, we have shown that either genetic deletion or siRNA-mediated knockdown of Lin28b prolongs survival, validating it as an oncogene with essential roles in liver cancer and potentially other malignancies.

Lin28a/b are heterochronic genes that regulate the growth and metabolic rates of embryonic developmental stages (Moss and Tang, 2003; Shyh-Chang and Daley, 2013; Yang and Moss, 2003). When expressed at the wrong time or place, they have the potential to revert adult cell physiology to a more embryonic state (Shyh-Chang and Daley, 2013). Lin28a can act together with transcriptional regulators found in embryonic stem cells to reprogram adult cells to a pluripotent state (Yu et al., 2007). In cancer, expression of Lin28 in post-natal tissues appears to reactivate programs of embryonic growth and metabolism to drive tumorigenesis (Viswanathan et al., 2009; Yu et al., 2007). In support of this hypothesis, ApoE-LIN28B tumors exhibit a poorly differentiated histology with elements of fetal hepatoblastoma, poorly differentiated HCC, biliary differentiation, and fetal hematopoiesis that mirror the pathology of human liver cancers (Zhang et al., 2008). Moreover, FDG-PET imaging revealed that LIN28B-driven tumors share metabolic features with glucose-avid, poorly differentiated tumors. Just as tumor behavior reflects its cell of origin, it also reflects its temporal origins (Visvader, 2011). Our data gives credence to the idea that Lin28b levels reflect tumors that have reactivated an oncofetal program, which helps to explain why the Lin28 family is dysregulated in many pediatric tumors, and potentially why pediatric and adult cancers differ in behavior and in treatment response.

A major challenge in the RNA-processing field is assessing the impact of individual RNA-binding protein targets versus systems level effects on many targets. Whereas Lin28 has a defined and important target set in the let-7 family, our liver tumor models demonstrate that Lin28 likely acts through both let-7 dependent and independent mechanisms. The LIN28B-driven hepatoblastoma/HCC mouse model will facilitate future efforts at determining the critical molecular targets of LIN28B in cancer, let-7 and otherwise.

Most importantly, using robust endogenous tumor models, our studies show that LIN28B is sufficient to drive tumorigenesis, and show that either siRNA treatment against or conditional genetic deletion of Lin28b in mice with established liver tumors slows tumor growth and extends survival. Thus, Lin28 could represent a therapeutic target in refractory liver tumors, and potentially in other Lin28-positive tumors.

EXPERIMENTAL PROCEDURES

Mice

All animal procedures were based on animal care guidelines approved by the Institutional Animal Care and Use Committee at Children’s Hospital Boston and UTSW. Alb-Cre and ROSA26-rtTA-IRES-EGFP mice were purchased from The Jackson Laboratory. Ubiquitin-Cre/ERT2 mice were a generous donation from the Morrison Lab at UTSW. ApoE-rtTAM2; TRE2-Cre transgenic mice expressing Cre recombinase under control of a liver-specific, doxycycline- inducible system were generated by co-injection of 2 transgenes; an ∼11kb ApoE-rtTAM2 transgene and the 2.8 kb TRE2-Cre transgene, kindly provided by Jiafu Ou (UTSW Dept. of Molecular Genetics). Cre was not utilized in this report. ApoE-rtTA mice were crossed to TRE-LIN28B mice and administered 1mg/mL or 0.1mg/mL of dox to the drinking water to induce LIN28B expression. MYC-driven liver tumors were generated by crossing the TRE-MYC strain with either LAP-tTA mice (described previously (Beer et al., 2004)), or with RLSL-rtTA (Belteki et al., 2005); Albumin-Cre (obtained from Jackson Labs). Mice bearing the LAP-tTA; TRE-MYC genotype were maintained on 1mg/mL of dox, and MYC was induced by withdrawing dox. Mice bearing the RLSL-rtTA; Alb-Cre; TRE-MYC transgenes were induced at birth with administration of 1mg/ml dox. Design of conditional Lin28a and Lin28b knockout mice were reported previously (Shinoda et al., 2013; Zhu et al., 2011). Briefly, exon 2 encoding the functional cold shock domain is flanked by LoxP sites such that expression of Cre recombinase excises the loci. The mice used contained the PGK-neo selection cassette. For conditional deletion of Lin28b, tamoxifen (Sigma) was dissolved in corn oil to 20 mg/mL by sonication for 10 minutes at 25°C, then incubation at 50°C for 10 minutes. 0.2 mg/g of mouse weight was injected IP for 3 consecutive days. After a 6 day break, mice were given tamoxifen for 2 more consecutive days. Lin28b floxed and wild-type alleles were genotyped using these primers: AACGCACATTGCAAATACCC (forward) and TTCATCTGGCTCCTTTCTCG (reverse) primers. Excised alleles were detected using these primers: CAGAAAGCGAAGGAGCAAAG (forward) and TCTTCTCTGGCATTGCTTCA (reverse).

