Introduction

By either epidemiological studies or transgenic mice experiments, human hepatocarcinogenesis was found to be strongly associated with chronic hepatitis B and C (Beasley and Hwang, 1984; Chisari et al., 1989; Di Bisceglie, 1997; Moriya et al., 1998). On the other hand, genetic analysis of hepatocellular carcinoma (HCC) tissue revealed many chromosomal aberrations, such as deletion, duplication, and translocation (Marchio et al., 1997; Kussano et al., 1999). Aided by the methods of comparative genomic hybridization and genome-wide allelotyping, allelic imbalance can be identified on chromosomes 1p, 4q, 6p, 8p, 13q, 16q, and 17p (Boige et al., 1997; Nagai et al., 1997). Presumably, sequential accumulation of genetic alterations, possibly induced by viral infection or other carcinogens, may lead to HCC. Such a view is inspired by the occurrence of stepwise genetic mutations during the development of colon cancer (Vogelstein et al., 1988; Kinzler and Vogelstein, 1996). Identification of specific genetic mutations during the development of HCC is thus pivotal for us to understand the molecular mechanism of hepatocarcinogenesis.

One of the approaches to understand genetic alterations contributing to hepatocarcinogenesis was to identify differentially expressed genes between cancerous and noncancerous tissues. Several methods have been designed to achieve this goal, including subtractive hybridization, differential display polymerase chain reaction, and cDNA microarray technique (Wu et al., 1995; Scuric et al., 1998; Miyasaka et al., 2001; Okabe et al., 2001; Shirota et al., 2001; Wang et al., 2001). With the help of these tools, many differentially expressed genes have been identified. Presumably, some of the 'downregulated' genes were in fact located on the deleted chromosomal regions, which were lost during hepatocarcinogenesis. Alternatively, some of the downregulated genes were differentially expressed as a result of altered gene regulation. In such cases, mutations in their regulatory elements or functional change of the corresponding transcription factors could be investigated thereafter. Although these studies provided important information, data reported from different groups varied markedly. The profiles of differentially expressed genes identified by different groups were so different that no two of them were similar. Such a discrepancy could be attributed to different methods used or, alternatively, could be caused by great variations of genetic compositions and thus the expression profiles among different HCC tissues. For example, HCC derived from hepatitis B or C infection may possess distinct genetic expression profiles (Iizuka et al., 2002).

In this study, we attempted to identify genes that were most frequently downregulated among different HCC tissues. Firstly, a set of cDNA macroarrays was established allowing the identification of a panel of candidate downregulated genes. Then, two different methods were used to screen through these genes and the results were compared. Finally, we have characterized one of the identified downregulated genes, BMAL2, to understand whether it affects cell growth. BMAL2 belongs to the bHLH-PAS superfamily (Reddy et al., 1986; Crews et al., 1988; Hoffman et al., 1991; Taylor and Zhulin, 1999; Ikeda et al., 2000). RNA analysis revealed that expression of BMAL2 transcripts was restricted to the fetal brain and adult liver in humans (Ikeda et al., 2000). Although its physiological function in adult liver is not understood, some evidence suggests that it may serve as a regulator of circadian clock oscillation in brain cells (Okano et al., 2001). Members in this superfamily can be classified roughly into three groups (Hogenesch et al., 2000). The -class members often function as 'sensors' of environmental stimuli, the -class members act as partners for a broad range of -class molecules, and the -class members function as coactivators. Several environmental factors may activate/regulate pathways mediated by these molecules, such as the presence of polycyclic aromatic hydrocarbon, low oxygen stress, and circadian rhythmic change. On the other hand, several lines of evidence indicate that proteins in the bHLH-PAS superfamily are involved in regulation of cell growth and differentiation. For example, USF1/2 proteins are capable of inhibiting transformation of rat embryonic fibroblast mediated by Ras and c-Myc (Luo and Sawadogo, 1996). Myo D, another bHLH protein, can dominantly activate the myogenic differentiation program and is defective in several human tumor cell lines (Gerber and Tapscott, 1996). AhR induced p27(Ki1) cyclin/cdk inhibitor by altering Kip1 transcription, which caused suppression of 5L hepatoma cell growth (Kolluri et al., 1999). Another study demonstrated that AhR interacted with the retinoblastoma protein (pRb) and the interaction was crucial for maximal dioxin-induced G1 arrest in 5L hepatoma cells (Elferink et al., 2001). Furthermore, some bHLH proteins may negatively regulate these differentiation-related bHLH proteins and thus behave as oncoproteins (Norton, 2000). Based on these reports, we suspected BMAL2 might be able to regulate cell growth.

