Human cancers often exhibit attenuated microRNA (miRNA) biogenesis and global underexpression of miRNAs; thus, targeting the miRNA biogenesis pathway represents a novel strategy for cancer therapy. Here, we report that miR-26a enhances miRNA biogenesis, which acts as a common mechanism partially accounting for miR-26a function in diverse cancers including melanoma, prostate and liver cancer. miR-26a was broadly reduced in multiple cancers, and overexpression of miR-26a significantly suppressed tumor growth and metastasis both in vitro and in vivo, including melanoma, prostate and liver cancers. Notably, miR-26a overexpression was accompanied by global upregulation of miRNAs, especially let-7, and let-7 expression was concordant with miR-26a expression in cancer cell lines, xenograft tumors and normal human tissues, underscoring their biological relevance. We showed that miR-26a directly targeted Lin28B and Zcchc11—two critical repressors of let-7 maturation. Furthermore, we have demonstrated that Zcchc11 promoted tumor growth and metastasis, and it was prominently overexpressed in human cancers. Our findings thus provide a novel mechanism by which a miRNA acts as a modulator of miRNA biogenesis. These results also define a role of the miR-26a and Zcchc11 in tumorigenesis and metastasis and have implications to develop new strategies for cancer therapy.
MicroRNAs (miRNAs) are often deregulated in human malignancies and can function as either oncogenes or tumor-suppressor genes;1, 2 therefore, they are considered as emerging ‘hallmarks’ of cancer.3 Notably, human cancers often exhibit attenuated miRNA biogenesis and global underexpression of miRNAs,4, 5 highlighting the importance of miRNA biogenesis during the development of cancer. Indeed, critical components of miRNA biogenesis, including Dicer 1, Drosha, TARBP2 and XPO5, have been identified as haploinsufficient tumor suppressors.6, 7, 8, 9, 10 Therefore, there is considerable interest in understanding the mechanisms of miRNA biogenesis, and it is thought that targeting the miRNA biogenesis pathway may lead to the development of novel cancer therapies.11, 12, 13, 14
miRNA biogenesis consists of a series of biochemical steps, and disruption of any step could influence miRNA abundance and contribute to tumorigenesis. The best-studied repressor of miRNA biogenesis is Lin28, an RNA-binding protein initially identified in Caenorhabditis elegans, which has crucial and widespread roles in development and diseases.15, 16, 17, 18, 19 These roles are enacted at least partially through blocking the maturation of let-7,20, 21, 22 which is one of the most abundant miRNA families in mammals, and members of this family function as tumor suppressors by targeting multiple oncogenes, including RAS, HMGA2 and c-Myc.23, 24, 25, 26, 27 Mammals have two Lin28 homologs, Lin28A (also known as Lin28) and Lin28B, which have similar suppressive roles in let-7 biogenesis.28, 29, 30, 31 Given the importance of the Lin28/let-7 regulatory circuit in development, pluripotency and tumorigenesis, this circuit provides novel therapeutic opportunities for cancer treatment.22 However, the mechanisms controlling the Lin28/let-7 circuit remain largely unknown.
In the studies described herein, we discovered that miR-26a directly targets Zcchc11 and Lin28B, thereby disrupting the Lin28/let-7 circuit, enhancing miRNA biogenesis and consequently inhibiting tumorigenesis and metastasis of several cancers, including melanoma, prostate and liver cancers. We also defined Zcchc11 as an oncogene for the development of these cancers and established a direct link between Zcchc11 and human liver and colorectal cancers. These findings not only provide novel insights into the mechanism by which miRNAs mediate tumor suppression but also offer new targets for cancer treatments.
