β₂-Microglobulin Promotes the Growth of Human Renal Cell Carcinoma through the Activation of the Protein Kinase A, Cyclic AMP–Responsive Element-Binding Protein, and Vascular Endothelial Growth Factor Axis

Materials and Methods

Reagents. Recombinant human β2M was purchased from Sigma (St. Louis, MO). Forskolin (Alexis Biochemicals, San Diego, CA) and N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide (H-89; Alexis Biochemicals) were used for the analysis of cAMP-dependent signaling pathway.

Cell culture. Human renal cell carcinoma SN12C cells (19) were cultured in MEM (Life Technologies, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (Bio Whittaker, Walkersville, MD), with 50 IU/mL penicillin and 50 μg/mL streptomycin (Life Technologies) in 5% CO₂ at 37°C.

Plasmid construction. For construction of the β2M expression vector, β2M cDNA was isolated by reverse transcription-PCR (RT-PCR) from cells and flanked with HindIII and XbaI cloning sites. After RT-PCR, the β2M cDNA (427 bp) was sequenced and subcloned into pcDNA3.1 expression vector (Invitrogen, Carlsbad, CA). The empty pcDNA3.1 expression vector was used as the control (Neo).

Stable transfection of β2M expression vector. Transfection was conducted using LipofectAMINE 2000 (Invitrogen) as recommended by the manufacturer. Briefly, SN12C cells were seeded in six-well plates at a density of 3 × 10⁵ per well 24 h before transfection in medium supplemented with fetal bovine serum. Cells were transfected using 4 μg of plasmid DNA (pcDNA3.1-β2M or pcDNA3.1-Neo) and 10 μL LipofectAMINE 2000 per well. To obtain stable transfectants, the pcDNA3.1-β2M or empty vector–transfected cells were cultured in the presence of 800 μg/mL geneticin sulfate. Several isolated clones were selected, and positive clones were identified by ELISA, RT-PCR, and immunoblot analyses.

ELISA. β2M protein secreted by the cells was measured in culture media by a commercial ELISA kit (R&D Systems, Minneapolis, MN). Briefly, exponentially growing cells were seeded at a density of 6 × 10⁵ per well in six-well dishes containing fetal bovine serum and cultured until 80% confluent. The medium was replaced with serum-free MEM, and conditioned medium was collected after 72 h. Expression of β2M was assessed according to the manufacturer's protocol (R&D Systems). Color develops by addition of tetramethylbenzidine, and the intensity is measured at 450 nm with dual wavelength correction at 620 nm. The concentration of total protein in conditioned medium was determined by the Bradford method using Coomassie plus protein reagent (Pierce, Rockford, IL). Secretion of vascular endothelial growth factor (VEGF) protein secreted by the cells was also measured in the conditioned medium by a commercial VEGF ELISA kit (R&D Systems).

Cell viability assay. Cell viability was determined with a colorimetric 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS; Promega, Madison, WI) assay. Briefly, exponentially growing cells were seeded at a density of 2 × 10³ per well in 96-well plates and cultured for 4 days. Twenty microliters of MTS reagent was added to each well containing 100 μL of fresh culture medium. After a 1-h incubation period, optical absorbance at 490 nm was measured.

Flow cytometric analysis. For cell cycle analysis, flow cytometric analysis of propidium iodide–stained nuclei was done. Briefly, cells were plated at a density of 5 × 10⁵ in a 60-mm dish overnight. The cells were collected by trypsinization and fixed with 70% ethanol. The fixed cells were incubated with 100 μg/mL RNase A (Sigma) for 30 min and stained with 25 μg/mL propidium iodide (Chemicon, Temecula, CA) for 30 min. Cell cycle was analyzed with a FACScan flow cytometer and CellQuest software (Becton Dickinson Labware, Lincoln Park, NJ). The data were expressed as a mean percentage from three independent experiments.

Colony-forming assay. Intrinsic anchorage-independent growth activity in vitro closely reflects the tumorigenicity of the epithelial cells (20). The cells were cultured in a two-layer Matrigel system (BD Biosciences, Bedford, MA) to prevent their attachment to the plastic surface. Cells (1 × 10³) were trypsinized to single-cell suspensions, resuspended in Matrigel (150 μL, 3.7 mg/mL), and added to a pre-set Matrigel layer in 24-well plates. The top Matrigel cell layers were covered with culture medium containing 10% fetal bovine serum. Colonies >200 μm in 10 randomly selected fields were scored 14 days after plating.

Invasion and migration assays. Cancer cell invasion and migration were assayed in Companion 24-well plates (Becton Dickinson Labware) with 8-μm-porosity polycarbonate filter membrane. A suspension of 5 × 10⁴ cells in 100 μL of medium was layered in the upper compartment of a well with 15 μg/mL collagen I (BD Biosciences) as attractant in the lower compartment. After incubation at 37°C for 48 h, the cells on the upper surface of the filters were removed by swabbing with a cotton swab, and cells that had migrated to the lower surface were stained with 0.5% crystal violet. After washing, stain was eluted from migratory cells with Sorensen's solution. The absorbance of each well was measured at 590 nm. For invasion assays, the same procedures described above were used, except filters were precoated with Matrigel (BD Biosciences) at a 1:4 dilution with medium.