Results

Strategy to identify frequently downregulated genes in HCC

To identify frequently downregulated genes in HCC, we have designed a strategy combining different methods (Figure 1). Firstly, cDNA macroarrays containing 12 000 clones from a normal human liver cDNA library (purchased from Invitrogen) were established. According to the manufacturer, these macroarrays should contain about 4000 independent clones totally. Probes were then generated from cytoplasmic RNA of an HCC tissue and the clones where no significant hybridization occurred were considered as candidates of downregulated genes (Figure 2a). All selected clones were further verified using gel electrophoresis followed by Southern analysis (Figure 2b). Sequence analysis for these clones was performed. In all, 12 clones containing inserts of E. coli genes and 10 clones containing inserts of chromosomal DNA with repeated sequences were obtained using this method. Identification of these rare clones suggested that this macroarray method could indeed identify genes that were present in the cDNA library but were not expressed in the HCC tissue. Totally 195 independent clones were identified as candidates of downregulated genes.

Figure 1.

Strategy to identify frequently downregulated genes in HCCs

Full figure and legend (60K)

Figure 2.

Selection of downregulated genes using a cDNA macroarray method. (a) Colony hybridization of cDNA macroarrays using probes derived from an HCC tissue. Arrowheads, clones where no significant hybridization occurred. (b) Southern blot analysis (lower panels) following agarose gel electrophoresis (upper panels) for clones where no significant hybridization occurred. Solid triangles, the position of cloning vector; solid dots, confirmed clones where no significant hybridization occurred; P, positive control; M, molecular weight marker

Full figure and legend (159K)

Verification of downregulated genes in HCC

Two methods were then used to screen through these candidate genes. Firstly, semiquantitative RT–PCR was performed using five different pairs of cancerous and noncancerous tissues (T1/N1 to T5/N5). Genes of which differential expression was detected in at least three pairs of tissues are listed in Table 1. Northern blot was thus performed for these listed genes. Differential expression can be shown for 17 of them in at least one pair of tissue (Figure 3). Different patterns of Northern blot results were observed. Single or multiple mRNA species were detectable in noncancerous tissues, but were underexpressed in HCC (Figure 3a and b) or, alternatively, they were underexpressed as well as mis-spliced (alteration of the size in gel electrophoresis) in HCC (Figure 3c). Alternatively, these 195 clones were PCR amplified, gelpurified, and mechanically dotted onto a customized cDNA microarray biochip. Three pairs of cancerous and noncancerous tissues (T6/N6 to T8/N8) were used for comparison. It is noteworthy that all genes confirmed to be differentially expressed by the cDNA microarray method were also verified to be so by Northern analysis. Seven genes with inferred function were confirmed to be downregulated by both methods in our study: albumin, fibrinogen polypeptide, alcohol dehydrogenase 1B, pre-B-cell colony-enhancing factor, ELL-related RNA polymerase II elongation factor, tubulin subunit, and transcription factor BMAL2 genes. The latter two could only be verified in one pair of tissue using cDNA microarray technique. Four genes with unknown function were verified by quantitative RT–PCR and Northern analysis: GK001 protein, CLONE25003 hypothetical protein, cDNA FLJ32642 fis, and FLJ13409 hypothetical protein genes. The first two were also verified by cDNA microarray assay. Interestingly, the cDNA FLJ32642 fis gene was located in a frequently deleted region of HCC, 8p22. Finally, no gene was assigned to three of our clones.

Figure 3.

Northern analysis for downregulated genes in HCCs. (a) Single mRNA species that were detectable in noncancerous tissue (N) but were diminished in cancerous (T) tissue. (b) Multiple mRNA species that were detectable in noncancerous tissue but were diminished in cancerous tissue. (c) multiple or single mRNA species that were detectable in noncancerous tissue but were diminished and presumably mis-spliced in cancerous tissue

Full figure and legend (99K)

Table 1 - Downregulated genes in hepatocellular carcinoma obtained by a cDNA macroarray screening method.

Full table

Six additional comparisons were also made: T6, T7, T8, N6, N7, N8 were compared with a completely normal liver tissue (noncirrhotic) for differential gene expression (data not shown). Differential expression of these genes was not significant between noncancerous (N6-8) and the normal liver tissue, whereas the differential expression profiles between T6-8 and the normal liver tissue were similar to those of T6/N6 to T8/N8 (Table 1).

