Usp18 Regulates Epidermal Growth Factor (EGF) Receptor Expression and Cancer Cell Survival via MicroRNA-7*

Epidermal growth factor receptor (EGFR) is involved in development and progression of many human cancers. We have previously demonstrated that the ubiquitin-specific peptidase Usp18 (Ubp43) is a potent regulator of EGFR protein expression. Here we report that the 3′-untranslated region (3′-UTR) of the EGFR message modulates RNA translation following cell treatment with Usp18 siRNA, suggesting microRNA as a possible mediator. Given earlier evidence of EGFR regulation by the microRNA miR-7, we assessed whether miR-7 mediates Usp18 siRNA effects. We found that Usp18 depletion elevates miR-7 levels in several cancer cell lines because of a transcriptional activation and/or mRNA stabilization of miR-7 host genes and that miR-7 acts downstream of Usp18 to regulate EGFR mRNA translation via the 3′-UTR. Also, depletion of Usp18 led to a decrease in protein levels of other known oncogenic targets of miR-7, reduced cell proliferation and soft agar colony formation, and increased apoptosis. Notably, all of these phenotypes were reversed by a specific inhibitor of miR-7. Thus, our findings support a model in which Usp18 inhibition promotes up-regulation of miR-7, which in turn inhibits EGFR expression and the tumorigenic activity of cancer cells.

Despite advances in cancer treatments, improvement of overall patient survival remains poor. One of the primary reasons for this low success is the inherent complexity of oncogenic pathways such as those driven by the epidermal growth factor (EGF) 5 receptor (EGFR). EGFR is a member of the receptor tyrosine kinase (RTK) family. Upon binding of ligand, EGFR autophosphorylates at tyrosine residues and triggers an intracellular signal transduction cascade, which ultimately promotes cellular survival and division (1,2). Dysregulation of EGFR and downstream signaling events is found in a multitude of tumor types (3)(4)(5)(6). Therefore, targeting the tyrosine kinase activity of EGFR with small molecule inhibitors or targeting EGFR with antibodies has been a focus in the treatment of several tumors, including brain (glioblastoma), cervical, lung, and head and neck (squamous cell carcinoma). However, this strategy has resulted in minimal success. A major limitation of these approaches is that tumor cells eventually develop resistance to the current therapeutics. The resistance develops through increased ligand expression, additional somatic mutations in the EGFR tyrosine kinase domain, and increased heterodimerization with other RTKs (3,(7)(8)(9).
As an alternative to developing approaches to directly inhibit EGFR signaling, our recent efforts focused on identifying allosteric modulators of EGFR protein levels. Inhibition of these modulators has the potential to significantly decrease EGFR protein levels irrespective of ligand levels or EGFR mutational status. Using a library of small interfering RNAs (siRNAs) that target deubiquitinase enzymes (DUBs), a class of proteins known to regulate receptor trafficking and expression (10 -12), we identified a number of candidate proteins which regulate EGFR protein levels. One of these candidates is Usp18 (Ubp43) (13). Ubiquitin specific peptidase 18 (Usp18) is a cysteine protease which has been shown to remove ubiquitin and the ubiquitin-like molecule interferon stimulated gene 15 (ISG15) from substrates (14,15). siRNA knockdown of Usp18 resulted in a 50 -90% reduction in EGFR protein levels in a variety of cancer cell lines (13). Interestingly, this decreased synthesis occurs despite no change to EGFR mRNA levels (13). Such an observation hints that Usp18 regulation of EGFR protein occurs through EGFR 3Ј-and/or 5Ј-untranslated regions, suggesting the involvement of microRNAs (miRNAs) (16 -18). In fact, miRNAs miR-128a,b (19) and miR-7 (20) have been shown to regulate EGFR.