Small Animal PET/CT

Mouse PET/CT imaging was performed using Siemens Inveon PET/CT Multimodality System (Siemens Medical Solutions, Knoxville, TN) with effective spatial resolution of 1.4 mm at the center of field of view (FOV). Mice were fasted for 12 hours prior to PET imaging and each mouse received 140 µCi of FDG in150 µL in saline intravenously via the tail vain. PET images were acquired one hour post-injection. CT images were acquired immediately after PET with the FOV centered at the shoulder of the mouse. CT projections (360 steps/rotation) were acquired with a power of 80 kVp, current of 500 µA, exposure time of 145 ms, binning of 4, and effective pixel size of 102 µm. PET images were reconstructed into a single frame using the 3D Ordered Subsets Expectation Maximization (OSEM3D/MAP) algorithm. The CT reconstruction protocol with a down sample factor of 2, which was set to interpolate bilinearly, used a Shepp-Logan filter. PET and CT images were co-registered by the manufacturer’s software for analysis.

Cell Lines, Constructs and Transfection Experiments

BE(2)C and H2.35 cells were purchased from ATCC. Cells were cultured using standard methods as described by the vendor. Reverse transfection was used for siRNA knockdown of Igf2bp3 in H2.35 cells. siScramble (Life Technologies, #4390843) or siIgf2bp3 (Life Technologies, 4390771-s100444) in Lipofectamine 2000 (Life Technologies, #11668019) was used at the concentration of 25nM. Forward transfection was employed for siRNA knockdown of Lin28b in H2.35 cells. siScramble or siLin28b 1 or 2 (Life Technologies #4390771-s117291, #4390771-s117292) in Lipofectamine 2000 was used at the concentration of 25nM.

Histology

Tissue samples were fixed in 10% neutral buffered formalin or 4% paraformaldehyde (PFA) and embedded in paraffin. In some cases, frozen sections were made.

Immunohistochemistry and Immunofluorescence

Immunostaining of frozen sections

Liver or tumor issue was fixed in 4% paraformaldehyde for 1.5 hours at RT, then dehydrated overnight in 30% w/v sucrose. Tissue was embedded in Cryo-gel cutting temperature compound (Tissue-Tek, #62806-01) and frozen on dry ice. Frozen sections were cut at 7 µm, incubated with 3% H2O2 in methanol for 10 minutes, blocked with 5% goat serum for 1 hour, labeled with primary antibodies overnight at 4°C and secondary antibody for 30 minutes at room temperature. Nuclear DNA was stained with 300 nM Dapi (Vector Biolabs, #NC9029229).

Immunostaining of paraffin sections

Tissue was fixed in 4% paraformaldehyde overnight, then put into in 75% ethanol overnight and embedded in paraffin. Slides were dewaxed with xylene and rehydrated through a series of washes with decreasing percentages of ethanol. Antigen retrieval was performed in 10 mM sodium citrate buffer (pH 6.0) by placement in a microwave on high for 20 minutes. Slides were treated with 3% hydrogen peroxide in methanol for 10 minutes to inhibit endogenous peroxidase activity. After blocking with 5% goat serum and the Avidin/Biotin Blocking Kit (Vector Laboratories, SP-2001), slides were incubated with primary antibody overnight at 4°C and secondary antibody for 30 minutes at room temperature. Detection was performed with the Elite ABC Kit and DAB Substrate (Vector Laboratories, SK-4100).

