To clone genes that are differentially expressed in hepatoblastomas, a cDNA library was constructed according to the manufacturer's recommendations. The SSH procedure (21) was performed with the PCR-Select cDNA Subtraction Kit (Clontech, Palo Alto, CA) using cancer and normal tissues as tester and driver, respectively. The tissue samples were washed with phosphate-buffered saline (PBS) and sliced into smaller pieces. Total RNA was extracted with Tri reagent (MRC, Cincinnati, OH). Poly(A)⁺ mRNA was isolated with the poly(A) Quick mRNA Isolation Kit (Stratagene, La Jolla, CA) and used as the template for cDNA synthesis. To clone the cDNA inserts unidirectionally, a cDNA synthesis primer with the HindIII recognition sequence, supplied in the kit, was replaced with a primer with a XhoI recognition sequence (5'-TTTTGTACCTCGAGTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT-3'). First-strand cDNA was synthesized with the modified primer, followed by second-strand cDNA synthesis using T4 DNA polymerase. The tester and driver cDNAs, prepared according to the protocol, were digested with RsaI, a four-base cutting restriction enzyme that yields blunt ends. The tester cDNAs were then subdivided into two groups, and separately ligated to group-specific cDNA adaptors (Adaptor 1: 5'-ACCTGCCCGG-3', complementary to a portion of 5'-CTAATACGACTCACTATAGGGCTCGAGCGGCCGCCCGGGCAGGT-3'; Adaptor 2: 5'-ACCTCGGCCG-3', complementary to a portion of 5'-CTAATACGACTCACTATAGGGCAGCGTGGTCGCGGCCGAGGT-3'). Two hybridization reactions were performed, followed by suppression PCR to amplify the differentially expressed sequences using a pair of primers (forward 5'-TCGAGCGGCCGCCCGGGCAGGT-3' and reverse 5'-AGCGTGGTCGCGGCCG AGGT-3') and the Advantage cDNA polymerase mix (Clontech, Palo Alto, CA). The subtracted cDNA pool was cloned into the NotI and XhoI restriction sites of the multiple cloning site of the pBluescript (pBS) SK (-) vector (Stratagene, La Jolla, CA) in the same direction as the LacZ gene using the standard procedure (22). The cDNA plasmid library was transformed into Epicurian Coli XL-10 Gold Ultracompetent Cells (Stratagene, La Jolla, CA) by the calcium-chloride method (22).

To determine the total number of primary transformants, 1 or 10 µl of 1 ml pilot transformation was plated onto LB ampicillin agar plates. After overnight incubation at 37℃, the number of ampicillin-resistant colonies was counted and multiplied by a factor of 1000 or 100. To determine the percentage of vectors with inserts and the average insert size, recombinant plasmids of 40 randomly selected individual clones were purified by the alkaline lysis method (22) and digested by the same restriction enzymes used in cloning. Further, to determine if unidirectional cloning had been successfully, cDNA regions of purified plasmids derived from 100 randomly selected individual transformants were sequenced from the 5' end of the (+) strand by employing the T3 primer in an automated DNA sequencer.

Bacterial competent cells containing recombinant pBS SK (-) phagemids were plated on LB agar plates containing 50 µg/ml ampicillin and 50 µg/ml tetracycline and incubated at 37℃ for 16 h. Well-isolated colonies were seeded in a well of 96 deep-well plates containing 1.4 ml of 2× YT liquid media (tryptone 16 g, yeast extract 10 g and NaCl 10 g per 1000 ml)with and additional 50 µg/ml ampicillin. Cells were cultured for 7 h at 37℃ with vigorous shaking. To produce LC-antisense molecules from each phagemid, 20 µl of the bacterial culture was transferred to each well that had been prefilled with 1.4 ml of fresh 2× YT liquid media containing 9 µl of helper bacteriophages, M13K07 (New England Biolabs, Bevelry, MA). After 1 h incubation, 4.2 µl kanamycin (70 µg/ml) was added and cultured at 37℃ for 12 h. The superinfection was carried out in triplicate for each clone to maximize the yield of antisense molecules in a single purification step. Single-stranded LC-antisense molecules were purified from the culture supernatant of bacterial competent cells using QIAprep 96 M13 Kits and QIAVAC vacuum manifolds (Qiagen, Hilden, Germany) according to the manufacturer's instructions. To test both the quantity and purity, the prepared LC-antisense molecules were run on a 1% agarose gel along with control LC-molecules derived from pBS SK (-) phagemid without a cDNA insert.

