Cell culture media, fetal bovine serum (FBS), G418 and Lipofectamine 2000 reagent were purchased from Invitrogen Life Technologies (Carlsbad, CA, USA). Primary antibodies against Rb and phospho-Rb (Ser807/811) were purchased from Cell Signaling Technology (Beverly, MA, USA). Primary antibodies against LKB1 and GAPDH were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Horseradish peroxidase (HRP)-conjugated secondary antibodies were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA, USA). Unless otherwise noted, all chemicals and reagents were purchased from Sigma-Aldrich (Taufkirchen, Germany) and were of the highest grade.

Human hepatic cell line L02 was purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai) and cultured in RPMI-1640 medium containing 20% FBS. HEK-293T and HeLa cells, and HUVECs, were maintained in minimum essential medium (MEM) supplemented with 10% FBS. Cells were cultured in a humidified incubator at 37°C containing 5% CO₂. Cultured cells were used between passage 3 and 10. Transfection was performed using Lipofectamine 2000 according to the manufacturer's protocols. L02 cells were transfected with either an empty mock plasmid or a plasmid expressing wild-type LKB1, as previously described (19). After selection for 16 days in G418 (600 µg/ml)-containing medium, the resistant isolated single clones were selected, expanded and propagated. LKB1 expression was verified by northern blot and western blot analyses, respectively. One maximally expressing clone from the LKB1-transfected L02 cells was obtained and maintained in the same manner as the parental cells.

Cell proliferation was examined using the CellTiter 96 AQueous One Solution Cell Proliferation Assay (Promega, Madison, WI, USA) according to the manufacturer's protocol. Cells (5,000/well) were plated into 96-well plates in triplicates for 96 h and incubated in culture medium containing MTS every day for 4 days. SpectraMax M5 (Molecular Devices, Sunnyvale, CA, USA) was used to detect the absorbance at 490 nm. Results are presented as the average absorbance of 3-wells/experiment.

Cells were seeded in complete media at a density of 1×10³ cells in 60-mm dishes containing a top layer of 0.35% low melting point (LMP) agar and a bottom layer of 0.6% LMP agar. The plates were incubated at 37°C in 5% CO₂ for 4 weeks. Every 7 days, fresh medium was added to each plate. Crystal violet 0.2% was used to stain the plates. Visible colonies were photographed and counted manually.

Total proteins were analyzed by SDS-PAGE electrophoresis and blotted as previously described (19,20). Specific primary anti bodies recognizing LKB1 (1:500), Rb (1:1,000), phospho-Rb (1:1,000) and GAPDH (1:1,000) were used at the indicated dilutions. HRP-conjugated secondary antibodies were used at 1:10,000 dilutions. The signals of HRP were detected by Substrate SuperSignal West Pico Chemiluminescent Substrate kit (Pierce Biotechnology, Rockford, IL, USA). The protein signal was quantified using Quantity One software (Bio-Rad, Hercules, CA, USA).

Total RNA was isolated using TRIzol (Invitrogen Life Technologies), resolved in 1% formaldehyde-agarose gels, transferred to nylon membranes and cross-linked. Blots were probed with digoxigenin (DIG)-labeled cDNAs. The LKB1 probe was prepared using the forward primer (5′-TCGGTGGGTATGGACACGTTC-3′) and the reverse primer (5′-GTTCGTACTCAAGCATCCCTTTCA-3′), and labeled with DIG using End Tailing kit (Roche Applied Science, Indianapolis, IN, USA). β-actin gene was used as a loading control. Hybridization signals were detected with chemiluminescence and recorded on x-ray film.

Total RNA (2 µg) was transcribed with M-MLV reverse transcriptase and used for subsequent PCR amplification with Taq polymerase (both from Promega). PCR products were analyzed on 2% agarose gels and verified by DNA sequencing. Housekeeping gene β-actin was amplified as an internal control. LKB1-specific primers designed to detect LKB1 transcriptional variants are described in Table I.