Immunofluorescence of frozen sections

Liver or tumor issue was fixed in 4% paraformaldehyde for 1.5 hours at RT, then dehydrated overnight in 30% w/v sucrose. Tissue was embedded in Cryo-gel cutting temperature compound (Tissue-Tek, #62806-01) and frozen on dry ice. Frozen sections were cut at 7 µm. Nuclear DNA was stained with 300 nM Dapi (Vector Biolabs, #NC9029229). Antibodies used: anti-Lin28a (Cell Signaling, #8641), anti-Lin28b (Cell Signaling, #5422), anti-Ki-67 (Abcam, #15580), anti-Phospho-H3 (Cell Signaling, #9706), anti-Cleaved Caspase-3 (Cell Signaling, #9661), anti-c-MYC (Sigma, #M4439), and anti-Afp (R&D Systems, MAB1368).

Western Blot Assay

Cells and tissues were lysed in RIPA buffer and macerated with a pestle. Proteins were separated by 10% polyacrylamide gel and transferred to a PVDF membrane (BioRad, #170-4159). The membrane was blocked overnight at 4°C in PBST containing 5% milk, and subsequently probed with primary antibodies overnight at 4°C. After incubating the membrane with donkey-anti-rabbit or sheep-anti-mouse HRP-conjugated secondary antibody (GE Healthcare), protein levels were detected with SuperSignal West Pico or Femto Luminol reagents (Thermo Scientific, #34095). Primary antibodies were prepared in 5% milk in PBST. The following primary antibodies were used: c-MYC (Santa Cruz, sc788), Igf2bp1 (Cell signaling, #8482), IGF2BP3 (Proteintech, #14642-1-AP), β-Tubulin (Cell Signaling, #2128).

RNA Extraction and qPCR

Total RNA was isolated using Trizol reagent (Invitrogen, #15596018). For qRT-PCR of miRNA or mRNA, cDNA synthesis was performed with 1 ug of total RNA using miScript II Reverse Transcription Kit (Qiagen, #218161) or iScript RT Supermix (BioRad, #1708840) . miRNA expression was evaluated using Qiagen miScript primers. 12.5 ng of total RNA was used to measure mRNA expression levels via qRT-PCR, following the protocol as specified in the miScript kit. mRNA levels were normalized to β-Actin gene expression, whereas miRNA levels were normalized to expression of U6 miRNA. Gene expression levels were measured using the ΔΔCt method as described previously (Zhu et al., 2011).

In vivo siRNA treatment

Ambion® in vivo siScramble (Life Technologies, #4457289), siLin28b 1 and 2 (Life Technologies, #4457308-s117291, and #4457308-s117292) were diluted to 3mg/ml in RNase-free water, mixed with Complex Buffer and Invivofectamine (Invivofectamine kit, Invitrogen #1377501), incubated for 30 minutes at 50°C, and dialyzed with FLOATALYZER (Thermo Fisher Scientific, #08607021) for 2 hours in PBS at 25C. Each 20g mouse received 0.7 mg/ml per dose. Each mouse was injected with 100ul of the prepared solution intraperitoneally at p14, retro-orbitally at p24, and through the tail vein at p31. Mice were monitored regularly for tumor growth and survival.

Statistical analysis

Data is presented as mean ± SEM, and Student’s t-test (two-tailed distribution, two-sample unequal variance) was used to calculate p values. Statistical significance is displayed as p < 0.05 (*) or p < 0.01 (**) unless specified otherwise. The tests were performed using Microsoft Excel.

Supplementary Material

01
02

HIGHLIGHTS.

  • LIN28B overexpression in the liver is sufficient to drive liver cancer development

  • MYC-induced liver tumors exhibit Lin28b overexpression and let-7 suppression

  • Lin28a/b deletion in mice reduces liver tumor burden and prolongs survival

  • Conditional Lin28b knockdown/deletion validate its tumor maintenance requirement

SIGNIFICANCE.