miRNAs are a class of noncoding RNAs that regulate protein expression by binding to the 3Ј-UTR of mRNA targets (17,18). They play critical roles in controlling cellular processes such as proliferation, apoptosis, development, and differentiation (16,17,20,21). miRNAs are first transcribed in the cell nucleus as long primary transcripts (pri-miRNAs), typically several hundred nucleotides long, and then capped, spliced, and polyade-nylated (22). These transcripts are processed in the nucleus by the ribonuclease enzyme Drosha into a precursor pre-miRNA which is about 70 nucleotides in length (16 -18, 22). The pre-miRNA is exported to the cell cytosol and further processed by the enzyme Dicer to 19 -23 nucleotide miRNA. The resultant siRNA-like mature miRNA molecule is incorporated into the RISC complex, where it directs mRNA translational inhibition and/or degradation (16 -18, 22).
In the present study, we have identified the mechanism by which Usp18 controls EGFR down-regulation. We found that Usp18 knockdown leads to increased miR-7 levels as a result of increased transcriptional activation and/or mRNA stabilization of miR-7 host genes, mediating the effect on EGFR expression and other known oncogenic targets of miR-7. This is the first study which demonstrates a role for a deubiquitinase enzyme in the regulation of a miRNA. Furthermore, we determined that tumor cells depleted of Usp18 undergo apoptosis through the activation of miR-7. These data suggest that inhibiting Usp18 may serve as a means of activating miR-7 and ultimately as a therapy for tumors with dysregulated EGFR.

MATERIALS AND METHODS
Cell Culture-Glioma cell lines U87MG and T98G, and the cervical cell line, HeLa, were acquired from American Type Culture Collection (ATCC). Head-and-neck squamous cell carcinoma UMSCC2 (referred to in this study as SCC2) cells originated from Dr. T. Carey (University of Michigan). All cell lines were grown under previously described conditions (13,23).
MicroRNA-7 and Control Inhibitor-Control and miR-7 specific inhibitor were obtained from Dharmacon (001005-01 and 300546-08, respectively). Briefly, the miR-7 inhibitor is a chemically engineered hairpin small RNA that includes a sequence that is perfectly complementary to the miR-7 sequence (24-UGGAAGACUAGUGAU UUUGUUGU-46) and is modified by the addition of a 2Ј-O-methyl group. The perfect complementarity allows it to bind specifically to and inhibit only miR-7. The control miR hairpin inhibitor is designed similarly to the miR-7 inhibitor, but its sequence is based on two Caenorhabditis elegans miRNA sequences that have been confirmed to have minimal sequence identity with miRNAs in rats, mice and humans (cel-miR-67 and cel-miR-239b: UCACAAC-CUCCUAGAAAGAGUAGA and UUGUACUACACAAAA-GUACUG) (24,25).
Plasmid Construction-Wild-type Usp18 plasmid used in this study encoded mouse Usp18 fused to an HA epitope tag and cloned into a pcDNA3.1 background (Addgene plasmid 12454). The mutant, enzymatically inactive C61S mutant (15) plasmid was generated from the wild-type construct above using site-directed mutagenesis with the forward caacatcggacagacgtcttgccttaactccttgcttc and reverse gaagcaaggagttaaggcaagacgtctgtccgatgttg primers. miR-7-reporter plasmid and EGFR 3Ј-UTR reporter plasmid were constructed as previously described (20). The native EGFR-3Ј-UTR contains three sites for miR-7 with 1-7 and 1-8 seed matches (20). Mutations in these regions render the 3Ј-UTR unresponsive to treatment with miR-7 (20,27).
Usp18 Rescue of Usp18 siRNA-HeLa cells were transfected with 1 g of pcDNA3.1, pHA-Usp18 Wt , or pHA-Usp18 C61S plasmids using TransIT-LT1 lipid reagent as per manufacturer protocol (Mirus Bio, Madison, WI). After 24 h they were then transfected with 30 nM control or Usp18 siRNA using Lipofectamine 2000 (Invitrogen). Total time of exposure to DNA was 72 h while exposure to siRNA was 55 h. T98 cells were transfected with control or Usp18 siRNA #5 together with either 0.5 g of pcDNA3.1, pHA-Usp18Wt, or pHA-Usp18C61S using Fugene HD (21,28). Cell extracts were performed as detailed previously (13) and subjected to SDS-PAGE and immunoblot analysis.