Downregulation of miR-26a in multiple cancer types
Previously, we identified miR-26a, but not miR-26b, as one of several miRNAs that are highly expressed in the mouse liver.32 Here, we sought to further determine the expression of miR-26a in human tissues. Quantitative real-time PCR (qRT–PCR) analysis revealed that miR-26a was universally expressed at high levels in normal human tissues (Figure 1a). Moreover, expression of miR-26a was dramatically decreased in human cancer cell lines when compared with the corresponding normal tissues/cell lines (Figure 1b). Then we determined miR-26a function in multiple cancer cell lines of diverse origins, including melanoma (A375, A2058, SK-MEL-5 and SK-MEL-28), prostate liver (DU145, LNCAP and PC-3) and liver cancer (HepG2, Huh 7 and SK-HEP-1). miR-26a overexpression, driven by transfection with its miRNA precursor (pre-26a) modestly, but significantly, inhibited cell proliferation, whereas knockdown of miR-26a expression by an antisense oligonucleotide (anti-26a) enhanced cell proliferation (Figures 1c and d; Supplementary Figure S1). These modest effects on tumor-cell growth may result from the short-term treatment of miR-26a overexpression/knockdown.
As another approach, we introduced miR-26a to three cancer cell lines—A375, DU145 and SK-HEP-1—through a retroviral vector MDH-PGK-eGFP 2.0 (henceforth referred to as MPG) with an insertion of a human miR-26a precursor sequence.33 This approach led to four- to sixfold increase in miR-26a in these three cell lines (Figure 1e). We found that this moderate upregulation of miR-26a was sufficient to suppress tumor-cell proliferation (Figure 1f) and to decrease the efficiency of colony formation in all the three cell lines (Figure 1g). Taken together, these results suggest that miR-26a has a broad role in suppressing the growth and transformation of multiple cancer types, namely melanoma, prostate and liver cancer.
miR-26a inhibits the growth and metastasis of human-tumor xenografts
To determine the potential roles of miR-26a in tumor growth and metastasis simultaneously in vivo, we used a modified mouse-xenograft system in which fewer cancer cells than are used in the classic xenograft system were injected subcutaneously into NOD-SCID (nonobese diabetic-severe combined immunodeficient) mice and sufficient time was allowed for spontaneous metastases to grow.34 We injected NOD-SCID mice with tumor cells (5 × 105) of diverse origins (melanoma (A375), and prostate (DU145) and liver cancers (SK-HEP-1)) transduced with either the MPG control vector or the MPG-miR-26a expression vector. Tumor growth was monitored regularly and spontaneous metastases in the liver and lung were examined when mice were euthanized. miR-26a-overexpressing tumors were smaller and weighed less than the control tumors (Figures 2a–f). A375 tumor growth was very rapid (Figure 2a) and no metastasis could be observed in either livers or lungs of xenograft mice (Supplementary Figures S2a and b). In contrast, DU145 tumor growth was slow (Figure 2c), and a few tiny metastases emerged in the lung of 50% control mice but not in mice injected with miR-26a-overexpressing cells, as evidenced by hemotaxylin–eosin staining (Supplementary Figures S2c–f), suggesting a role for miR-26a in tumor metastasis.
The probable function of miR-26a in tumor growth and metastasis was further supported by the analysis of SK-HEP-1-derived tumors. miR-26a overexpression not only significantly suppressed the growth of SK-HEP-1-derived tumors (Figures 2e and f) but also almost completely blocked spontaneous metastasis in both the liver and lung (Figures 2g–f; Supplementary Figures S2g–j). All control mice had many visible metastases in their livers and lungs (Figures 2g and h). However, only 17% of mice injected with miR-26a-overexpressing cells displayed metastases, and these metastases were less frequent and much smaller (Figure 2h). In line with pulmonary metastatic burden, mice injected with miR-26a-overexpressing cells maintained an almost normal lung:body weight ratio compared with a twofold increase in the ratio in control mice (Figure 2h). Concordantly, cell proliferation, as measured by staining for proliferating cell nuclear antigen, was much greater in the livers and lungs of control mice than in mice injected with miR-26a-overexpressing SK-HEP-1 cells (Figures 2i and j).