To identify genes involved in the growth of liver cancer cells, antisense molecules of the liver cancer specific LC-antisense library were mixed with a cationic liposome and transfected into the liver cancer cell line, HepG2. Briefly, the cells (7×10³), washed twice with Opti-MEM (Invitrogen, Carlsbad, CA), were seeded in each well of 96-well plates in 100 µl of Opti-MEM supplemented with 10% FBS (JBI, Daegu, Korea). HepG2 cells were incubated for a further 12~18 h at 37℃ in an incubator with 5% CO₂. LC-antisense molecules (0.1 µg) were complexed with Lipofectamine 2000 (Invitrogen, Carlsbad, CA) in a ratio of 1:3 (w/w), and the antisense molecule+liposome complexes were added to the cultured cells according to the manufacturer's instructions. The cultures were exchanged with fresh media 24 h after transfection and incubated for a further 4 days. To compare the effects of the LC-antisense molecules on cell proliferation, identical quantities of liposome alone and control DNA+ liposome complexes were also added to the cells in a different 96-well plate and simultaneously assayed. Control DNA was single-stranded phage genomic DNA derived from pBS SK (-) phagemid without a cDNA insert.

Light microscopic observation and an MTT reduction assay were performed to study the effect of antisense molecules on the proliferation of HepG2 cells. For an MTT reduction assay, cell culture media was replaced with 50 µl fresh media and 25 µl 5 mg/ml MTT reagent (3-(4,5-dimethylthazol-2-yl)-2,5-diphenyltetrazolium bromide, in PBS) (Sigma-Aldrich, St. Louis, MO). The MTT reaction was carried out in a 37℃ incubator for 4 h followed by addition of 150 µl isopropanol, containing 0.1 N HCl, and further incubation at room temperature for 1 h. The absorbance was measured at 570 nm with a SpectraMAX 190 (Molecular Devices, Sunnyvale, CA) to quantify the surviving cells. The percentage growth inhibition of the cells in each well treated with liposome alone, antisense+liposome complexes or control DNA+liposome complexes was calculated by comparing the optical density to that of a sham treatment, using the following formula: 1-(absorbance of an experimental well/absorbance of a sham control well)×100.

To identify genes complementary to the LC-antisense molecules that inhibited the proliferation of liver cancer cells, recombinant phagemids obtained by alkaline lysis were sequenced from the 5' upstream of the (+) strand of cDNA inserts using the T3 primer. The revealed cDNA sequences were compared with those of the GenBank database (http://www.ncbi.nlm.nih.gov/genebank). Polypeptides deduced from the cDNA sequences were then searched for amino acid motifs using the ProfileScan Server (http://hits.isb-sib.ch/cgi-bin/PFSCAN?).

HepG2 cells (5×10⁴) were seeded in each well of 24-well plates. Cells were trypsinized and suspended in 250 µl PBS buffer in 48 h treatments of 0.8 µg of selected LC-antisense molecules combined with 2.4 µg of Lipofectamine 2000. As controls, sham treated cells, with liposome alone or control DNA+liposome complexes, were also prepared. Cell suspensions were fixed with 750 µl 95% ethanol by vortexing and incubating for at least 1 h on ice. Cells were centrifuged, washed twice with 1 ml PBS, and treated with 5 µl RNaseA (1 mg/ml) at room temperature for 5 min. Cells were stained with propidium iodide (PI), at a final concentration of 10 µg/ml, followed by incubation at room temperature for 30 min. The cell cycle patterns were analyzed on an equal number (10⁴) of cells by flow cytometry using a FACSCalibur System (Becton Dickinson, Franklin Lakes, NJ).

HepG2 cells (1×10⁶) were seeded in a 6 cm-dish one day prior to antisense transfection. Eight µg of each LC-antisense species was complexed with 24 µg of Lipofectamine 2000 and added to the cells. To compare the effects of the LC-antisense molecules on apoptosis, equal amounts of the control DNA+ liposome complexes were added to the cells for control analysis. In addition, cisplatin (Sigma-Aldrich, St. Louis, MO), an apoptosis-inducing positive-control drug, was also added, to a final concentration of 16 µg/ml. Cells were lysed with a cell lysis buffer (1% NP-40, 25 mM EDTA and 50 mM Tris-HCl, pH 6.6) 48 h post-transfection. Cell lysates were treated with RNaseA (final conc. 20 µg/ml) and incubated for 3 h at 58℃. Proteinase K was then added to a final concentration of 20 µg/ml and incubated for 3 h at 37℃. The cell lysates were mixed with 2.5 vol. of ice-cold ethanol and 0.1 vol. of 3 M ammonium acetate and stored overnight at -20℃. Precipitated DNA was resuspended in 25 µl TE buffer (pH 8.0) and run on a 1.6% agarose gel.