Real-time PCR-The mRNA levels of Usp18 and EGFR were determined as previously described (13). Briefly, T98, HeLa, and SCC2 cells were treated with control or Usp18 siRNA (50 -75 nM) using Dharmafect 2 transfection reagent. At 72 h post-transfection, total RNA was isolated using Qiagen RNeasy Mini protocol. Samples were analyzed using an Applied Biosystems (Foster City, CA) 7900HT and Taqman Expression Assays (EGFR: Hs00193306_m1; Usp18: Hs00276441_m1; 18 S ribosomal RNA: 4308329). Usp18, luciferase, and EGFR mRNA qPCR values were normalized to 18 S ribosomal RNA qPCR values in each sample. For the determination of the levels of miR-7 and its host genes, cells were transfected with either control or Usp18 siRNA and lysed using Qiazol and QIAshredder columns (Qiagen). RNA was isolated using the miRNeasy kit and RT-PCR performed using miScript (Qiagen). cDNA was generated and quantitative real-time PCR analysis for miR-7 and U6B were performed using miR-7 and U6B specific forward primers and a universal reverse primer (Qiagen). For Usp18, pituitary gland specific factor 1a (PGSF1a), pri-miR-7-2 and heterogeneous nuclear ribonucleoprotein K (hnRNPK), specific primers were used. All samples were analyzed with an Applied Biosystems (StepOnePlus) realtime PCR system and normalization to U6B small RNA (28). Data analysis was carried out using StepOne Software v2.1 (Applied Biosystems).
Determination of Cell Numbers (Cell Count)-Cells transfected with either control or Usp18 siRNA together with either control or miR-7 specific inhibitor were cultured for 2-3 days. The cells were trypsinized, washed with 1ϫ PBS, and their numbers determined with microscopy (Carl Zeiss, Axiovert 40 C, Thornwood, NY) via hemocytometer.
Caspase-3/7 Assay-40 l of cells transfected with either control or Usp18 siRNAs together with either control or miR-7 specific inhibitor resuspended in 1ϫ PBS were mixed with 40 l of caspase-Glo 3/7 reagent (Promega), and luciferase activity was measured as previously described using a luminometer and normalized to protein concentration (23,28).
Annexin V Assay-Cells were transfected with control or Usp18 siRNA as well as control or miR-7 specific inhibitor. After 72 h, cells were treated with annexin V and propidium iodide (PI) reagent as per manufacturer's instructions (BD Pharmingen, San Diego, CA). The number of annexin V-positive and PI-negative cells is presented as a percent of total number of cells as determined by flow cytometry.
Luciferase Assay-Luciferase reporter assays were performed as previously described on a Promega Glomax 20/20 luminometer (23). EGFR 3Ј-UTR-luciferase or miR-7-reporter activities were normalized by dividing luciferase activity in each well by ␤-galactosidase or Renilla luciferase activity. Fold change in luciferase output was computed as previously described (23,28).
Soft Agar Colony Forming Assay-A 0.6% agar/medium base layer was made and 3 ml was added to each well (6-well plate). Culture medium (1 ml) containing cells transfected with either control or Usp18 siRNA was mixed with 1 ml of 0.6% agar/ medium, poured on the base layer, and placed in an incubator at 37°C and 5% CO 2 . The number of colonies was counted as previously described (21,28).
Statistical Analysis of Data-Data presented in figures as graphs are always from a minimum of three independent experiments and expressed as a mean Ϯ S.E. For calculating statistical significant differences between groups of data, Student's two-way pair or unpaired t test were used.