The above-described xenograft system was unable to generate spontaneous metastases using A375 melanoma cells because of their extremely rapid growth. We therefore determined the effects of miR-26a in melanoma metastasis using an experimental metastasis model. We injected NOD-SCID mice, via the tail vein, with A375 cells (1 × 106) transduced with either MPG-miR-26a expression or MPG control vector and evaluated metastasis 3 weeks after the injection. Melanomatic metastatic lesions, which were composed of solid white masses, were detected in the livers and lungs of all mice. However, control mice had much more lung and liver metastases than mice that received miR-26a-overexpressing cells (Figures 2k and l). Collectively, these results demonstrate that miR-26a represses the growth and metastasis of multiple cancer types, including melanoma, and liver and prostate cancers in vivo.
miR-26a significantly upregulates let-7 expression in human-tumor xenografts
Strikingly, expression of many miRNAs, especially members of the let-7 family, was significantly greater in miR-26a-overexpressing A375 tumors than in control tumors (Figure 3a and Supplementary Figure S3a). This led us to profile the expression of various miRNAs in three melanoma cell lines (A2058, A375 and SK-MEL-5) transduced with either the MPG control vector or MPG-miR-26a expression vector. We found that the expression of these miRNAs was also increased in miR-26a-overexpressing melanoma cells (Figure 3b and Supplementary Figure S3b). As global underexpression of miRNAs is a common feature of human cancer types, we hypothesized that the global increase in miRNA expression induced by miR-26a overexpression may be a common mechanism that partially accounts for the effects of miR-26a in the development of tumors of diverse origins.
Given that let-7 is a potent tumor suppressor in various cancer types and its expression was consistently upregulated by miR-26a in melanoma tumors and cell lines, we focused our attention on this family and found that let-7 was also upregulated in both DU145- and SK-HEP-1 miR-26a-overexpressing tumors relative to control tumors (Figures 3c and d). We next correlated the let-7 level with the miR-26a level by using a panel of 20 diverse normal human tissues. There was a strong positive correlation between let-7d and miR-26a levels (Figure 3e, left panel, R2=0.5372, P=0.0002), as well as let-7f and miR-26a levels (Figure 3e, right panel, R2=0.3795, P=0.0038), further underscoring their biological relevance.
Next, we examined the contribution of let-7 in miR-26a-mediated inhibition on tumor growth in vitro. To this end, we transfected miR-26a-overexpressing A375, DU145 and SK-HEP-1 cells by the let-7 family inhibitor (designated as let-7 FI), which can inhibit the expression of all let-7 family members. We found that inhibition of let-7 in these cells reversed the decreased growth caused by miR-26a overexpression (Figures 3f and g), indicating that the tumor-suppressor effects of miR-26a are mediated, at least partially, through the regulation of let-7.
Let-7 expression can be regulated at either the transcriptional or the post-transcriptional level. To clarify this, we assessed the expression of pri-let-7 in xenograft tumors. The expression of pri-let-7-a/f/d and pri-let-7g was unchanged obviously between miR-26a-overexpressing tumors and control tumors (Figure 3h), supporting a post-transcriptional regulation. As a control, the levels of mature miR-16a and pri-miR-16a exhibited no significant changes (Figures 3a–d and h). Collectively, these data indicate that miR-26a overexpression leads to the upregulation of let-7, which may function as a common mechanism underlying the inhibitory effect of miR-26a in tumorigenesis and metastasis.