Identification of the 5' end sequences of BMAL2 mRNA in normal liver tissue

To further understand whether this method could identify genes capable of regulating cell proliferation, we have studied the effect of a transcription factor, BMAL2, on cell growth. Firstly, we tried to identify the 5' end of the mRNAs using the 5' RACE method. Two different sequences were identified using normal liver tissue, BMAL2-L1 and BMAL2-L2 (Figure 4a). Searching the GenBank data, the BMAL2-L1 sequence was previously identified, whereas the BMAL2-L2 sequence was novel. Two other species, BMAL2-1 and BMAL2-2, were also identified previously using other tissues. These sequence variations were possibly resulting from alternative splicing or differential promoter usage. Immunofluorescence study indicated that both BMAL2-L1 and L2 were nuclear proteins, although a significant proportion of BMAL2-L2 remained in the cytoplasm (Figure 4b).

Figure 4.

Subcellular localization of BMAL2 proteins derived from liver tissue. (a) Four distinct transcripts of BMAL2 genes. Sequences of BMAL2-1, BMAL2-2, and BMAL2-L1 were previously reported (GenBank accession number attached). BMAL2-L1 and BMAL2-L2 mRNA could be detected in normal liver tissue. (b) Subcellular localization of BMAL2-L1 and BMAL2-L2 by immunofluorescence analysis

Full figure and legend (108K)

Overexpression of antisense RNA of BMAL2 in 293EBNA cells

To investigate whether BMAL2 proteins affect cell growth, DNA fragments encoding BMAL2 were expressed in 293EBNA cells either in antisense (pDR2-BMAL2-rev) or sense (pDR2-BMAL2-L2) orientation (Figure 5a). Northern blot analysis clearly demonstrated the presence of antisense RNA, but not the mRNA in sense orientation (Figure 5b). RT–PCR showed the presence of sense BMAL2 mRNA in all the three cell lines (mock, pDR2-BMAL2-rev, and pDR2-BMAL2-L2 transfected cells). Interestingly, the detected bands were shorter than expected. Sequence analysis after gel purification of this product showed another previously unreported splicing site (nt 1491–1678). The DNA fragment with the expected length can still be detected if a large amount of PCR product was loaded on gel. However, the amount is extremely small in comparison to the spliced species. All three cell lines were then subjected to flow cytometric analysis. No significant difference was found between the mock and pDR2-BMAL2-L2 transfected cells in terms of cell distribution among various cell cycle phases, whereas a significantly reduced proportion of cells in the G2 phase with a concomitantly increased proportion of cells in the S phase was observed in pDR2-BMAL2-rev transfected cells (Figure 5c).

Figure 5.

Overexpression of antisense BMAL2 RNA. (a) Plasmids and probes used for expression and detection of BMAL2 RNAs in sense and antisense orientation. DNA fragments of BMAL2-L2 and BMAL2-rev were inserted into pDR2 in sense and antisense orientation, respectively. Probe-1 and Probe-2 were generated to detect BMAL2 RNA (by Northern blot) in sense and antisense orientation, respectively. P1 and P2 were used to detect BMAL2 mRNA (by RT–PCR) in sense orientation. (b) Northern analysis of BMAL2 RNA in 293EBNA cells. Single-stranded DNA fragment (10 pg), either in sense (upper panel) or antisense orientation (lower panel), was generated and used as a positive control (lane 1). Total RNAs extracted from 293EBNA cells transfected by pDR2 (lane 2), pDR2-BMAL2-L2 (lane 3), and pDR2-BMAL2-rev (lane 4) were subjected to Northern analysis. BMAL2 mRNA was also detected using RT–PCR (right panel; M, molecular weight marker; lanes 1–3, pDR2, pDR-BMAL2-L2, and pDR2-BMAL2-rev transfected cells). Empty triangle, antisense BMAL2 RNA; solid triangle, RT–PCR product of BMAL2 mRNA. (c) Flow cytometry analysis of 293EBNA cells transfected by pDR2 (shaded area) or pDR2-BMAL2-rev (solid area)

Full figure and legend (81K)

Antisense overexpression of BMAL2 enhances cell proliferation

Since antisense overexpression of BMAL2 RNA altered the cell distribution in the S and G2 phases of the cell cycle, we further investigated whether it affected cell proliferation (Figure 6a–c). Three assays were performed in this study. By performing [³ H]thymidine incorporation assay in cells synchronized by serum starvation, we discovered that the cell cycle time of the first synchronized cycle after replenishment of serum was shortened (Figure 6a). By performing TNF- stimulation assay, it was found that TNF- could induce an increment of CPP32/caspase-3 activity in mock-transfected 293EBNA cells. The increment was diminished in pDR2-BMAL2-rev transfected cells (Figure 6b). Furthermore, after removal of hygromycin in the culture medium, allowing the loss of extrachromosomal replicating plasmids (i.e. pDR2-BMAL2-rev), the cells became susceptible to the TNF- induced increment of CPP32/caspase-3 activity again. Finally, the plating efficiency in soft agar was determined. It was found that the plating efficiency increased significantly from 82 to 94% for the antisense overexpressed cells (Figure 6c). These pieces of evidence suggested that antisense overexpression of BMAL2 RNA was advantageous for cell proliferation.