Liver cancer is the third most common cause of cancer death worldwide, but scientific and clinical understanding of it remains poor. Existing treatments fail to exploit a molecular or cellular understanding of the disease, and the best drug for hepatocellular carcinoma has a 3% response rate. Thus, understanding new mechanisms behind liver carcinogenesis is critical for the development of more effective treatments. Here we show that the RNA-binding protein LIN28B is sufficient to drive aggressive liver tumorigenesis in mice. Additionally, we show that both genetic deletion of Lin28a/b and siRNA therapy against Lin28b reduces tumor burden and prolongs survival, highlighting this oncogene as a potential therapeutic target for cancers that overexpress Lin28b.

ACKNOWLEDGEMENTS

We thank S.J. Morrison for providing the Ubiquitin-Cre/ERT2 mice, Joshua Mendell for providing the LAP-tTA and TRE-MYC mice, Roderick Bronson and the Harvard Medical School Rodent Histopathology Core, John Shelton and the UTSW Richardson Molecular Pathology Core for mouse tissue pathology and diagnostic interpretation. D.R. (Dinesh Rakheja) is supported by a Cancer Prevention and Research Institute of Texas grant (RP101195-C3) to study the genetics and biology of liver tumorigenesis in children. This work was supported by a grant from the Ellison Medical Foundation to G.Q.D. G.Q.D. is an investigator of the Howard Hughes Medical Institute and the Manton Center for Orphan Disease Research. This work was also supported by an American Cancer Society Postdoctoral Fellowship, an NIH K08 (1K08CA157727-02), a Burroughs Welcome Career Medical Award, and a Cancer Prevention and Research Institute of Texas New Investigator grant to H.Z.

Footnotes

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

AUTHOR CONTRIBUTIONS

L.N., D.A.R., M.T.S., and H.Z. performed the experiments. D.A.R., H.Z., and G.Q.D. wrote the manuscript. G.S., T.Y.D., and K.T. helped with various experiments and mouse husbandry. D.R. interpreted and wrote the descriptions of mouse liver histology. S.C. and R.H. provided the ApoE-rtTA transgenic mouse. S.R. and X.S. designed and performed the PET/CT experiments and interpreted the imaging data. G.Q.D. and H.Z. designed and supervised experiments.

The authors declare no competing financial interests.