Lin28B and Zcchc11 are directly targeted by miR-26a
Let-7 biogenesis could be post-transcriptionally regulated by Lin28 and Zcchc11 (zinc-finger CCHC domain-containing 11, also known as TUT4),28, 31 we thus pursued the potential relationship between Lin28/Zcchc11 and miR-26a. Interestingly, both human Lin28B and Zcchc11 are predicted as targets of miR-26a by public algorithms (TargetScan)35 (Supplementary Figure S4). To test it directly, we cloned the 3′-untranslated regions (UTRs) of Lin28B and Zcchc11 into the psiCheck2.2 luciferase expression vector and examined luciferase activity upon miR-26a overexpression. Ectopic expression of miR-26a significantly suppressed the luciferase activity of wild-type Lin28B and Zcchc11 3′-UTR, but not their mutant 3′-UTR, indicating that miR-26a suppresses Lin28B and Zcchc11 expression through direct binding to their 3′-UTR (Figure 4a). miR-26a overexpression inhibited the mRNA expression of both Lin28B and Zcchc11 in HEK293 cells (Figure 4b). Moreover, miR-26a overexpression inhibited the endogenous expression of both Lin28B and Zcchc11 proteins in cancer cell lines (Figure 4c) and xenograft tumors (Figure 4d). Collectively, these data demonstrate that Lin28B and Zcchc11 are direct targets of miR-26a.
Zcchc11 promotes tumorigenesis and metastasis
The importance of Lin28B in tumorigenesis has been extensively investigated; however, the role of Zcchc11 in cancer development remains poorly understood.36, 37 Moreover, direct evidence linking Zcchc11 to human cancer types has not yet been provided. We therefore focused on the function of Zcchc11 in tumorigenesis.
Initially, we measured the expression of Zcchc11 in various cancer cell lines. QRT–PCR and immunoblot analysis showed that Zcchc11 was broadly expressed (Figures 4e and f). Then we introduced specific siRNA against Zcchc11 into three cancer cell lines—A375, DU145 and SK-HEP-1—and found that Zcchc11 depletion impaired tumor-cell growth (Figure 5a and Supplementary Figure S5a). As another approach, we designed four short-hairpin RNAs against the human Zcchc11 gene and initially determined their efficiency in HEK293T cells (Supplementary Figure S5b). We then established A375, DU145 and SK-HEP-1 cell lines with stable Zcchc11 depletion by using two of these short-hairpin RNAs, sh-Zcc3 and sh-Zcc4 (Supplementary Figure S5c) and evaluated the common aspects of tumorigenesis and transformation in these cells. Zcchc11 depletion repressed proliferation (Figure 5b) and colony formation (Figures 5c–e) of A375, DU145 and SK-HEP1 cells. Transwell invasion assays demonstrated that Zcchc11 depletion and miR-26a overexpression strongly suppressed the invasive capacity of SK-HEP-1 cells (Figure 5f and Supplementary Figure S5d). Taken together, these data suggest that Zcchc11 promotes tumorigenesis and transformation in vitro.
Knockdown of Zcchc11 represses the growth and metastasis of human liver-tumor xenograft
As shown above, the SK-HEP-1 xenograft model is excellent for examining tumor growth and metastasis simultaneously. Therefore, we assessed the effect of Zcchc11 downregulation in this model and found that mice bearing Zcchc11-depleting SK-HEP-1 xenografts had smaller tumors that weighed less than those of control mice (Figure 5g). Notably, all control mice had many visible metastases in the livers and lungs, whereas only 25 and 37.5% mice injected with Zcchc11-depleted cells had liver and lung metastases, respectively, and these metastases were less frequent and much smaller (Figures 5h and i; Supplementary Figures S5e and f). In line with pulmonary metastatic burden, mice injected with Zcchc11-depleted cells exhibited significant decrease in lung:body weight ratio compared with the control mice (Figure 5i). Furthermore, tumors of control mice had a high level of cellular proliferation, as measured by proliferating cell nuclear antigen staining, in livers and lungs relative to mice injected with Zcchc11-depleted cells (Figures 5j and k). Taken together, these data demonstrate that Zcchc11 depletion has similar effects as miR-26a overexpression and suggest that Zcchc11 promotes tumorigenesis and metastasis in vivo.