Figure 6.

Enhancement of cell proliferation by antisense overexpression of BMAL2. (a) Shortened cell cycle time in 293EBNA cells transfected by pDR-BMAL2-rev. Cells were synchronized by serum deprivation. After replenishment of serum, cells were labeled with [³H]thymidine every 3 h. Solid circle, cells transfected by pDR2-BMAL2-rev; empty square, cells transfected by pDR2. (b) Diminished TNF- induced increment of CPP32/caspase-3 activity in 293EBNA cells transfected by pDR2-BMAL2-rev. Cells were either transfected by pDR2 (mock) or pDR2-BMAL2-rev (Rev, Hyg+). The latter transformants were then cultured in medium without hygromycin for 1 month (Rev, Hyg-), which allowed elimination of the episomal plasmids, pDR2-BMAL2-rev. Cellular caspase-1 and -3 activities were then measured after TNF- treatment. Activities were calculated as folds of increase comparing to cells without TNF- treatment. (c) Increased plating efficiency in 293EBNA cells transfected with pDR2-BMAL2-rev. Cells transfected with pDR2 and pDR2-BMAL2-rev were plated out in 0.35% soft agar. The percentages of cells that grew into colonies were calculated. Vertical bars in (b) and (c), standard deviations of three experiments. P-values were calculated by the one-tailed Student's t-test

Full figure and legend (84K)

Discussion

It is generally believed that deciphering the changes of gene expression profiles in HCCs might bring to light the molecular mechanism of hepatocarcinogenesis. However, studies from different groups revealed various sets of gene expression profiles that could not be matched with each other. Presumably, the presence of multiple oncogenic pathways and multiple steps of hepatocarcinogenesis were responsible. The different methods used in these studies also contributed to the great variations of results. In this study, we tried to identify frequently underexpressed genes in HCCs by combining different methods to test multiple pairs of tissues. Theoretically, underexpressed genes identified in this way should be associated with and the oncogenic mechanism shared by different pathways. Our results, however, were again difficult to be matched with those from other studies. Two of the downregulated genes, albumin and fibrinogen polypeptide, were reported in at least one other study, while the other downregulated genes appeared only in our list (Xu et al., 2001). Many of our identified genes were located in frequently deleted chromosomal regions in HCCs, such as 4q, 6p, 6q, 8p, and 18p. Conceivably, 'downregulation' of these genes was in fact caused by deletion of the genes. The function of several genes in this list has not yet been reported. It would be interesting to study whether these genes affect cell proliferation.

In this study, we discovered that BMAL2 mRNA was expressed in a very small amount in 293EBNA cells. Furthermore, the amount did not increase even after transfection of a plasmid expressing BMAL2 mRNA in sense orientation. In contrast, the antisense BMAL2 RNA fragment can be expressed to a high level (detectable by Northern blot) using the same vector. Conceivably, BMAL2 mRNA was maintained at a lower level presumably through a mechanism of post-transcriptional regulation. The majority of BMAL2 mRNA in 293EBNA cells was constituted of a transcript harboring a novel splicing site. The resulting transcript encoded a truncated BMAL2 protein lacking 107 amino-acid residues of the carboxyl terminal portion. However, both the bHLH and PAS domains are intact. The role of this spliced product was unclear at this stage. Since we could not overexpress the sense BMAL2 mRNA, we thus studied the effect of BMAL2 on cell growth through overexpression of antisense RNA. All our assays indicated that overexpression of antisense BMAL2 RNA promoted cell proliferation. By performing TNF- stimulation assay, we discovered that the increment of caspase-3 but not caspase-1 activity was diminished in antisense BMAL2 RNA overexpressed cells. Caspase-1 functions mainly in cytokine processing (Kuida et al., 1995), whereas caspase-3 is involved in the regulation and execution of apoptosis (Kuida et al., 1996). The present data suggested that antisense BMAL2 RNA likely exerted an antiapoptosis effect and promoted cell growth. It is thus highly likely that BMAL2 itself caused suppression of cell growth as did many bHLH-PAS members, and the mRNAs of BMAL2 were kept to a lower level so that the cells could proliferate better.

