UNC50 Prompts G1/S Transition and Proliferation in HCC by Regulation of Epidermal Growth Factor Receptor Trafficking

Background UNC50 has long been recognized as a Golgi apparatus protein in yeast, and is involved in nicotinic receptor trafficking in Caenorhabditis elegans, but little is known about UNC50 gene function in human biology despite it being conserved from yeast to high eukaryotes. Objectives We investigated the relation between UNC50 and human hepatocellular carcinoma (HCC) and the potential mechanisms underlying HCC development. Methods UNC50 mRNA expression patterns in 12 HCC and adjacent non-cancerous tissues determined using northern blotting were confirmed by real-time PCR in another 44 paired tissues. Microarray experiments were used to screen for global effects of UNC50 knockdown in the Hep3B cell line, and were confirmed by real-time PCR, western blotting, flow cytometry, and tetrazolium assay in both UNC50 overexpression and knockdown Hep3B cells. Results UNC50 expression levels were upregulated in HCC tissues in comparison with the adjacent non-cancerous tissues. UNC50 knockdown reduced mRNA levels of the downstream targets of the epidermal growth factor receptor (EGFR) pathway: cyclin D1 (CCND1), EGF, matrix metalloproteinase-7 (MMP7), aldose reductase-like 1 (AKR1B10), cell surface–associated mucin 1 (MUC1), and gastrin (GAST). Moreover, UNC50 influenced EGF, inducing cell cycle entry by affecting cell surface EGFR amounts. Conclusions UNC50 may plays some roles in HCC progression by affecting the EGFR pathway.


Introduction
Hepatocellular carcinoma (HCC) is one of the most malignant cancers worldwide, accounting for millions of deaths every year [1]. Although HCC has been intensively studied, the molecular basis underlying HCC progression remains largely elusive. DNA microarrays have aided in the discovery of novel genes that are differentially expressed in HCC in comparison to non-cancerous adjacent tissues [2]. However, many of these genes have not been investigated closely, and their functions in humans are largely unknown. Such genes should be characterized to uncover novel tumor markers, oncogenes, and therapeutic targets for HCC. To obtain this information, we must find the crosstalk between these genes and well-known pathways.
The overexpression of receptor tyrosine kinases (RTK) is a hallmark of many cancers with poor prognosis [3]. Epidermal growth factor receptor (EGFR) is the earliest known RTK family protein, and is overexpressed in a wide range of cancers, playing important roles in cell growth and survival [4]. Specific EGFR inhibitors, such as erlotinib, gefitinib, and the monoclonal antibody cetuximab are effective for treating cancer [5]. Unfortunately, the tumor cells of most HCC patients have developed other mechanisms to bypass the EGFR pathway [6], and erlotinib is inefficient for disease control in the majority of patients with advanced HCC [7,8]. Even so, both gefitinib and erlotinib prevented HCC development in different animal models of cirrhosis; thus, the EGFR pathway is a promising target for HCC prevention [9,10].
UNC50 is conserved in nearly all eukaryotic organisms and is located mainly in the Golgi apparatus membrane [11]. The genes abbreviated to "UNC" followed by a numeral were first discovered during screening for mutants with the same uncoordinated motor behavior phenotype in Caenorhabditis elegans, and were thus named. The protein products of these genes usually have no amino acid sequence or protein structure similarity. In C. elegans, mutants of the UNC50 gene are resistant to the acetylcholine receptor (AChR) agonist levamisole [12,13]. Moreover, it has been demonstrated that UNC50 plays a role in levamisole-sensitive nicotinic AChR (levi-AChR) rafficking [14] in C. elegans, indicating its crucial role in nerve signal transmission at neuron-muscle junctions. However, homologs for the UNC50 gene are conserved in most eukaryotic organisms, including yeast and plants, which do not express AChRs; in humans, UNC50 is ubiquitously expressed in cells outside the nervous system, suggesting that the UNC50 gene plays wider roles.
Our group first cloned and submitted the full-length sequence of human UNC50 mRNA to the National Center for Biotechnology Information database (GeneBank ID: AY017215.1 in December 2000). In this study, we illustrate that UNC50 is overexpressed in HCC, and aim to uncover one of the roles UNC50 plays in HCC progression and the potential underlying molecular mechanisms.