Overexpression of miR-26a inhibits Lin28B and Zcchc11 in human-tumor xenografts
As shown in Figure 4, miR-26a directly targets Lin28B and Zcchc11. To further confirm this regulation in vivo, we assessed the expression of Lin28B and Zcchc11 using immunohistochemistry (IHC) in A375, DU145 and SK-HEP-1 tumor xenografts. This analysis revealed decreased Lin28B and Zcchc11 staining in miR-26a-overexpressing tumors as compared with control tumors (Figures 6a–c), consistent with an inhibitory role of miR-26a on their expression. Of note, IHC analysis also showed that Lin28B was primarily localized in the nucleus, whereas Zcchc11 was predominantly localized in the cytoplasm, consistent with previous reports.28, 36 These results demonstrate that miR-26a represses Lin28B and Zcchc11 in the development of cancer types in vivo.
We also performed IHC analysis on Zcchc11-depleting SK-HEP-1 tumor xenografts. As expected, Zcchc11-depleting tumors exhibited much lower Zcchc11 staining than that exhibited by control mice (Figure 6d). In line with reduced Zcchc11 staining, Zcchc11-depleting tumors exhibited decreased Lin28B staining (Figure 6d), probably resulting from an indirect effect of Zcchc11 inhibition on Lin28B expression. For example, reduced Zcchc11 leads to increased activities of let-7 and miR-26a, thereby decreasing the Lin28B expression.
The effects of miR-26a and Zcchc11 on tumor-cell proliferation and apoptosis
To explore whether the slowing of tumor growth induced by miR-26a overexpression was because of impaired cell proliferation, increased apoptotic cell death or both, we used immunohistochemical analysis to evaluate tumor-cell proliferation and apoptosis in xenografts (A375, DU145 and SK-HEP-1). For all xenografts, the percentage of proliferating cell nuclear antigen-positive cells was significantly lower in miR-26a-overexpressing tumors compared with control tumors (Figures 6a–c). In contrast, miR-26a-overexpressing tumors exhibited considerably more apoptotic bodies, as determined using TUNEL assay, than did control tumors (Figures 6a–c). As a direct target of miR-26a, Zcchc11 was speculated to have opposite effects on tumor-cell proliferation and apoptosis. Consistent with this idea, tumors derived from Zcchc11-depleted SK-HEP-1 cells exhibited reduced tumor-cell proliferation and increased apoptosis compared with control tumors (Figure 6d). These results demonstrate the opposing effects of miR-26a and Zcchc11 on tumorigenesis and indicate that this effect is associated with tumor-cell proliferation and apoptosis.
Upregulation of Zcchc11 in human liver and colorectal cancers
To determine the relevance of our experimental findings on Zcchc11 to human cancer patients, we examined Zcchc11 expression in primary human cancer types. qRT–PCR analysis revealed that the expression of Zcchc11 was increased at least 1.5-fold in 65% liver tumors (six out of nine paired hepatocellular carcinoma (HCC) samples) (Figure 7a). We further assessed the expression of Zcchc11 by IHC in another liver cancer cohort including nine pairs of liver cancer samples and matched peri-tumor tissues. This analysis revealed elevated Zcchc11 staining in seven out of nine tumor tissues as compared with adjacent nontumor tissues (Figure 7b). We also evaluated Zcchc11 expression in 10 paired samples of colorectal cancers with normal tissue through IHC and found that Zcchc11 was significantly upregulated in eight out of ten tumor tissues compared with adjacent nontumor colon tissues (Figure 7c). Notably, IHC analysis also showed that Zcchc11 was predominantly localized in the cytoplasm (Figures 7b and c), consistent with our IHC analysis on xenografts and previous reports.28, 36 Together, these data indicate that Zcchc11 has an oncogenic role in human liver and colorectal cancers, consistent with the results obtained from in vitro experiments.
miRNAs have been identified as the key regulators in tumorigenesis and several stages of metastasis.14, 38 Previous studies have demonstrated the suppressive role of miR-26a in tumor growth.39, 40, 41, 42 Our study has expanded these findings by revealing a specific function of miR-26a in metastasis. That is, miR-26a not only inhibits tumorigenesis but also strongly prevents spontaneous and distant metastasis, indicating that miR-26a is involved in the development of both primary tumors and metastatic disease, at least including melanoma, and prostate and liver cancers. This expansion of miR-26a function to metastasis is of clinical importance, as metastasis is responsible for >90% of cancer-associated mortality. Of note, miR-26a is universally expressed at high level in human tissues. Taken together, these data suggest that miR-26a represents a potential therapeutic target to treat both tumor growth and metastasis.