Usp18-dependent Regulation of EGFR Protein
Occurs through the 3Ј-UTR of EGFR mRNA-We previously showed that Usp18 siRNA-mediated depletion of Usp18 in SCC2 and COS-1 cells leads to a decrease in EGFR protein synthesis despite eliciting no changes to EGFR mRNA (13). These observations strongly suggest that depletion of Usp18 in these cells leads to suppression of EGFR mRNA translation. To investigate this effect in additional cell types and further study the mechanism we employed several other cancer cell lines with various levels of EGFR expression, including the established glioma line T98 and the human cervical cancer cell line HeLa (3)(4)(5). Transfection of T98 and HeLa cells with two different siRNA oligonucleotides effectively decreased Usp18 mRNA (supplemental Fig. S1, A, B, and D), as previously observed in SCC2 cells (13). Furthermore, we confirmed that Usp18 siRNA treatment of T98 cells led to depletion of the Usp18 protein ( Fig. 1A and supplemental Fig. S1C). We then compared EGFR protein levels in T98 and HeLa cells depleted of Usp18 using the above tested siRNAs. Immunoblot analysis demonstrated a dramatic decrease in EGFR protein levels in Usp18 siRNA-transfected T98 (Fig. 1, A, D, and E) and HeLa (Fig. 1, B and C) cells as compared with control siRNA-transfected cells.
The decrease in EGFR protein levels as a result of Usp18 siRNA treatment of Hela and T98 cells could be rescued with the simultaneous transfection of a wild-type mouse Usp18 construct that is insensitive to siRNA targeting human Usp18 (Fig.  1, B-E). Such an observation confirms the specificity of the siRNA effects. To test whether the catalytic activity of Usp18 is necessary for this rescue effect, enzymatically inactive mouse Usp18 was co-transfected with Usp18 siRNA. As shown in Fig.  1, B-E this Usp18 mutant did not restore EGFR levels in either Hela or T98 cells. These data suggest that the effect of Usp18 on EGFR protein expression is dependent on the ability of Usp18 to remove ISG15 or ubiquitin from substrates.
Despite EGFR protein levels decreasing following Usp18 depletion there was little to no change in EGFR mRNA levels (supplemental Fig. S1, A and B). This finding in T98 and Hela cells is consistent between two different Usp18 siRNA oligonucleotides and with observations made previously in SCC2 and COS-1 cells (13). This strongly suggests that Usp18 regulates EGFR protein levels by controlling the process of EGFR mRNA translation. Interestingly, Usp18 siRNA #5 elicited a 25% decrease in EGFR mRNA, but only in HeLa cells (supplemental Fig. S1B). This oligo did not elicit the same phenotype in T98 (supplemental Fig. S1A) and SCC2 (13) cell lines used in this study nor in U87 and COS-1 cell lines (data not shown). Therefore, an additional mechanism to down-regulate EGFR translation may occur in HeLa cells that involves mRNA cleavage (20,27).
To test the prediction that the regulation of EGFR translation occurs through the EGFR 3Ј-UTR, we measured the luciferase output of a pGL3-luciferase vector bearing the EGFR 3Ј-UTR (20). T98 and HeLa cells transfected with Usp18 siRNA exhibited a decreased luciferase output as compared with control siRNA (Fig. 1, F and G). Measurements of luciferase mRNA levels revealed no changes when we compared Usp18 siRNA to

MicroRNA-7 Mediates the Effects of Usp18 Knockdown
control siRNA-transfected cells (supplemental Fig. S1D). These data confirm that EGFR down-regulation by Usp18 depletion occurs through the EGFR 3Ј-UTR.