In summary, we have utilized a strategy combining different methods to harvest a list of frequently downregulated genes in HCCs. Antisense overexpression of one of them, that is, BMAL2, in 293EBNA cells enhanced cell proliferation.

Materials and methods

Liver tissues

One HCC tissue obtained from a patient positive for hepatitis B virus surface antigen (HBsAg) but negative for hepatitis C virus antibody (anti-HCV) was used for the macroarray screening assay. Five pairs of cancerous and adjacent noncancerous liver tissues obtained from five patients (three were positive for HBsAg and two were positive for anti-HCV; T1/N1 to T5/N5) were included for semiquantitative RT–PCR and Northern analysis. Three more pairs of liver tissues obtained from three patients (two were positive for HBsAg and one was positive for anti-HCV; T6/N6 to T8/N8) were included for cDNA microarray analysis. A piece of normal liver tissue, obtained from a patient with large hemangioma, was also included for comparison. This patient is negative for either HBsAg or anti-HCV. All tissues were surgically removed, subjected to histopathological examination, and immediately frozen in -70°C until assayed.

Cell lines

Human embryonic kidney cells (named 293EBNA cells) constitutively expressing Epstein–Barr virus nuclear antigen-1 (EBNA-1) protein from Epstein–Barr virus (293EBNA cells; Invitrogen, Carlsbad, CA, USA) were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and 250 g of G418 per ml. Huh-7 cells were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. Transfection was performed by the standard CaPO₄ precipitation method.

cDNA macroarray assay

A cDNA library derived from normal human liver (Clontech Laboratories, Inc., Palo Alto, CA, USA) was constructed by inserting the cDNAs into a vector, pDR2. E. coli carrying this library was spread on Luria-Bertani agar plates, and each individual colony was picked and arranged to form arrays. The arrays of colonies were lifted onto nitrocellulose filters and lysed in 10% SDS to liberate DNA, which was then fixed to the filter by baking. Totally 12 000 clones were blotted in this study. A single-stranded probe was then generated from cytoplasmic RNA extracted from an HCC tissue. The tissue was minced into small pieces, and lysed in a buffer containing 10 mM Tris hydrochloride (pH 7.2), 150 mM NaCl, and 0.5% Nonidet P-40. After centrifugation at 1500 g for 5 min, the supernatant was used for RNA extraction. Reverse transcription (RT) was performed using SuperScript II RNase H minus Reverse Transcriptase (Invitrogen Corporation, Calsbad, CA, USA) and oligo(dT) was used as the RT primer. One-third of dTTP in the dNTP mixture was replaced by digoxigenin-11-dUTP (Boehringer Mannheim, Germany) to generate digoxigenin labeled probes. The probes were mixed (1 : 2 in molar ratio) with oligo(dA) in 40°C for 1 h before hybridization. The hybridization signal was detected by a DIG Luminescent Detection Kit (Boehringer Mannheim, Germany). For each batch of hybridization, 1 ng of pDR2 without cDNA insert was used as a negative control and 1 pg of pDR2 containing a fragment of human albumin gene (Hs.184411) was used as a positive control. To confirm the result derived from the aforementioned macroarray hybridization method, plasmids from colonies where no significant hybridization occurred on the filter were extracted and analysed by electrophoresis on an agarose gel followed by Southern blotting.

cDNA microarray assay

Primers flanking the inserted cDNAs in the pDR2 vector were used for PCR to generate a high concentration of purified cDNA fragments. These cDNA fragments were mechanically placed onto a Biochip (ABC Human UniversoChip 8k-1, Asia Bioinnovation Corporation, CA, USA). Total RNA extracted from cancerous and noncancerous liver tissues were used for reverse transcription in the presence of fluorescence dyes, Cy3 and Cy5. After hybridization, the image was obtained using a chip scanner (GenePix 400B Array Scanner, Axon Instruments, USA) and the data were analysed using a software package, GenePix Pro 3.0.5.56. The data were normalized by the method of standard median centering. Differential expression was considered as significant when the ratio of signals was greater than 2.