Tissue specimen collection
Fresh surgical specimens of HCC, which comprised tumor tissues and adjacent non-cancerous liver tissues, were obtained from 56 HCC patients at Zhongshan Hospital, Shanghai, China. All samples were immediately frozen in liquid nitrogen after surgery and then stored at -80°C for further analysis.

Ethics statement
Prior to surgery, patients signed written informed consent forms for surgery and were presented the following option: "I agree to donate my resected tissue samples to Zhongshan Hospital and relevant research groups for research." The tissue samples of those who had selected this option were stored for research purposes. The written informed consent forms for surgery were stored by Zhongshan Hospital. In addition, verbal informed consent was obtained and recorded by our research group from all patients during our telephone follow-up. The present study was approved by the Ethics Committee of the Fudan University, Shanghai, China.

Cell culture and transfection
The human HCC cell line Hep3B was purchased from ATCC and cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS; HyClone) in a humidified incubator (5% CO 2 ) at 37°C.
Plasmids were transfected into Hep3B cells with Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Puromycin (1 μg/ml) was added to the cells for two days to purify UNC50 knockdown cells; 200 μg/ml G418 was added to the cells for two weeks to purify UNC50 overexpression cells.

Cell cycle analysis
Cultured cells were trypsinized, washed with phosphate-buffered saline (PBS) twice, and then permeabilized with 0.05% Triton X-100 in PBS for 10 minutes. The permeabilized cells were stained with 10 μg/ml propidium iodide solution for 10 minutes in a dark box. The treated cells were analyzed by flow cytometry (FACSCalibur; BD Bioscience) to determine DNA content and cell cycle status. We collected 10,000 gated cells for each sample.
RNA preparation, reverse transcription, and quantitative real-time PCR Total RNA was extracted from tissues or cultured cells using TRIzol (Invitrogen) according the manufacturer's protocol. RNA (1 μg) was used for reverse transcription with ReverTra Ace (Toyobo) using random hexameric oligos at 42°C for 1 hour, and the complementary DNA products were appropriately diluted with Milli-Q water. To rule out contamination, additional reverse transcription reactions containing no RNA were prepared as the negative controls for real-time PCR. Quantitative real-time PCR was carried out using a SYBR qPCR Mix (Toyobo) in a 10-μl reaction volume with a LightCycler 480 II (Roche) in 384-well plates (Axygen). Each reaction was repeated in three wells. We confirmed the specificity of the real-time PCR products by melting curve analysis. All gene expression levels were normalized to β-actin.

Microarray analysis
We analyzed total RNA from UNC50 knockdown and control Hep3B cells with Agilent Gene Expression oligo microarrays. The results have been uploaded to the Gene Expression Omnibus (GEO) database (GEO accession number: GSE63322). Genes with an average negative or positive fold change of 1.5 times were analyzed further. Datasets were taxonomized with Fun-Net (http://www.funnet.info/). Gene interaction networks were drawn using STRING (http:// string-db.org/).

Northern blotting
Total tissue RNA (10 μg) was separated on 1.5% agarose gels containing 2.2 M formaldehyde and transferred onto Hybond XL membranes. The membranes were probed with α-32P-UTP (800 Ci/mmol; PerkinElmer)-labeled minus or plus strand-specific full-length UNC-50 riboprobes and exposed to a phosphorimager screen.

Western blotting
Cells were lysed in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer and boiled in water for 10 minutes. Approximately 20 μg total protein extracts were loaded and separated by SDS-PAGE, and then transferred onto nitrocellulose membranes (Millipore). The membranes were blocked with 5% non-fat milk in Tris-buffered saline for 1 hour at room temperature, and then incubated with specific antibodies against EGFR, phosphorylated EGFR (pEGFR), cyclin D1 (CCND1) (Cell Signaling Technology), β-actin, mycepitope (Sigma-Aldrich), and UNC50 (Abgent) at the appropriate dilutions at 4°C overnight. This was followed by incubation with a corresponding horseradish peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology) at room temperature for 1 hour. The membrane was immersed in chemiluminescence reagent (7SeaPharmTech) and exposed with G:BOX (Syngene).

Tetrazolium assay
Cells (1000 cells per well) were seeded onto 96-well plates (Falcon) and underwent tetrazolium (MTT) assay (Dojindo) every day. The absorbance of each well was measured 1 hour after incubation using a microtiter reader (Bio-Rad) at 530 nm.