Usp18 Depletion Leads to miR-7 Up-regulation-miRNAs have been shown to regulate protein levels by binding to the 3Ј-UTR regions of their mRNA targets (18). Based on the above observation and our previous data showing that forced expres-sion of miR-7 in cells decreased EGFR levels (20), we hypothesized that the effects of Usp18 depletion occurred through the up-regulation of miR-7. Quantitative RT-PCR analysis of miR-7 levels in T98, HeLa, and SCC2 cells transfected with Usp18 siRNA revealed a 1.5-2-fold increase when compared with control siRNA (p Ͻ 0.05 and p Ͻ 0.01) ( Fig. 2A and supplemental Fig. S2A). Such increases are within the range of miR C and E, immunoblot of EGFR expression was quantified and normalized to tubulin and presented as bar graphs (*, p Ͻ 0.05 and **, p Ͻ 0.01) (E for T98 and C for Hela). F and G, normalized luciferase activity of an EGFR 3Ј-UTR luciferase reporter plasmid in T98 (F) and HeLa (G) cells 48 -72 h post-transfection with either control or Usp18 siRNA (#5) (**, p Ͻ 0.01 and *, p Ͻ 0.05). All bar graphs in this study illustrate data from a minimum of three independent experiments. changes that are considered to be capable of eliciting potent effects on mRNA translation (31). miR-7 activity was also analyzed using a miR-7 luciferase reporter to assess whether the changes in miR-7 levels correlated with changes in miR-7 activity (24,32). Treatment of T98 and HeLa cells with Usp18 siRNA resulted in significantly reduced miR-7 reporter luciferase output (Fig. 2B). This reduced luciferase output was completely reversed in the presence of a miR-7 specific inhibitor in both T98 and HeLa cells (Fig. 2B). Also, the miR-7 inhibitor reduced the basal activity of miR-7 (at 72 h p Ͻ 0.05), as indicated by an increased reporter output (supplemental Fig. S2B). Thus, HeLa and T98 cells express biologically active miR-7 under steady state conditions, and Usp18 depletion promotes an increase in both miR-7 expression and activity.
To further assess the involvement of Usp18 in the regulation of miR-7 expression, we compared the steady state levels of Usp18 and miR-7 in the cell lines used in this study. Our observations reveal that T98 and HeLa cells normally express high levels of Usp18 mRNA and corresponding low levels of miR-7 (Fig. 2C). In contrast, SCC2 cells express low levels of Usp18 mRNA and high levels of miR-7 (Fig. 2C). Thus, under steady state conditions in these cell lines lower Usp18 mRNA levels correlate with higher miR-7 levels, and vice versa. We did not observe this same correlation between Usp18 and EGFR mRNA (data not shown). However, this is consistent with the model that changes in Usp18 are affecting translation from EGFR mRNA and not changes in EGFR mRNA levels themselves. While analysis of a larger set of cell lines is necessary to definitively establish correlation between steady-state levels of Usp18 and miR-7, the observations made in these three cancer cell lines are consistent with the data above showing that depletion of Usp18 leads to increased levels of miR-7. Together the data strongly suggest that endogenous Usp18 is an important regulator of miR-7 expression and activity. miR-7 Mediates the Effects of Usp18 Depletion on EGFR Expression-Based on our previous work (13,20) and data above ( Fig. 2 and supplemental Fig. S2A), we hypothesized that the reduced levels of EGFR protein observed in cells lacking Usp18 is due to increased miR-7 expression (Figs. 1, 2 and supplemental Fig. S2A). We tested this hypothesis using miR-7specific inhibitors and analyzing EGFR expression in T98, HeLa, and SCC2 cells. Transfection of a miR-7-specific inhibitor in T98, HeLa, or SCC2 cells slightly increased the basal levels of the EGFR protein (Fig. 3, A and B). Importantly, a specific miR-7 inhibitor was able to completely rescue the decrease in EGFR protein expression associated with Usp18 depletion (Fig. 3, A and B).