A library of LC-antisense molecules was constructed with an M13 phagemid vector, utilizing the salient feature of easy cDNA library construction (Fig. 1). Messenger RNA was prepared from normal liver and hepatoblastoma tissues, and differentially amplified for liver cancer specific mRNA by a subtractive hybridization method. Amplified cDNAs were unidirectionally cloned into the pBS SK (-) vector using the NotI and XhoI restriction sites, and the cDNA library was transformed into E. coli competent cells. The primary library size was determined as 2×10⁴ cfu/ml. By performing restriction analysis, it was determined that 97.5% of the tested transformants had cDNA inserts, and the average insert size was approximately 300 bp (data not shown). When sequenced for cloning integrity using 100 randomly selected clones, 96% of the recombinant phagemids had inserts in the intended orientation (data not shown). From the cDNA library of 9,600 transformants, 1,200 clones with cDNA inserts of more than 500 base pairs were selected with a simplified procedure for plasmid preparation (23).

LC-antisense molecules were purified from the culture supernatant of bacterial competent cells superinfected with helper bacteriophages. After confirming adequate purification of the antisense molecules by gel electrophoresis, the random gene LC-antisense library of 1,200 member species specific for liver cancer was arrayed in thirteen 96-well plates for further use in functional analysis.

LC-antisense molecules from the antisense library were transfected into HepG2 cells in order to screen genes affecting the growth of the liver cancer cell line. To examine the optimal transfection conditions in HepG2 cells, various cationic liposomes were complexed with recombinant plasmid, containing the GFP or luciferase gene in various ratios, and treated to the cells (data not shown). Based on these preliminary results, each LC-antisense molecule (0.1 µg) was complexed with a cationic liposome at a ratio of 1:3 (w/w) and transfected into 7×10³ HepG2 cells in each well of 96-well plates. Fresh media was added 24 h post-transfection, and the cells incubated for a further 4 days prior to the functional analysis. Cells were inspected for morphological changes under light microscopy (Fig. 2) and the growth inhibition measured quantitatively with MTT reduction assays. Of the 1,200 antisense species, 153 (~13%) were found to have varying inhibitory degrees on the growth of HepG2 cells. In contrast, cells treated with single-stranded control DNA (without antisense insert sequences) exhibited a mild level of growth inhibition, which could also be seen in cells treated with the double- stranded DNA-liposome complex. To eliminate redundancy, cDNA clones reversely complementary to the 153 growth- inhibiting LC-antisense molecules were sequenced and then searched for in the GenBank database.

There were 58 unique sequences out of the 153 cDNAs, which were designated as clones WGSL 1 to 58 (Table 1). From the cDNA sequence analysis, the antisense region of each clone was revealed. For instance, LCAS 3, 5, 8, 9, 10 and 23 contained the antisense sequences reversely complementary to the cDNA site 1367-1885 of WGSL 3, 21-489 of WGSL 5, 1448-2017 of WGSL 8, 421-863 of WGSL 9, 2119-2446 of WGSL 10 and 321-512 of WGSL 23, respectively. The LC-antisense molecules of the 58 clones were designated as LCAS 1 to 58. The growth inhibiting activities of the LC- antisense molecules were further confirmed by repetitive MTT assays (Fig. 3). As shown in Table 1, the GenBank annotation and motif-based searches suggested that 18 out of the 58 genes encode proteins of novel functions. The remaining 40 have an additional function to those previously characterized (e.g. protein synthesis, translation factors, metabolism and so on). These 58 genes may have functions directly or indirectly related to liver cancer cell growth.

The 58 LC-antisense species inhibitory to cell growth were studied further to reconfirm their roles in cell growth inhibition and to understand the underlying molecular mechanisms of their inhibitory effects. To accomplish this, flow cytometric analysis was used to detect changes of the cell cycle patterns in HepG2 cells treated with the selected LC-antisense species for 48 h. When compared to the sham or control treatments employing liposome alone or single-stranded control DNA+liposome complexes, 53 (~91%) out of the 58 LC-antisense molecules exhibited varying degrees of an increasing percentage of cells with a sub-G0-G1 DNA content (Fig. 4).

Whether these cell deaths, caused by the antisense treatments, reflected the induction of apoptosis were then analyzed. To do this, HepG2 cells treated with LC-antisense molecules were subjected to a DNA fragmentation assay. Twenty seven (LCAS 1, 3, 5, 8, 11, 12, 14, 15, 16, 20, 21, 22, 24, 29, 31, 33, 34, 35, 40, 41, 42, 43, 47, 49, 50, 52 and 54) out of the 58 LC-antisense species were found to cause characteristic DNA ladder formation 48 h after the transfection, indicating apoptotic progression due to the antisense treatment (Fig. 5). These results suggest that the LC-antisense library system is an effective means for en mass identification and validation of genes involved in cancer cell growth.