Aliquots of the particular cell populations were counted and injected into the axillary region (2×10⁶ cells/injection site) of 6-week-old male BALB/c nude mice (Animal Center of Chongqing Medical University, Chongqing, China). The mice were housed in sterile cages under laminar flow hoods in a temperature-controlled room and were fed autoclaved chow and water ad libitum. All mice (n=4/group) were weekly weighed and examined every other day for morbidity and for tumor growth (measured using a digital caliper). After 6 weeks, all of the mice were euthanized with excess CO₂ and necropsied to determine possible gross metastases.

All experiments using mice were in accordance with the Guide for the Care and Use of Laboratory Animals and was approved by the Institutional Animal Care and Use Committee (IACUC) of Chongqing Medical University (Chongqing, China).

Tumor tissues were fixed in 4% paraformaldehyde and embedded in paraffin wax before sectioning at 5-µm thickness. For histopathological examinations, the sections were stained with hematoxylin and eosin (H&E). For immunohistochemistry, serial sections were deparaffinized, treated with 0.3% H₂O₂ in methanol to inactivate endogenous peroxidase, and incubated with anti-LKB1 monoclonal antibody at a 100-fold dilution for 60 min. Immunodetection of LKB1 was performed using Vector Stain Elite kit (Vector Research, Burlingame, CA, USA). A positive control was included with each batch of staining to ensure consistency between consecutive runs. Immunostaining was evaluated by a pathologist.

All cell culture experiments were repeated at least 3 times, unless otherwise indicated, and paired t-tests were used to determine statistical significance.

We first examined the protein expression of LKB1 in the L02 cell line by western blot analysis. The HeLa cell line, which has been extensively characterized as endogenous LKB1-deficient (21), was included as a negative control. Human embryonic kidney 293T cells (HEK-293T cells) and human umbilical vein endothelial cells (HUVECs) were used as positive controls. As depicted in Fig. 1A, a specific band for LKB1 was detected in 25 µg total protein of HEK-293T cells and HUVECs, whereas the LKB1 signal was not detectable in 180 µg total protein of L02 cells and 140 µg total protein of HeLa cells, suggesting that LKB1 protein expression is severely compromised in L02 cells. In concert with the deficiency of LKB1 protein, northern blot analysis indicated that LKB1 mRNA levels were barely detectable in the L02 and HeLa cells, in contrast to the HEK-293T cells which expressed abundant LKB1 transcript (Fig. 1B), supporting the lack of LKB1 expression in the L02 cell line.

To better evaluate the genetic defects underlying LKB1 deficiency in the L02 cell line, we performed LKB1 mutation analysis. The transcript was screened for putative mutations by sequencing of products obtained by reverse transcription followed by polymerase chain reaction (RT-PCR) from the L02 cell line. cDNA from L02 cells was amplified with 6 sets of primers covering the entire coding region (exons 1–9) of the LKB1 gene. These primers are described in Table I. As shown in Fig. 1C, compared with the HEK-293T cells and HUVECs, LKB1 transcription was severely affected in the L02 cells particularly for the 5′ end region. LKB1 mRNA expression was reduced to ~5% of the HEK-293T cells with upstream primer P1, 10% with midstream primer P2, and 25% with downstream primer P3. The majority of the LKB1 coding sequence (93%) was amplified from the L02 cells at very low levels by the primers chosen, except for the region containing the initial 89 bases, which was unable to be obtained by RT-PCR. Direct sequencing of the PCR products did not detect any abnormalities in nucleotides 90-1302 of the LKB1 open reading frame, suggesting the putative mutation region may be narrowed down to 89 bases downstream of the start codon. Another possibility for silencing of LKB1 expression is epigenetic inactivation caused by promoter hypermethylation (22). However, treatment with the demethylating agent 5-aza-2′-deoxycytidine could not restore mRNA or protein expression of LKB1 in our research (data not shown).

To obtain further insights into the biological role of LKB1, we attempted to reconstruct the in vivo situation of L02 cells by stably expressing wild-type LKB1. L02 cells were transfected with an expression vector encoding both LKB1 and a neomycin resistance gene or a vector encoding the selection marker only. Transfectants were subjected to G418 selection for 16 days. After the selection, a strongly reduced number of colonies were detected in the LKB1 transfectants compared with the mock vector transfectants of the L02 cells. Individual colonies were selected and analyzed for inducible expression of LKB1 by northern blot and western blot analyses. As shown in Fig. 2, recombinant LKB1 was readily detected at high levels in the L02 cells stably transfected with the LKB1 construct, but not in the cells transfected with the control vector. Thus, the reconstruction of LKB1 gene expression was successfully achieved in the L02 cells.