Here we show that overexpression of miR-26a results in increased expression of let-7, and knockdown of let-7 reverses the inhibitory effects of miR-26a overexpression in tumor-cell growth. Moreover, let-7 expression is concordant with miR-26a expression in normal human tissues, cancer cell lines and human-xenograft tumors, underscoring coordinated regulation of these two miRNAs. Let-7 has been regarded as a bona fide tumor suppressor, and accumulating evidence demonstrates its crucial roles in the development of cancer. For example, let-7 targets multiple oncogenes, including RAS, HMGA2 and c-Myc.23, 24, 25, 26, 27 Recently, let-7 has been shown to act in a metastasis-signaling cascade involving the RAF kinase inhibitory protein.43, 44 Thus, miR-26a-induced let-7 provides a conserved mechanism for the suppressive role of miR-26a during tumorigenesis and metastasis of diverse cancers. However, it is likely that miR-26a also directly represses distinct oncogenes within different cancer context,39, 42 which contributes to its function.
Mechanistically, miR-26a directly targets Lin28B and Zcchc11, which are critical repressors of the maturation of miRNAs, particularly let-7.21, 22 Previous studies have defined a regulatory loop consisting of Lin28 and let-7, in which Lin28A/Lin28B suppresses let-7 maturation, whereas let-7 directly targets them.21, 22 Recently, it has been shown that Lin28A is predominantly cytoplasmic and recruits Zcchc11 to pre-let-7 to induce 3′-uridylation, thereby blocking Dicer cleavage and let-7 maturation.36 Unlike Lin28A, Lin28B localizes primarily to the nucleus and blocks let-7 maturation by binding to pri-let-7 and interfering with Drosha, processing through a Zcchc11-independent mechanism.36 Our finding that miR-26a targets both Lin28B and Zcchc11 suggests that miR-26a may modulate Lin28/let-7 regulatory loop through either Zcch11-dependent (Lin28A) or Zcchc11-independent (Lin28B) mechanism.
Compared with Lin28B, much less is known about the function of Zcchc11 in cancer development.36, 37 Piskounova et al.36 suggested that Zcchc11 enhances the tumorigenicity and invasiveness of Lin28A-expressing cancer cells but has no effect on these properties in Lin28B-expressing cancer cell lines, indicating that the function of Zcchc11 in tumorigenesis depends on its uridyltransferase activity for let-7 maturation. In contrast, Blahna et al.37 showed that Zcchc11 promotes the proliferation of lung-cancer cells independent of its uridyltransferase activity. Here, we provide several lines of evidence to demonstrate the oncogenic function of Zcchc11 in melanoma, and prostate and liver cancers in vitro, and in liver xenograft-mouse models. Overall, Zcchc11 displays an opposite function to that of miR-26a in tumor development and metastasis, suggesting reciprocal regulation of these two genes. Furthermore, we reveal that Zcchc11 expression is consistently elevated in human liver and colorectal cancers, which establishes the first clinical link between Zcchc11 and human cancers. It will be of great interest for future investigations to explore the underlying mechanism for Zcchc11 function in tumorigenesis.
In summary, we find that miR-26a directly targets Lin28B and Zcchc11 and define the suppressive functions of miR-26a and Zcchc11 in tumor growth and metastasis. These findings reveal new insights into an miRNA-mediated tumor suppression through enhancing miRNA biogenesis, which may provide novel approaches for cancer treatments.