REFERENCES

  1. Ambros V, Horvitz HR. Heterochronic mutants of the nematode Caenorhabditis elegans. Science. 1984;226:409–416. doi: 10.1126/science.6494891. [DOI] [PubMed] [Google Scholar]
  2. Balzer E, Moss EG. Localization of the developmental timing regulator Lin28 to mRNP complexes, P-bodies and stress granules. RNA Biol. 2007;4:16–25. doi: 10.4161/rna.4.1.4364. [DOI] [PubMed] [Google Scholar]
  3. Beachy SH, Onozawa M, Chung YJ, Slape C, Bilke S, Francis P, Pineda M, Walker RL, Meltzer P, Aplan PD. Enforced expression of Lin28b leads to impaired T-cell development, release of inflammatory cytokines, and peripheral T-cell lymphoma. Blood. 2012;120:1048–1059. doi: 10.1182/blood-2012-01-401760. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Beer S, Zetterberg A, Ihrie RA, McTaggart RA, Yang Q, Bradon N, Arvanitis C, Attardi LD, Feng S, Ruebner B, et al. Developmental context determines latency of MYC-induced tumorigenesis. PLoS Biol. 2004;2:e332. doi: 10.1371/journal.pbio.0020332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Belteki G, Haigh J, Kabacs N, Haigh K, Sison K, Costantini F, Whitsett J, Quaggin SE, Nagy A. Conditional and inducible transgene expression in mice through the combinatorial use of Cre-mediated recombination and tetracycline induction. Nucleic Acids Res. 2005;33:e51. doi: 10.1093/nar/gni051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Boyerinas B, Park SM, Hau A, Murmann AE, Peter ME. The role of let-7 in cell differentiation and cancer. Endocrine-related cancer. 2010;17:F19–F36. doi: 10.1677/ERC-09-0184. [DOI] [PubMed] [Google Scholar]
  7. Boyerinas B, Park SM, Shomron N, Hedegaard MM, Vinther J, Andersen JS, Feig C, Xu J, Burge CB, Peter ME. Identification of let-7-regulated oncofetal genes. Cancer Res. 2008;68:2587–2591. doi: 10.1158/0008-5472.CAN-08-0264. [DOI] [PubMed] [Google Scholar]
  8. Cairo S, Armengol C, De Reynies A, Wei Y, Thomas E, Renard CA, Goga A, Balakrishnan A, Semeraro M, Gresh L, et al. Hepatic stem-like phenotype and interplay of Wnt/beta-catenin and Myc signaling in aggressive childhood liver cancer. Cancer Cell. 2008;14:471–484. doi: 10.1016/j.ccr.2008.11.002. [DOI] [PubMed] [Google Scholar]
  9. Canberk S, Uludokumaci A, Sonmez C, Cakalir C, Gulsen F, Ozbay G. Mixed hepatocellular carcinoma and hepatoblastoma: cytohistopathologic findings and differential diagnosis. Acta cytologica. 2013;57:91–95. doi: 10.1159/000339275. [DOI] [PubMed] [Google Scholar]
  10. Chang TC, Zeitels LR, Hwang HW, Chivukula RR, Wentzel EA, Dews M, Jung J, Gao P, Dang CV, Beer MA, et al. Lin-28B transactivation is necessary for Myc-mediated let-7 repression and proliferation. Proc Natl Acad Sci U S A. 2009;106:3384–3389. doi: 10.1073/pnas.0808300106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chen LT, Lin LJ, Zheng LL. The correlation between insulin-like growth factor II mRNA binding protein 3 expression in hepatocellular carcinoma and prognosis. Hepato-gastroenterology. 2013a;60:553–556. doi: 10.5754/hge12855. [DOI] [PubMed] [Google Scholar]
  12. Chen YL, Jeng YM, Hsu HC, Lai HS, Lee PH, Lai PL, Yuan RH. Expression of insulin-like growth factor II mRNA-binding protein 3 predicts early recurrence and poor prognosis in intrahepatic cholangiocarcinoma. International journal of surgery (London, England) 2013b;11:85–91. doi: 10.1016/j.ijsu.2012.11.021. [DOI] [PubMed] [Google Scholar]
  13. Cheng S-W, Tsai H-W, Lin Y-J, Cheng P-N, Chang Y-C, Yen C-J, Huang H-P, Chuang Y-P, Chang T-T, Lee C-T, et al. Lin28B is an oncofetal circulating cancer stem cell-like marker associated with recurrence of hepatocellular carcinoma. PloS one. 2013;8:e80053. doi: 10.1371/journal.pone.0080053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Cho J, Chang H, Kwon SC, Kim B, Kim Y, Choe J, Ha M, Kim YK, Kim VN. LIN28A is a suppressor of ER-associated translation in embryonic stem cells. Cell. 2012;151:765–777. doi: 10.1016/j.cell.2012.10.019. [DOI] [PubMed] [Google Scholar]