Semiquantitative RT–PCR and Northern analysis

To verify whether a particular RNA species was underexpressed in HCC tissue, semiquantitative RT–PCR was performed. For each clone, cDNA was first sequenced up to at least 500 bp using an automatic DNA sequencer (CEQ 2000; Beckman Instruments, Inc., Fullerton, CA, USA). Primers flanking 200–500 bp of each cDNA fragment were synthesized accordingly. Another set of primers flanking a fragment of -actin mRNA, 5'-CACCAACTGGGACGACATGG-3' (nt 301–320, sense) and 5'-AGGATCTTCATGAGGTAGTC-3' (nt 651–532, antisense), was also synthesized and used as an internal control. Equal amounts of total RNA extracted from cancerous and noncancerous liver tissues were used for RT–PCR. RT was performed using random primers. PCR was performed in a mixture containing the cDNA flanking primers and the -actin primers at 94°C for 1 min, 55°C for 1 min, and 72°C for 1 min for each cycle. PCR product (10 l) obtained from the 10th, 15th, 20th, 25th, and 30th cycles of PCR mixture was subjected to agarose gel electrophoresis. Differential expression was considered as significant when the target cDNA was detected in noncancerous but not the cancerous tissue in at least one of these products, while -actin cDNA was detected in both the tissues in equal densities. The probes for Northern analysis were generated by PCR using the same sets of cDNA flanking primers. A probe used to detect -actin mRNA was also generated using the -actin primers. Differential expression was considered as significant when the ratio of target mRNA signals was greater than 2 in the noncancerous and cancerous tissue, while the -actin mRNA signals were the same.

5' Rapid amplification of cDNA ends (RACE) assay

To study the 5' end sequences of BMAL2 mRNA, the 5' RACE assay was performed using a 5'/3' RACE kit (Boehringer Mannheim Biochemica, Mannheim, Germany). Total RNA was extracted from normal liver. The primer, R1: 5'-CCTCTTTCACATCCAACCAC-3'(nt 490–509, antisense), was used for cDNA synthesis. After being tailed with dATP homopolymer by a terminal deoxynucleotidyl transferase, the tailed cDNA was amplified by PCR with an oligo(dT)-anchor primer and a primer, R2:5'-AGCCTTCTGCAGTCTTAAGG-3' (nt 462–481, antisense), located slightly upstream of R1. Finally, a second step of PCR was performed with the anchor primer and R3:5'-GTCTGAGCTCATTATCCTGAAG-3' (nt 433–454, antisense), located further upstream of R2. The details of the experimental procedure and the sequences of oligo(dT)-anchor primer have been described previously (Yeh et al., 1997).

Immunofluorescence staining analysis

The coding region of BMAL2 cDNAs was inserted in-frame with the sequence of a V5 epitope in pcDNA3.1/V5-His (Invitrogen, San Diego, CA, USA). The plasmid was transfected into Huh-7 cells. The methods of cell fixation and staining have been described previously (Yeh et al., 2001). The primary antibody used was mouse anti-V5 monoclonal antibody (Invitrogen). The secondary antibody used was fluorescein isothiocyanate-conjugated goat anti-mouse antibody (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA).

Overexpression of antisense RNA of BMAL2

Expression of Epstein–Barr nuclear antigen-1 in 293 cells allows extrachromosomal replication of plasmids carrying the Epstein–Barr virus replication origin (OriP) (Hambor et al., 1998; Hung et al., 2001). An antisense fragment constructed using this system expressed a high level of antisense RNA in EBNA-1 expressing cells. A fragment of BMAL2 gene (nt 420–1280) was inserted in antisense orientation into pDR2, downstream of a Rous sarcoma virus long terminal repeat promoter. The resulting plasmid, pDR2-BMAL2-rev, was transfected into 293EBNA cells and the transformants were selected under 150 g of hygromycin B per ml. Two other plasmids, pDR2 and pDR2-BMAL2-L2, containing no insert and an insert of BMAL2-L2 coding sequence, respectively, were also transfected into 293EBNA cells to establish transformants. To establish single-stranded probes for specific detection of sense or antisense mRNA, primers flanking the insert in pDR2-BMAL2-rev were used for single-sided PCR. The details of this procedure have been described previously (Yeh et al., 1998). To detect BMAL2 mRNA by RT–PCR, two primers were synthesized, P1: 5'-GTATGAGTGTACCTGGAATG-3' (nt 1400–1419, sense) and P2: 5'-GGCCCCCCTCTGCTTCTAAG-3' (nt 1752–1771, antisense). To detect the mRNA in sense orientation, P2 was used for RT.

Flow cytometric analysis

Cells were detached from the plate by trypsinization, washed in ice-cold PBS, and fixed by 100% ethanol. After staining with propidium iodide, cells were analysed with a Coulter Epics Altra flow cytometer. To calculate the percentages of cells in different cell phases, the readings of DNA content for the mid-G1 and G2 phases were assigned as one and two copies/cell. Cells with DNA content between 0.75–1.25, 1.25–1.75, and 1.75–2.25 copies/cell were calculated as the G1,S, and G2 phase cells.