Materials and methods
Melanoma (A2058, A375, SK-MEL-5 and SK-MEL-28), hepatoma (HepG2, Huh7 and SK-HEP-1), prostate cell lines (DU145, LNCAP and PC-3), colon (HCT116(p53+) and HCT116 (p53−)), kidney (786-O), lung (A549), breast (MB-MDA-231) and acute myelocytic leukemia (K562) cell lines were obtained from American Type Culture Collection and cultured according to instructions. Normal human dermis fibroblasts were kindly provided by Dr J Wu (City of Hope).
RNA extraction and real-time PCR
Total RNA was extracted using Tri-Reagent (Molecular Research Center, Inc, Cincinnati, OH, USA) following the supplier’s instructions. miR-26a and let-7 expression levels in human tissues were measured with First-Choice Human Total RNA Survey Panel (Applied Biosystems, Foster City, CA, USA). For mRNA analysis, complementary DNA was synthesized using MMLV reverse transcriptase (Invitrogen, Carlsbad, CA, USA). For miRNA analysis, complementary DNA was synthesized using Superscript III reverse transcriptase (Invitrogen) as described previously.45 Both mRNA and miRNA quantitative real-time PCR analyses were performed with the Power SYBR Green PCR Master Mix (Applied Biosystems). Expression of mRNA and miRNA was normalized to actin and 5S RNA, respectively. Sequences of the qRT–PCR primers are provided in Supplementary Information.
miRNA and siRNA transfections
miRNA mimic, miRCURY LNA miR-26a inhibitor and let-7 family inhibitor were purchased from Ambion (Austin, TX, USA) and Exiqon (Vedbaek, Denmark), respectively. esiRNA human ZCCHC11 was purchased from Sigma (St Louis, MO, USA). Transfections of miRNA or siRNAs were performed using HiPerfect (Qiagen, Valencia, CA, USA) according to the manufacturer’s protocol.
Retroviral and lentiviral transductions
The human miR-26a-2 gene was amplified using PCR and directionally inserted into the MDH1-PGK-GFP 2.0 retrovirus vector (Addgene, cat. no. 11375). Retrovirus was generated and it infected cancer cells as described previously.33 A multiplicity of infection of 3–6 was used to generate miR-26a-overexpressing stable cell lines. For Zcchc11 knockdown, the pLKO.1 lentivirus vector (Addgene, cat. no. 8453) was used. Transductions with lentiviral particles were performed using 2 mg/ml of polybrene (Sigma), and positive cells were selected with puromycin (Sigma). A list of Zcchc11 short-hairpin RNA sequences is provided in the Supplementary Information.
Cells (5000 cells/well) were seeded in a 96-well plate and transfected with miR-26a precursors/inhibitors. Forty-eight hours post transfection, rates of cell proliferation were determined using CellTiter 96 AQueous Non-Radioactive Cell Proliferation Assay (MTS) (Promega, Madison, WI, USA).
Colony formation assay
miR-26a-overexpressing, Zcchc11-depleted stable cell lines and their corresponding controls were used for colony formation assay. Briefly, cells were seeded at low density and allowed to grow for 14 days. Colonies were photographed and counted.
Transwell invasion assays were performed according to the manufacturer’s protocol (BD Biosciences, San Jose, CA, USA). Briefly, cells (5 × 104) were added to the upper chamber of transwell inserts (BD Biocoat Matrigel 24-well invasion chamber) in serum-free medium containing 0.1% bovine serum albumin, and the number of cells that invaded into the lower chamber were stained with Diff-Quik stain and counted after 24 h of incubation at 37 °C with 5% CO2. The cell nucleus stained purple and the cytoplasm stained pink. Each experimental group had two replicates, and three fields in each replicate were randomly chosen for quantification of invasive cells.
Total cell lysis and western blot analysis were performed as described previously.46 Western blot analysis used the following antibodies: anti-Zcchc11 (1:1000 dilution), anti-Lin28B (1:1000 dilution) (Abcam, Cambridge, UK) and anti-β-actin (1:5000 dilution) (Sigma).