  15. European Association for the Study of the Liver, E. O. f. R. a. T. o. C. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol. 2012;56:908–943. doi: 10.1016/j.jhep.2011.12.001. [DOI] [PubMed] [Google Scholar]
  16. Gu S, Jin L, Zhang Y, Huang Y, Zhang F, Valdmanis PN, Kay MA. The loop position of shRNAs and pre-miRNAs is critical for the accuracy of dicer processing in vivo. Cell. 2012;151:900–911. doi: 10.1016/j.cell.2012.09.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Guo Y, Chen Y, Ito H, Watanabe A, Ge X, Kodama T, Aburatani H. Identification and characterization of lin-28 homolog B (LIN28B) in human hepatocellular carcinoma. Gene. 2006;384:51–61. doi: 10.1016/j.gene.2006.07.011. [DOI] [PubMed] [Google Scholar]
  18. Hafner M, Max KE, Bandaru P, Morozov P, Gerstberger S, Brown M, Molina H, Tuschl T. Identification of mRNAs bound and regulated by human LIN28 proteins and molecular requirements for RNA recognition. RNA. 2013;19:613–626. doi: 10.1261/rna.036491.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Ishak KG, Glunz PR. Hepatoblastoma and hepatocarcinoma in infancy and childhood. Report of 47 cases. Cancer. 1967;20:396–422. doi: 10.1002/1097-0142(1967)20:3<396::aid-cncr2820200308>3.0.co;2-o. [DOI] [PubMed] [Google Scholar]
  20. Jeng YM, Chang CC, Hu FC, Chou HY, Kao HL, Wang TH, Hsu HC. RNA-binding protein insulin-like growth factor II mRNA-binding protein 3 expression promotes tumor invasion and predicts early recurrence and poor prognosis in hepatocellular carcinoma. Hepatology. 2008;48:1118–1127. doi: 10.1002/hep.22459. [DOI] [PubMed] [Google Scholar]
  21. Jin J, Jing W, Lei XX, Feng C, Peng S, Boris-Lawrie K, Huang Y. Evidence that Lin28 stimulates translation by recruiting RNA helicase A to polysomes. Nucleic Acids Res. 2011;39:3724–3734. doi: 10.1093/nar/gkq1350. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Johnson SM, Grosshans H, Shingara J, Byrom M, Jarvis R, Cheng A, Labourier E, Reinert KL, Brown D, Slack FJ. RAS is regulated by the let-7 microRNA family. Cell. 2005;120:635–647. doi: 10.1016/j.cell.2005.01.014. [DOI] [PubMed] [Google Scholar]
  23. Li N, Zhong X, Lin X, Guo J, Zou L, Tanyi JL, Shao Z, Liang S, Wang LP, Hwang WT, et al. Lin-28 homologue A (LIN28A) promotes cell cycle progression via regulation of cyclin-dependent kinase 2 (CDK2), cyclin D1 (CCND1), and cell division cycle 25 homolog A (CDC25A) expression in cancer. J Biol Chem. 2012;287:17386–17397. doi: 10.1074/jbc.M111.321158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Liang L, Wong CM, Ying Q, Fan DN, Huang S, Ding J, Yao J, Yan M, Li J, Yao M, et al. MicroRNA-125b suppressesed human liver cancer cell proliferation and metastasis by directly targeting oncogene LIN28B2. Hepatology. 2010;52:1731–1740. doi: 10.1002/hep.23904. [DOI] [PubMed] [Google Scholar]
  25. Mayr C, Hemann MT, Bartel DP. Disrupting the pairing between let-7 and Hmga2 enhances oncogenic transformation. Science. 2007;315:1576–1579. doi: 10.1126/science.1137999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. McMahon BJ. The influence of hepatitis B virus genotype and subgenotype on the natural history of chronic hepatitis B. Hepatology international. 2009;3:334–342. doi: 10.1007/s12072-008-9112-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Molenaar JJ, Domingo-Fernandez R, Ebus ME, Lindner S, Koster J, Drabek K, Mestdagh P, van Sluis P, Valentijn LJ, van Nes J, et al. LIN28B induces neuroblastoma and enhances MYCN levels via let-7 suppression. Nat Genet. 2012;44:1199–1206. doi: 10.1038/ng.2436. [DOI] [PubMed] [Google Scholar]