To further confirm the increased miR-7 activity we analyzed the protein levels of other known targets of miR-7 such as IRS-1 and c-Raf-1 ((20, 27) and supplemental Fig. S3, A and B). Following treatment with Usp18 siRNA we observed a significant decrease in c-Raf-1 protein levels in T98 (supplemental Fig.  S3C), and a significant decrease in IRS-1 protein levels in SCC2

MicroRNA-7 Mediates the Effects of Usp18 Knockdown
and HeLa cells (supplemental Fig. S3, D-F). Additionally, the decrease in IRS-1 was rescued with addition of the miR-7-specific inhibitor (supplemental Fig. S3, E and F). Thus, Usp18 controls the protein levels of EGFR, Raf-1, and IRS-1 in a miR-7-dependent manner.
Knowing that the EGFR 3Ј-UTR is necessary for Usp18-dependent regulation of EGFR (Fig. 1, F and G and Ref. 13) and that this regulation is also dependent on activation of miR-7 ( Fig. 2 and supplemental Fig. S2A), we assessed the necessity of the EGFR 3Ј-UTR in the ability of the inhibitor to rescue Usp18 depletion phenotypes. We found that the Usp18 siRNAdependent decrease in EGFR 3Ј-UTR-luciferase output was completely abrogated by miR-7 inhibitor in both T98 and HeLa cells (Fig. 3C). These data implicate miR-7 and the EGFR 3Ј-UTR as mediators of the Usp18-dependent regulation of EGFR protein expression.
Usp18 Regulation of miR-7 Likely Occurs at the Transcriptional Level-The three primary transcripts of mature miR-7 in the human genome originate from an intergenic region and the intronic regions of two different genes. In general, parallel transcription of host genes and miR-7 expression is likely to occur and thus similar expression patterns would be expected. Such a phenomenon has been observed for several other microRNAs and their host genes (33)(34)(35). Our studies here compared the mRNA levels of miR-7 host genes, hnRNPK and PGSF1a, as well as the abundance of intergenic pri-miR-7-2 (a long nonprotein coding transcript). Quantitative RT-PCR analysis of T98, HeLa, and SCC2 cells depleted of Usp18 revealed significant increases in hnRNPK, PGSF1a, and pri-miR-7-2 mRNA levels relative to control siRNA cells (Fig. 4, A-C). For hnRNPK and pri-miR-7-2, the fold inductions were more pronounced in HeLa and T98 compared with SCC2 cells (Fig. 4, A and C); for PGSF1a the increase was greater in SCC2 and HeLa when compared with T98 cells (Fig. 4B). This variation suggests that Usp18 depletion increases the expression of miR-7 from all three loci and possibly through different mechanisms in the different cell lines.
Usp18 Depletion Significantly Reduces Cellular Proliferation-As seen above, Usp18 depletion activates miR-7 in a number of different tumor cell lines (Figs. 2 and 3 and supplemental Fig.  S2A). Previous studies from our laboratory showed that forced expression of miR-7 reduced the proliferation and viability of established glioblastoma cell lines including T98 cells (20). Therefore, we hypothesized that Usp18 depletion would pro- mote a miR-7-dependent reduction in cell proliferation in cancer cells. Transfection of T98 and HeLa cells with Usp18 siRNA significantly reduced cell densities as compared with control siRNA transfected cells (Fig. 5A). This reduction in cell numbers was significantly reversed upon the introduction of a miR-7 specific inhibitor (Fig. 5A). The inability of miR-7 inhibition to completely reverse the reduction in cell densities suggests that Usp18 controls cellular proliferation through miR-7 and another pathway(s).
Usp18 Depletion Induces Apoptosis of Cancer Cells through miR-7 Up-regulation-The reduced cell numbers in cells depleted of Usp18 could be due to a decrease in cell cycle progression or an increase in cell death. To investigate potential increases in apoptosis we examined the effects of Usp18 depletion on activation of pro-apoptotic enzymes, caspases 3 and 7, and annexin V staining (36). Depletion of Usp18 significantly increased the combined activity of caspases 3 and 7 in T98 and HeLa cells (Fig. 5B). The addition of the miR-7-specific inhibitor completely reversed caspase-3/7 activation following Usp18 depletion (Fig. 5B). Similarly, Usp18 knockdown increased annexin V staining (Fig. 5C) while simultaneous inhibition of miR-7 almost completely negated this increase in apoptosis. These observations demonstrate that Usp18 depletion induces apoptosis in T98 and HeLa cells in a miR-7-dependent manner.