Downregulation of LKB1 in HeLa and G361 cells provided a growth advantage to these cells (21,23). To investigate this possibility in L02 cells, we performed MTS assay, which detects the activity of mitochondrial dehydrogenase enzymes of viable cells and is used as a measure for cell proliferation in cell cultures (24). Indeed, growth curve experiments clearly demonstrated that LKB1 expression conferred a proliferative disadvantage to L02 cells when compared with the parental L02 and mock vector-transfected L02 cells (Fig. 3A), which suggests that ectopic LKB1 expression strongly inhibited the growth of L02 cells. No difference was noted in the cellular growth between the parental L02 and the vector-transfected L02 cells. Thus, introducing wild-type LKB1 into the L02 cells led to growth suppression.

Cell cycle machinery controls cell proliferation, and the retinoblastoma (Rb) protein is a well-known gate keeper of cell cycle progression (25). Compatible with the decrease in cell proliferation rate in vitro, stable transfection with the LKB1 plasmid, but not with the mock vector, considerably reduced phosphorylation of Rb at Ser807 and Ser811 in the L02 cells, with no differences observed in the total Rb protein level (Fig. 3B). These results are consistent with our previous study (19), and indicate that activation of LKB1 signaling leads to hypophosphorylation of Rb, which in turn contributes to reduced cell cycle progression and cell proliferation in L02 cells.

Anchorage-independent growth in soft agar is one of the hallmark characteristics of cellular transformation and uncontrolled cell growth, with normal cells typically not capable of growth in semi-solid matrices (26). Recent studies have provided compelling evidence that loss of LKB1 activity is a critical step in oncogenesis (27–30). To explore the oncogenic potential of LKB1-deficient L02 cells, a soft agar colony formation assay was carried out. Notably, L02 cells and the control vector transfectants grew efficiently in soft agar and formed numerous colonies, whereas the L02 cells with constitutive LKB1 expression exhibited a significant reduction in anchorage-independent growth on soft agar (Fig. 3C). These results unambiguously suggest that L02 cells gained induced tumorigenic phenotypes in vitro, and LKB1 expression is an obstacle for anchorage-independent growth of L02 cells.

These encouraging results prompted us to further study the potential carcinogenesis of L02 cells in vivo. A limiting dilution tumorigenicity assay was conducted utilizing subcutaneous injections into athymic nude mice without any protein support. As shown in Fig. 4A, both L02 parental cells and mock vector transfected L02 cells were capable of developing subcutaneous tumors in all of the inoculated nude mice. In contrast, L02 cells with stable LKB1 expression were unable to form any tumor in the mouse model. These data demonstrated that the L02 cell line is tumorigenic in vivo, which could be completely abolished by sustained LKB1 expression. No distant metastases to spleen, liver, kidneys, lungs or intestine in the sets of mice were observed upon necropsy.

H&E staining of tumor tissues revealed a trabecular pattern of cell growth, where tumor cells were arranged in plates of various thickness, separated by sinusoid vascular spaces. There was no discernable normal lobular architecture, although vascular structures were present. Cytologic features of malignancy were noted, with irregular nuclear contours, giant multinucleated cells, and increased nuclear/cytoplasmic ratio (Fig. 4B). These morphological appearances indicated the hepatic histogenesis of the malignant tumors in agreement with the cell lines inoculated, excluding the formation of the primary tumors. All histologic sections were analyzed by one of our expert pathologists.

A major question concerning the mechanism of tumor formation of the L02 cells was whether the tumors still expressed LKB1. Immunohistology revealed that the LKB1 protein was not detected in the subcutaneous tumors from the nude mice (Fig. 4B), with HUVECs grown on coverslips as a positive control. These results suggest that LKB1 inactivation had occurred prior to tumor emergence and loss of LKB1 expression is required as a precondition of tumor formation.