  28. Moss EG, Tang L. Conservation of the heterochronic regulator Lin-28, its developmental expression and microRNA complementary sites. Dev Biol. 2003;258:432–442. doi: 10.1016/s0012-1606(03)00126-x. [DOI] [PubMed] [Google Scholar]
  29. Nischalke HD, Schmitz V, Luda C, Aldenhoff K, Berger C, Feldmann G, Sauerbruch T, Spengler U, Nattermann J. Detection of IGF2BP3, HOXB7, and NEK2 mRNA expression in brush cytology specimens as a new diagnostic tool in patients with biliary strictures. PloS one. 2012;7:e42141. doi: 10.1371/journal.pone.0042141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Page JM, Harrison SA. NASH and HCC. Clinics in liver disease. 2009;13:631–647. doi: 10.1016/j.cld.2009.07.007. [DOI] [PubMed] [Google Scholar]
  31. Pant V, Sen IB, Soin AS. Role of 18F-FDG PET CT as an independent prognostic indicator in patients with hepatocellular carcinoma. Nuclear medicine communications. 2013;34:749–757. doi: 10.1097/MNM.0b013e3283622eef. [DOI] [PubMed] [Google Scholar]
  32. Peng S, Chen LL, Lei XX, Yang L, Lin H, Carmichael GG, Huang Y. Genome-wide studies reveal that Lin28 enhances the translation of genes important for growth and survival of human embryonic stem cells. Stem Cells. 2011;29:496–504. doi: 10.1002/stem.591. [DOI] [PubMed] [Google Scholar]
  33. Polesskaya A, Cuvellier S, Naguibneva I, Duquet A, Moss EG, Harel-Bellan A. Lin-28 binds IGF-2 mRNA and participates in skeletal myogenesis by increasing translation efficiency. Genes Dev. 2007;21:1125–1138. doi: 10.1101/gad.415007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Qiu S, Adema CM, Lane T. A computational study of off-target effects of RNA interference. Nucleic Acids Res. 2005;33:1834–1847. doi: 10.1093/nar/gki324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Roush S, Slack FJ. The let-7 family of microRNAs. Trends in cell biology. 2008;18:505–516. doi: 10.1016/j.tcb.2008.07.007. [DOI] [PubMed] [Google Scholar]
  36. Sampson VB, Rong NH, Han J, Yang Q, Aris V, Soteropoulos P, Petrelli NJ, Dunn SP, Krueger LJ. MicroRNA let-7a down-regulates MYC and reverts MYC-induced growth in Burkitt lymphoma cells. Cancer Res. 2007;67:9762–9770. doi: 10.1158/0008-5472.CAN-07-2462. [DOI] [PubMed] [Google Scholar]
  37. Shachaf CM, Kopelman AM, Arvanitis C, Karlsson A, Beer S, Mandl S, Bachmann MH, Borowsky AD, Ruebner B, Cardiff RD, et al. MYC inactivation uncovers pluripotent differentiation and tumour dormancy in hepatocellular cancer. Nature. 2004;431:1112–1117. doi: 10.1038/nature03043. [DOI] [PubMed] [Google Scholar]
  38. Shinoda G, Shyh-Chang N, de Soysa TY, Zhu H, Seligson MT, Shah SP, Abo-Sido N, Yabuuchi A, Hagan JP, Gregory RI, et al. Fetal Deficiency of Lin28 Programs Life-Long Aberrations in Growth and Glucose Metabolism. Stem Cells. 2013 doi: 10.1002/stem.1423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Shyh-Chang N, Daley GQ. Lin28: primal regulator of growth and metabolism in stem cells. Cell Stem Cell. 2013;12:395–406. doi: 10.1016/j.stem.2013.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Vadla B, Kemper K, Alaimo J, Heine C, Moss EG. lin-28 controls the succession of cell fate choices via two distinct activities. PLoS genetics. 2012;8:e1002588. doi: 10.1371/journal.pgen.1002588. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Visvader JE. Cells of origin in cancer. Nature. 2011;469:314–322. doi: 10.1038/nature09781. [DOI] [PubMed] [Google Scholar]