Flow cytometry
Cultured cells (1 × 10 6 ) from each sample were trypsinized, resuspended in PBS, and fixed in 4% formaldehyde for 10 minutes at room temperature. Cells from each sample were divided into two portions for either extracellular or whole-cell staining. For the latter, cells were additionally permeabilized in chilled 90% methanol for 30 minutes on ice. Both portions were then incubated in 0.5% bovine serum albumin/PBS solution for 5 minutes twice. Subsequently, the cells were resuspended for 1 hour at room temperature away from light in 100 μl 1:10 diluted Alexa Fluor 488-conjugated EGFR antibody (Santa Cruz Biotechnology) recognizing the cell surface epitope of EGFR. The cells were washed by centrifugation in PBS twice, resuspended, and analyzed using a flow cytometer (FACSCalibur; BD Biosciences). We collected 10,000 gated cells from each sample.

Statistical analysis
Stata 10 statistical software (StataCorp LP) was used for statistical analysis. All data are expressed as the mean ± standard error of the mean. Statistical analysis of the data was performed using Student's t-test. P < 0.05 was considered significant. Materials and methods for metaanalyses have been included in S1 Text.

UNC50 was upregulated in HCC
We identified UNC50 as a potential upregulated gene in human HCC compared to the adjacent non-cancerous tissues during our initial screening for HCC-relevant genes by northern blotting (Fig. 1A). Quantitative PCR of UNC50 mRNA expression levels in the 44 paired HCC samples confirmed this. Overall, 20 of 44 (45.5%) HCC cases showed significant UNC50 upregulation, 22 of 44 (50%) showed no alteration, and only two of 44 (4.5%) showed reduced UNC50 (Fig. 1B). In agreement with this, western blotting revealed that UNC50 was detectable in eleven of 12 cancer tissues and in only six of 12 non-cancerous tissues (Fig. 1C). Following systematic review of 16 independent microarray experiments in the GEO database, our metaanalysis further showed that UNC50 was significantly upregulated in HCC tissues in comparison with the paired adjacent non-cancerous liver tissues (p = 0.005) (S1 Fig.). Begg's funnel plot was used to illustrate publication bias (S1 Fig.). The studies used in the meta-analysis have been included in S1 Table. The influence of UNC50 knockdown on gene expression patterns To gain insight into the role UNC50 plays in HCC progression, we used microarray analyses to identify indirect evidence of UNC50 gene function via the knockdown strategy in Hep3B cells. Hep3B cells transfected with the shRNA expression plasmids shR-554, shR-749, and shR-MOCK were purified with 1 μg/ml puromycin, and the total RNA from each cell was extracted and analyzed with oligo microarrays. Fig. 2A shows that UNC50 knockdown or overexpression successfully altered UNC50 levels in comparison with the control. Following statistical selection of the regulated transcripts in the group, we found that 94 genes were regulated. Among them, several altered genes are the downstream targets of the EGFR pathway, including CCND1, EGF, matrix metalloproteinase-7 (MMP7) [15], aldose reductase-like 1 (AKR1B10) [16], cell surface-associated mucin 1 (MUC1) [17], and gastrin (GAST) [18]. Table 1 lists the microarray data. The microarray results were confirmed by real-time PCR analysis of UNC50, CCND1, EGF, MMP7, AKR1B10, MUC1, and GAST (Fig. 2B) in UNC50 overexpression and knockdown Hep3B cells.