[³H]thymidine incorporation assay

The procedure for this assay has been described previously (Yeh et al., 1993). Briefly, the cells were labeled with [³H]thymidine at a concentration of 25 Ci/ml for 30 min at selected time points after synchronization. The cells were rinsed with PBS twice and lysed with lysis buffer containing 13.3 mM Tris-HCl (pH 8.0), 6.7 mM EDTA, 0.7% SDS, and 200 g/ml of proteinase K. After heating at 55°C for 4 h, chromosomal DNA was extracted and counted using a scintillation counter.

Measurement of TNF--induced increment of caspase-1 and -3 activities

The procedure has been described previously (Yeh et al., 2000). Briefly, 5 10⁴ cells were incubated in minimal essential medium containing 10% fetal bovine serum and 1 g/ml of actinomycin D (Boehringer Mannheim). Recombinant human TNF- (40 ng/ml, 2.86 10⁷ U/mg, R&D systems) was added into the medium and the cells were incubated for 7 h at 37°C before being harvested for assay. The ICE/caspase-1 and CPP32/caspase-3 activities were assayed using the CaspACE Assay System (Promega, Madison, WI, USA) according to the manufacturer's instructions.

Plating efficiency in soft agar

Plating efficiency was calculated as the number of colonies formed divided by the number of cells seeded in 0.35% soft agar.