  42. Viswanathan SR, Daley GQ, Gregory RI. Selective blockade of microRNA processing by Lin28. Science. 2008;320:97–100. doi: 10.1126/science.1154040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Viswanathan SR, Powers JT, Einhorn W, Hoshida Y, Ng TL, Toffanin S, O’Sullivan M, Lu J, Phillips LA, Lockhart VL, et al. Lin28 promotes transformation and is associated with advanced human malignancies. Nat Genet. 2009;41:843–848. doi: 10.1038/ng.392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Wachter DL, Kristiansen G, Soll C, Hellerbrand C, Breuhahn K, Fritzsche F, Agaimy A, Hartmann A, Riener MO. Insulin-like growth factor II mRNA-binding protein 3 (IMP3) expression in hepatocellular carcinoma. A clinicopathological analysis with emphasis on diagnostic value. Histopathology. 2012;60:278–286. doi: 10.1111/j.1365-2559.2011.04091.x. [DOI] [PubMed] [Google Scholar]
  45. Wang YC, Chen YL, Yuan RH, Pan HW, Yang WC, Hsu HC, Jeng YM. Lin-28B expression promotes transformation and invasion in human hepatocellular carcinoma. Carcinogenesis. 2010;31:1516–1522. doi: 10.1093/carcin/bgq107. [DOI] [PubMed] [Google Scholar]
  46. Wilbert ML, Huelga SC, Kapeli K, Stark TJ, Liang TY, Chen SX, Yan BY, Nathanson JL, Hutt KR, Lovci MT, et al. LIN28 binds messenger RNAs at GGAGA motifs and regulates splicing factor abundance. Mol Cell. 2012;48:195–206. doi: 10.1016/j.molcel.2012.08.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wurmbach E, Chen YB, Khitrov G, Zhang W, Roayaie S, Schwartz M, Fiel I, Thung S, Mazzaferro V, Bruix J, et al. Genome-wide molecular profiles of HCV-induced dysplasia and hepatocellular carcinoma. Hepatology. 2007;45:938–947. doi: 10.1002/hep.21622. [DOI] [PubMed] [Google Scholar]
  48. Xu B, Zhang K, Huang Y. Lin28 modulates cell growth and associates with a subset of cell cycle regulator mRNAs in mouse embryonic stem cells. RNA. 2009;15:357–361. doi: 10.1261/rna.1368009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Yang DH, Moss EG. Temporally regulated expression of Lin-28 in diverse tissues of the developing mouse. Gene expression patterns : GEP. 2003;3:719–726. doi: 10.1016/s1567-133x(03)00140-6. [DOI] [PubMed] [Google Scholar]
  50. You X, Liu F, Zhang T, Lv N, Liu Q, Shan C, Du Y, Kong G, Wang T, Ye L, Zhang X. Hepatitis B virus X protein upregulates Lin28A/Lin28B through Sp-1/c-Myc to enhance the proliferation of hepatoma cells. Oncogene. 2013 doi: 10.1038/onc.2012.618. [DOI] [PubMed] [Google Scholar]
  51. Yu J, Vodyanik MA, Smuga-Otto K, Antosiewicz-Bourget J, Frane JL, Tian S, Nie J, Jonsdottir GA, Ruotti V, Stewart R, et al. Induced pluripotent stem cell lines derived from human somatic cells. Science. 2007;318:1917–1920. doi: 10.1126/science.1151526. [DOI] [PubMed] [Google Scholar]
  52. Zhang F, Chen XP, Zhang W, Dong HH, Xiang S, Zhang WG, Zhang BX. Combined hepatocellular cholangiocarcinoma originating from hepatic progenitor cells: immunohistochemical and double-fluorescence immunostaining evidence. Histopathology. 2008;52:224–232. doi: 10.1111/j.1365-2559.2007.02929.x. [DOI] [PubMed] [Google Scholar]
  53. Zhu H, Shah S, Shyh-Chang N, Shinoda G, Einhorn WS, Viswanathan SR, Takeuchi A, Grasemann C, Rinn JL, Lopez MF, et al. Lin28a transgenic mice manifest size and puberty phenotypes identified in human genetic association studies. Nat Genet. 2010;42:626–630. doi: 10.1038/ng.593. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zhu H, Shyh-Chang N, Segre AV, Shinoda G, Shah SP, Einhorn WS, Takeuchi A, Engreitz JM, Hagan JP, Kharas MG, et al. The Lin28/let-7 axis regulates glucose metabolism. Cell. 2011;147:81–94. doi: 10.1016/j.cell.2011.08.033. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS610967-supplement-01.pdf (16.8MB, pdf)
02
NIHMS610967-supplement-02.pdf (73.2KB, pdf)

RESOURCES