UNC50 activates EGFR pathway in a ligand-dependent manner
We assumed that UNC50 plays a role in the EGFR pathway. The EGFR pathway is crucial to cell cycle progression and proliferation. However, both total EGFR protein (Fig. 3A) and EGFR mRNA (data not shown) levels remained unchanged, suggesting that UNC50 acts in a posttranslational manner. Therefore, we evaluated EGFR pathway activity by detecting the phosphorylation levels of EGFR at tyrosine 1068 (pEGFR-1068) with immunoblot assays. As cellular pEGFR-1068 levels are sensitive to extracellular ligand stimulation (ligand-dependent) and cell-cell contact (ligand-independent), we seeded equal numbers of cells to reach similar confluence (approximately 40%) after adherence and starved the cells for 24 hours before the different treatments. As shown in Fig. 3A, the EGFR pathway activity of the cells was proportional to that of UNC50 expression levels when cultured in DMEM with 10% FBS for 8 hours. Moreover, such effects were enhanced when 1 ng/ml EGF was added. However, we were unable to detect the pEGFR-1068 levels in cells cultured in serum-free medium, indicating that ligandindependent activation was scarce. From this, we may infer that UNC50 affects EGFR pathway activity in a ligand-dependent manner.
UNC50 prompted cell cycle entry and proliferation through the EGFR pathway CCND1 is a key component of the G1/S checkpoint and is a downstream target of the EGFR pathway. As both CCND1 mRNA (Fig. 2B) and protein levels (Fig. 3) were altered in accordance with UNC50 levels, we examined the cell cycle distribution of cells in which UNC50 expression levels had been modified. Cells were synchronized in the G0 phase through serum starvation before treatment. Following release from the G0 phase by serum stimulation, more cells were arrested in G0/G1 phase after UNC50 knockdown, and vice versa; the addition of EGF further enhanced these differences (Fig. 4A). Erlotinib, a specific EGFR inhibitor, countered the effects of EGF stimulation, indicating that the role of UNC50 in the cell cycle relies on, at least partly, a functional EGFR pathway.
To support our cell cycle results, we evaluated the cell proliferation states using the MTT assay. No significant difference was observed under normal conditions (10% FBS in DMEM). Constantly supplying low levels of EGF to the cultured cells (approximately 1 ng every 12 hours) greatly enhanced the corresponding differences among the cells, further supporting our earlier results (Fig. 4B).

UNC50 affected cell surface EGFR amounts
As UNC50 regulates nicotinic AChR trafficking to the cell membrane [14], we hypothesized that UNC50 acts similarly with EGFR. To test this, we stained cells with fluorescence-labeled EGFR antibodies with or without cell membrane permeabilization and compared cell surface EGFR amounts using flow cytometry. To eliminate the influence of ligands on EGFR translocation, we cultured the cells in serum-free medium for 24 hours before staining. As expected, the fluorescence intensity of the non-permeabilized stained cells was weaker and proportional to UNC50 expression levels (Fig. 5A, B); the fluorescence intensity of the permeabilized cells was stronger and not correlated with UNC50 expression levels (Fig. 5C, D). Moreover, the immunofluorescence experiment showed that UNC50 knockdown dramatically altered EGFR location (Fig. 5E). We could not visualize the EGFR changes by immunofluorescence following UNC50 overexpression (Fig. not shown) because the effects of UNC50 overexpression (Fig. 5B, D) on EGFR translocation were much weaker than that of UNC50 knockdown (Fig. 5A, C). These data show that UNC50 may be crucial for EGFR distribution without much change in total cellular EGFR.