Effects of Usp18 Depletion on Cell Growth in Soft Agar-Because Usp18 depletion increases apoptosis and decreases the number of T98 and HeLa cells, we tested whether it also suppresses the tumorigenic activity of these cell lines. To investigate this possibility we quantified the ability of these cell lines to form colonies in soft agar, a well established in vitro tumorigenicity assay. Usp18 knockdown significantly reduced the number of colonies formed by both HeLa and T98 cells (Fig. 6). These data suggest that Usp18 may function as an oncogene and its inhibition may suppress proliferative and metastatic properties of tumor cells.

DISCUSSION
We show here that the mechanism of EGFR down-regulation following Usp18 depletion is dependent on the induction of miR-7 activity. Additionally, tumor cells depleted of Usp18 undergo apoptosis as a direct result of miR-7 activation and also exhibit a decreased ability to form colonies in soft agar, suggesting that Usp18 may have oncogenic properties. This is the first study to date demonstrating a role for Usp18 in the regulation of microRNAs and cancer cell apoptosis. Such observations not only provide additional insight into the complex mechanisms of EGFR and miR-7 regulation, but also suggest that Usp18 inhibition has potential for cancer therapeutics.
The determination that enzymatically inactive Usp18 could not rescue the EGFR down-regulation phenotype associated with Usp18 knockdown strongly suggests that the activity of Usp18 is critical to this function. In contrast, studies linking Usp18 to interferon signaling and viral response have shown that the activity of the enzyme is irrelevant (37). Thus, Usp18 may have varied roles in distinct cellular processes but it appears that the inhibition of its activity will affect only some of these processes, such as EGFR expression. Such a point is critical when evaluating any enzyme for therapeutic inhibition.
Usp18 is generally considered an enzyme which removes ubiquitin-like ISG15 from substrates (15,38), although it has also been shown to remove ubiquitin from substrates in vitro (14,39). Currently, the association between cancer and ISG15 is still quite uncertain. Some studies have observed altered levels of ISG15 in cancer cells (40 -42). However, it remains unclear as to whether increased ISG15 levels are a cause or effect of tumorigenesis. Clearly, future work investigating substrates of Usp18 and the ISG15 conjugation system will be helpful to the field of cancer.
We also show in this study that Usp18 depletion elicits a dramatic increase in miR-7 levels, which in turn affects cell viability and tumorigenicity. Earlier studies also provided in vivo and in vitro evidence for the roles of Usp18 in cellular death. These studies showed that mice lacking Usp18 suffered brain injury (44), and Usp18 knockdown sensitized cells to bortezomib, interferon ␥, and TRAIL-induced apoptosis (43). Similarly, overexpression of miR-7 has previously been shown to reduce cellular invasiveness and the viability of cancer cells, implying reduced tumorigenicity in response to increased miR-7 (20). Additionally, increased miR-7 expression was also shown to reduce the levels of pro-survival proteins such as IRS-1, IGF-1R, PAK1, and Raf-1 (20,27,45,46). This suggests that Usp18 and miR-7 share similar pro-survival targets and thus, demonstrates that a decrease in Usp18 leads to increased miR-7, and in turn to reduced cell viability and tumorigenicity.