Luciferase reporter assays
The Luciferase-Zcchc11-3′-UTR and Luciferase-Lin28B-3′-UTR constructs were based on psiCHECK2 vector (Promega). The predicted miR-26a-binding site was mutated using the QuickChange II XL Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA). HeLa cells were plated in a 96-well plate and cotransfected with the luciferase reporter vector and 20 nM miR-26 mimic (Ambion) using HiPerfect (Qiagen). Firefly and renilla luciferase activities were analyzed by the Dual-Luciferase Reporter Assay System (Promega) 24 h after transfection.
To study the effect of miR-26a overexpression on tumorigenesis in vivo, A375, DU145 and SK-HEP-1 cells (5 × 105) transduced with either the miR-26a expression vector or the control vector were trypsinized, resuspended in sterile PBS and injected subcutaneously into the flank of female NOD-SCID mice. To study the effect of Zcchc11 depletion in tumorigenesis in vivo, SK-HEP-1 cells (5 × 105) transduced with either the Zcchc11-depleting vector or the control vector were used as described above. Tumor size was measured with Vernier calipers and tumor volumes were calculated according to the following formula: L × W2/2, where L is the length and W is the width of the tumor. After mice were euthanized, liver and lung metastases were examined and tumor samples were collected for RNA and protein extraction processes.
In vivo metastasis assay
miR-26a-overexpressing stable A375 cells generated by retrovirus transduction were used. Briefly, 1 × 106 A375 cells were trypsinized, resuspended in sterile PBS and injected into female NOD-SCID mice through the tail vein. Three weeks after injection, mice were euthanized. Livers and lungs were then harvested and tumor nodules were counted.
Human liver and colorectal samples
Total RNAs of nine liver tumors and their adjacent normal livers were obtained from the City of Hope National Medical Center (Duarte, CA, USA). Formalin-fixed, paraffin-embedded liver cancer sections and colorectal cancer sections were obtained from the Third Military Medical University (Chongqing, China) and Zhejiang University (Hangzhou, China), respectively.
Histology and immunostaining
Samples were formalin-fixed, paraffin-embedded and sectioned. Hemotaxylin–eosin staining was performed according to the standard histopathological procedures. The number of proliferating cells was determined by staining paraffin sections with the proliferating cell nuclear antigen-staining kit (Invitrogen) according to the manufacturer’s protocol. Apoptosis was evaluated by staining paraffin sections with In Situ Cell Death Detection kit, TMR red terminal deoxynucleotidyl-transferase-mediated dUTP nick-end labeling (TUNEL) system (Roche, Indianapolis, IN, USA) according to the manufacturer’s protocol. Slides were mounted with Vectashield mounting medium with DAPI (Vector Laboratories). Immunohistochemical staining was performed by the City of Hope Pathology Core according to the manufacturer’s protocol using the anti-Zcchc11 antibody (1:50 dilution, Sigma) and anti-Lin28B antibody (1:50 dilution, Abcam).
Results of experiments are displayed as mean±s.e.m. in the Figures. Two-tailed ANOVA assay and Student's t-test were used to determine differences between data groups (two-tailed; P<0.05 was considered statistically significant) unless otherwise noted.
We thank Drs Arthur Riggs and Rama Natarajan for reviewing the manuscript, Dr Kyle Sousa and other members of the Huang laboratory for their helpful discussions and Dr Keely Walker for editing the manuscript. This work was supported by the National Cancer Institute (P30 CA033572) and the American Cancer Society (RSG-11-132-01-CCE).
About this article
Supplementary Information accompanies this paper on the Oncogene website (http://www.nature.com/onc)
miR-26a promotes hepatocellular carcinoma invasion and metastasis by inhibiting PTEN and inhibits cell growth by repressing EZH2
Laboratory Investigation (2019)