References

1. Beasley RP and Hwang LY. (1984). Semin. Liver Dis., 4, 113–121. | PubMed | ChemPort |
2. Boige V, Laurent-Puig P, Fouchet P, Flejou JF, Monges G, Bedossa P, Bioulac-Sage P, Capron F, Schmitz A, Olschwang S and Thomas G. (1997). Cancer Res., 57, 1986–1990. | PubMed | ISI | ChemPort |
3. Chisari FV, Klopchin K, Moriyama T, Pasquinelli C, Dunsford HA, Sell S, Pinkert CA, Brinster RL and Palmiter RD. (1989). Cell, 59, 1145–1156. | Article | PubMed | ISI | ChemPort |
4. Crews ST, Thomas JB and Goodman CS. (1988). Cell, 52, 143–151. | Article | PubMed | ISI | ChemPort |
5. Di Bisceglie AM. (1997). Hepatology, 26 (3 Suppl 1), 34S–38S. | Article | PubMed | ChemPort |
6. Elferink CJ, Ge NL and Levine A. (2001). Mol. Pharmacol., 59, 664–673. | PubMed |
7. Gerber AN and Tapscott SJ. (1996). Mol. Cell. Biol., 16, 3901–3908.
8. Hambor JE, Hauer CA, Shu HK, Groger RK, Kaplan DR and Tykocinski ML. (1998). Proc. Natl. Acad. Sci. USA, 85, 4010–4014.
9. Hoffman EC, Reyes H, Chu FF, Sander F, Conley LH, Brooks BA and Hankinson O. (1991). Science, 252, 954–958. | Article | PubMed | ISI | ChemPort |
10. Hogenesch JB, Gu YZ, Moran SM, Shimomura K, Radcliffe LA, Takahashi JS and Bradfield CA. (2000). J. Neurosci., 20, RC83 (1–5). | PubMed | ChemPort |
11. Hung SC, Kang MS and Kieff E. (2001). Proc. Natl. Acad. Sci. USA, 98, 1865–1870. | Article | PubMed | ChemPort |
12. Iizuka N, Oka M, Yamada-Okabe H, Mori N, Tamesa T, Okada T, Takemoto N, Tangoku A, Hamada K, Nakayama H, Miyamoto T, Uchimura S and Hamamoto Y. (2002). Cancer Res., 62, 3939–3944. | PubMed | ISI | ChemPort |
13. Ikeda M, Yu W, Hirai M, Ebisawa T, Honma S, Yoshimura K, Honma KI and Nomura M. (2000). Biochem. Biophys. Res. Commun., 275, 493–502.
14. Kinzler KW and Vogelstein B. (1996). Cell, 87, 159–170. | Article | PubMed | ISI | ChemPort |
15. Kolluri SK, Weiss C, Koff A and Gottlicher M. (1999). Genes Dev., 13, 1742–1753. | PubMed | ISI | ChemPort |
16. Kuida K, Lippke JA, Ku G, Harding MW, Livingston DJ, Su MS and Flavell RA. (1995). Science, 267, 2000–2003. | Article | PubMed | ISI | ChemPort |
17. Kuida K, Zheng TS, Na S, Kuan C, Yang D, Karasuyama H, Rakic P and Flavell RA. (1996). Nature, 384, 368–372. | Article | PubMed | ISI | ChemPort |
18. Kussano N, Shiraishi K, Kubo K, Oga A, Okita K and Sasaki K. (1999). Hepatology, 29, 1858–1862. | Article | PubMed | ISI | ChemPort |
19. Luo X and Sawadogo M. (1996). Proc. Natl. Acad. Sci. USA, 93, 1308–1313. | Article | PubMed |
20. Marchio A, Meddeb M, Pineau P, Dangiot G, Tiollais P, Bernheim A and Dejean A. (1997). Genes Chromosomes Cancer, 18, 59–65. | Article | PubMed | ISI | ChemPort |
21. Miyasaka Y, Enomoto N, Nagayama K, Izumi N, Marumo F, Watanabe M and Sato C. (2001). Br. J. Cancer, 85, 228–234. | Article | PubMed | ISI | ChemPort |
22. Moriya K, Fujie H, Shintani Y, Yotsuyanagi H, Tsutsumi T, Ishibashi K, Matsuura Y, Kimura S, Miyamura T and Koike K. (1998). Nat. Med., 4, 1065–1067. | Article | PubMed | ISI | ChemPort |
23. Nagai H, Pineau P, Tiollais P, Buendia MA and Dejean A. (1997). Oncogene, 14, 2927–2933. | Article | PubMed | ISI | ChemPort |
24. Norton JD. (2000). J. Cell Sci., 113, 3897–3905. | PubMed | ISI | ChemPort |
25. Okabe H, Satoh S, Kato T, Kitahara O, Yanagawa R, Yamaoka Y, Tsunoda T, Furukawa Y and Nakamura Y. (2001). Cancer Res., 61, 2129–2137. | PubMed | ISI | ChemPort |
26. Okano T, Yamamoto K, Okano K, Hirota T, Kasahara T, Sasaki M, Takanaka Y and Fukada Y. (2001). Genes Cells, 6, 825–836.
27. Reddy P, Jacquier AC, Abovich N, Peterson G and Rosbash M. (1986). Cell, 46, 53–61. | Article | PubMed | ChemPort |
28. Scuric Z, Stain SC, Anderson WF and Hwang JJ. (1998). Hepatology, 27, 943–950. | PubMed |
29. Shirota Y, Kaneko S, Honda M, Kawai HF and Kobayashi K. (2001). Hepatology, 33, 832–840. | Article | PubMed | ISI | ChemPort |
30. Taylor BL and Zhulin IB. (1999). Microbiol. Mol. Biol. Rev., 63, 479–506. | PubMed | ISI | ChemPort |
31. Vogelstein B, Fearon ER, Hamilton SR, Kern SE, Preisinger AC, Leppert M, Nakamura Y, White R, Smits AM and Bos JL. (1988). N. Engl. J. Med., 319, 525–532. | PubMed | ISI | ChemPort |
32. Wang ZX, Wang HY and Wu MC. (2001). Br. J. Cancer, 85, 1162–1167. | Article | PubMed | ISI | ChemPort |
33. Wu GS, Kar S and Carr BI. (1995). Life Sci., 57, 1077–1085.
34. Xu XR, Huang J, Xu ZG, Qian BZ, Zhu ZD, Yan Q, Cai T, Zhang X, Xiao HS, Qu J, Liu F, Huang QH, Cheng ZH, Li NG, Du JJ, Hu W, Shen KT, Lu G, Fu G, Zhong M, Xu SH, Gu WY, Huang W, Zhao XT, Hu GX, Gu JR, Chen Z and Han ZG. (2001). Proc. Natl. Acad. Sci. USA, 98, 15089–15094. | Article | PubMed | ChemPort |
35. Yeh CT, Chiu HT, Chu CM and Liaw YF. (1998). J. Med. Virol., 55, 42–50. | PubMed |
36. Yeh CT, Lai HY, Chen TC, Chu CM and Liaw YF. (2001). J. Virol., 75, 11017–11024.
37. Yeh CT, Lu SC, Chu CM and Liaw YF. (1997). J. Gen. Virol., 78, 2761–2770.
38. Yeh CT, Shen CH, Tai DI, Chu CM and Liaw YF. (2000). Oncogene, 19, 5213–5220. | Article | PubMed | ChemPort |
39. Yeh CT, Wong SW, Fung YK and Ou JH. (1993). Proc. Natl. Acad. Sci. USA, 90, 6459–6463.

Acknowledgements

This work was supported by a grant from the National Science Council, Taiwan (NSC 91-3112-B-182-002).