It was previously reported that increases in the levels of specific microRNAs may arise through increased transcriptional activation, mRNA stabilization of host genes, and/or increased processing of primary microRNA transcripts (46 -48). The observations made in this study strongly suggest that the Usp18 knockdown-dependent induction of miR-7 levels occurs through transcriptional activation and/or mRNA stabilization of miR-7 host genes. We observed that knockdown of Usp18 results in a significant increase in the expression of miR-7 host genes hnRNPK and PGSF1a, as well as the intergenic pri-miR-7-2. The fold increase in miR-7 host genes and intergenic pri-miR-7-2 was different for the different cancer cells. This suggests that the degree to which each genomic source contributes to mature miR-7 differs in each cell line. This intriguing observation suggests that the regulation of transcription and/or mRNA stabilization of miR-7 host genes and the subsequent increase in miR-7 by Usp18 depletion is a process that may involve several mechanisms. One possible mechanism is based on a recent study which suggested that the level of mature miR-7 is regulated at the transcriptional level by the transcription factor HoxD10 (46). However, we did not find changes in the protein levels of HoxD10 following Usp18 depletion FIGURE 5. Usp18 depletion induces apoptosis in cancer cells in a miR-7-dependent manner. A, bar graphs of cell counts performed 72 h after the transfection of T98 and HeLa cells with either control or Usp18 siRNA (#5) together with control miR-inhibitor or miR-7 specific inhibitor (**, p Ͻ 0.01 and *, p Ͻ 0.05). B, bar graphs of caspase-3 and -7 activity normalized with protein concentration in Hela and T98 cells 72 h after transfection with control or Usp18 siRNA (#5) together with control miR-inhibitor or miR-7 inhibitor (**, p Ͻ 0.01). C, bar graphs of annexin V-positive and propidium iodide-negative signals after the transfection of HeLa and T98 cells with Usp18 (#5 or #6) or control siRNA together with control miR-inhibitor or miR-7 specific inhibitor (***, p Ͻ 0.001; **, p Ͻ 0.01; and *, p Ͻ 0.05).
(supplemental Fig. S3D). Thus, different transcription factors or mechanism(s) may be involved in the transcriptional regulation and/or mRNA stabilization of miR-7 host genes in cells tested here. In the case of increased transcription, Usp18 would need to act directly or through an intermediary protein(s). Our previous study showed no YFP-Usp18 signal in the nucleus (13), an observation matched by cellular fractionation studies of endogenous Usp18 (supplemental Fig. S4). Thus, the mechanism for the increase in miR-7 host gene(s) would likely involve one or more targets of Usp18 traveling to the nucleus to influence transcription at the three miR-7 loci. In the case of mRNA stabilization in the cytoplasm Usp18 has the potential to function directly in that process. In both processes Usp18 enzymatic activity would be critical as inactive Usp18 did not rescue EGFR down-regulation following Usp18 knockdown.
This study has also shown that depletion of Usp18, and subsequent miR-7 up-regulation, leads to reduced cellular proliferation in two different cancer cell lines. It is important to point out that Usp18 and/or miR-7 are likely inhibiting multiple cel-lular processes and, in particular, the activity of multiple cell transformation proteins. For example, our studies and previous studies have shown that miR-7 has the ability to inhibit several pro-growth signaling proteins including EGFR, IRS-1, IGF-1R, PAK1, and Raf-1 (20,27,45,46). Furthermore, Usp18, and its substrate ISG15, have also been shown to control the stability of pro-growth proteins (49,50). Thus, it is likely that Usp18 activity is regulating cellular growth via a mechanism(s) that is additional to the Usp18/miR-7 pathway regulation of growth promoting proteins. This would explain the observation that miR-7 inhibition completely rescues Usp18 depletion-induced apoptosis but can only partially rescue Usp18 depletion-induced decreases in cell number.
In summary, the data presented here show that Usp18 is a potent regulator of miR-7 levels and activity. Depletion of Usp18 in numerous cancer cell lines leads to a dramatic increase in miR-7 activity, which in turn reduces the levels of EGFR and other oncogenic proteins. These changes ultimately lead to increased apoptosis and reduced cell numbers. Usp18 has now been identified, and miR-7 confirmed, as critical elements responsible for controlling cell growth and tumorigenicity of cancer cells.