Discussion
To our knowledge, this is the first study on UNC50 function in HCC and the second report on UNC50 in humans. The first study of the UNC50 gene in humans illustrated the role UNC50 may play in mechanical stress in periodontal tissues [19]. mRNA expression levels are normalized to β-actin; log2-transformed fold changes of HCC tissues compared to the adjacent non-cancerous tissues were calculated by the comparative cycle threshold (ΔΔCt) method. The paired tissues are ordered from low to high ΔΔCt values. Cutoff values were set to ±1 to identify the significance of changes. (C) Western blotting analysis of UNC50 levels in another 12 paired HCC tissues; βactin was used as the reference.  In this study, we found that UNC50 is upregulated in HCC. The expression levels of many genes are dramatically altered in HCC, but we are ignorant of their functions therein. Uncovering their roles in HCC may help us understand the molecular basis underlying HCC and thus provide novel therapy targets for HCC prevention and/or therapy, and UNC50 might be one such target.
It has been reported that UNC50 is a Golgi apparatus membrane protein [11]. Very few Golgi apparatus-related proteins are reported as being related to HCC progression despite the Golgi apparatus being a crucial cellular apparatus. GP73, a Golgi protein secreted into the serum, is a promising serum marker for diagnosing HCC [20].
Using 2D-SDS-PAGE, Yang et al. screened for Golgi proteins up-or downregulated in HCC [21]. They did not pick up the UNC50 protein spot in the matrix-assisted laser desorption/ionization-time-of-flight mass spectrometry. However, among the 17 differentially expressed proteins selected, COG8 (one of the eight subunits of the conserved oligomeric Golgi protein complex), has a yeast homolog, Cog8p, that interacts with Gmh1p (the yeast homolog for UNC50), as demonstrated in two independent large-scale screening experiments for protein interactions in yeast [22,23]. Table 2 contains the full list of Gmh1p interaction partners.
Despite these findings, current knowledge of Golgi apparatus proteins in HCC progression is scanty. Our research brings a new Golgi apparatus protein, UNC50, to more attention.
Using gene expression microarray analyses, we were able to screen for the oncological roles of UNC50 indirectly by interfering with UNC50 expression levels in Hep3B cells. The Hep3B cell line lacks the retinoblastoma (RB) and P53 genes, and has a hepatitis B virus (HBV) infection background. In our opinion, it is the appropriate cell model for Chinese HCC patients because most of these patients have a history of chronic HBV infection. The gene expression patterns revealed impaired expression levels of downstream targets of the EGFR pathway. We evaluated EGFR pathway activity by examining EGFR phosphorylation; western blotting showed that pEGFR levels were related to UNC50 expression, indicating that UNC50 plays a role in the EGFR pathway. In accordance with the molecular phenotype, we observed G0/G1 arrest and retardation of proliferation following UNC50 inhibition, and vice versa, upon EGF stimulation. We observed only marginal effects of UNC50 inhibition or overexpression on the cell cycle and no effect on proliferation because the ability of FBS to activate EGFR is very limited (Fig. 3A). Adding erlotinib, a specific EGFR inhibitor, abrogated the cell cycle differences, indicating that the impact of UNC50 on the cell cycle relies on a functional EGFR pathway.
Notably, total EGFR protein and EGFR mRNA levels did not change with UNC50 levels, thus UNC50 must act in a post-translational manner. EGFR staining with antibodies and flow cytometry analysis with or without permeabilization demonstrated that cell surface EGFR amounts are altered upon UNC50 overexpression or inhibition (Fig. 5), indicating that UNC50 influences EGFR translocation to the cell membrane despite total EGFR remaining unchanged. EGFR is rapidly internalized into endosomes and recycled to the cell surface even in the absence of ligands; endosomal retention is a key determinant for cell surface EGFR amounts [24]. UNC50 may play a role in facilitating EGFR recycling, thus increasing cell surface EGFR amounts. It is worth noting that EGFRs cycled between the cell surface and endosomes are less likely to be degraded [25]. That is why the total EGFR amount remained unchanged even when EGFR distribution was altered.  How could UNC50 influence the endosomal process if it is a Golgi apparatus membrane protein? We found that the recombinant UNC50 protein, with enhanced green fluorescent protein linked to its C-terminal, was located mainly in the Golgi apparatus and endoplasmic reticulum regions (Fig. 6). Moreover, an early study demonstrated that UNC50 is an inner nuclear membrane protein [26]. The Golgi apparatus is a hub for protein sorting and trafficking. A previous study showed that UNC50 affects receptor trafficking indirectly [14] and does not require colocalization with these receptors, and this might be the case for the relation between UNC50 and EGFR. The details regarding UNC50 location in HCC cells requires more evidence. There are many types of receptors, and for UNC50 to solely affect EGFR translocation would not be logical. Eimer et al., who studied levi-AChR, encountered the same puzzle regarding the specificity of UNC50 [14]. They clarified this by assuming that UNC50 plays a redundant role in protein trafficking and that levi-AChR exclusively relies on UNC50 functions. They also mentioned that apart from impaired movement, C. elegans UNC50 mutants Table 2. Gmh1p interaction partners in yeast.

Pubmed ID
Gene name exhibited only marginal phenotypes, which was greatly similar to our findings. Differing from their explanation for levi-AChR, we hypothesize that UNC50 is a global factor for protein trafficking. UNC50 plays certain roles for a specific protein (not confined to the receptor), and affects the counterpart of this protein, thus its effects are neutralized. Moreover, the effects of UNC50 on a small number of proteins are not totally neutralized; as the present study reveals, EGFR happens to be one of them. This may better explain the fact that UNC50 is conserved among all eukaryotic organisms, as both EGFR and AChR do not exist in all eukaryotic organisms. However, the implications of UNC50 in other protein trafficking processes require further extensive study. In summary, UNC50 enhances the EGFR pathway by facilitating EGFR translocation to the cell surface in the HCC cell line Hep3B, and promotes cell cycle entry and proliferation in the presence of EGF. This study relates the novel Golgi apparatus membrane protein UNC50 to an important tumor-promoting pathway that involves EGF. Further studies will determine its roles in HCC progression in detail.