RNA preparation and RT-PCR reaction. Total RNA was isolated from confluent cell monolayers using the RNeasy Mini kit (Qiagen, Valencia, CA). Total RNA (2-5 μg) was used as template, and oligo-dT (0.5 μg) was added for reverse transcription and amplification in a reaction volume of 20 μL according to the manufacturer's instructions (Invitrogen). After reverse transcription reaction, first-strand cDNA (2 μL) was used for PCR with a PTC-100 programmable thermal controller (MJ Research, Inc., Waltham, MA). The oligonucleotide primer sets used for PCR analysis of cDNA are human β2M (427 bp) 5′-ACGCGTCCGAAGCTTACAGCATTC-3′ (forward) and 5′-CCAAATGCGGCATCTAGAAACCTCCATG-3′ (reverse), human VEGF (404 bp) 5′-CGAAGTGGTGAAGTTCATGGATG-3′ (forward) and 5′-TTCTGTATCAGTCTTTCCTGGT-3′ (reverse), human neuropilin 1 (210 bp) 5′-AGGACAGAGACTGCAAGTATGAC-3′ (forward) and 5′-AACATTCAGGACCTCTCTTGA-3′ (reverse), and glyceraldehyde-3-phosphate dehydrogenase (450 bp) 5′-ACCACAGTCCATGCCATCA-3′ (forward) and 5′-TCCACCACCCTGTTGCTGT-3′ (reverse). The thermal profile for human β2M amplification is 40 cycles, starting with denaturation of 1 min at 94°C followed by 1 min of annealing at 64°C and 30 s for extension at 72°C. For human VEGF and neuropilin 1 amplification, the thermal profile is 35 cycles, starting with denaturation for 15 s at 94°C followed by 30 s of annealing at 60°C and 1 min of extension at 68°C. The program for glyceraldehyde-3-phosphate dehydrogenase amplification is 25 cycles, starting with denaturation for 30 s at 94°C followed by 30 s of annealing at 60°C and 1 min of extension at 72°C. The RT-PCR products were analyzed by 1.0% agarose gel electrophoresis. Quantity one-4.1.1 Gel Doc gel documentation software (Bio-Rad, Hercules, CA) or NIH Image (version 1.55) was used for quantification of each mRNA expression normalized by glyceraldehyde-3-phosphate dehydrogenase mRNA expression.

Immunoblot analysis. Protein was extracted from cell pellets with a lysis buffer [50 mmol/L Tris (pH 8), 150 mmol/L NaCl, 0.02% NaN₃, 0.1% SDS, 1% NP40, 0.5% sodium deoxycholate, and 1 mmol/L phenylmethylsulfonyl fluoride] in the presence of protease inhibitor cocktail (Roche Applied Science, Indianapolis, IN). Samples containing equal amounts of protein (30 μg) were electrophoresed on 8% to 16% Tris-Glycine gels (Invitrogen) and transferred to nitrocellulose membranes. After blocking with T-TBS containing 5% nonfat milk powder, the membranes were incubated with mouse monoclonal antibody against β2M (1:500 dilution; Santa Cruz Biotechnology, Santa Cruz, CA) or rabbit polyclonal antibodies against PKA C (1:1,000 dilution; Cell Signaling, Danvers, MA), phospho-PKA C (Thr¹⁹⁷; 1:1,000 dilution; Cell Signaling), CREB (1:1,000 dilution; Cell Signaling), phospho-CREB (Ser¹³³; 1:1,000 dilution; Cell Signaling), Akt (1:1,000 dilution; Cell Signaling), phospho-Akt (1:1,000 dilution; Cell Signaling), extracellular signal-regulated kinase (ERK; 1:1,000 dilution; Santa Cruz Biotechnology), phospho-ERK (1:1,000 dilution; BioSource, Camarillo, CA), VEGF (1:200 dilution; Santa Cruz Biotechnology), VEGFR-2 (Flk-1; 1:1,000 dilution; Cell Signaling), neuropilin 1 (1:500 dilution; Santa Cruz Biotechnology), caspase-3 (1:1,000 dilution; Cell Signaling), caspase-9 (1:500 dilution; Cell Signaling), and poly(ADP-ribose) polymerase (1:1,000 dilution; Cell Signaling), respectively, at 4°C overnight. After washing with T-TBS, the membranes were incubated with corresponding secondary antibodies, which were conjugated with horseradish peroxidase (Santa Cruz Biotechnology). The blots were stripped and reprobed with anti-β-actin antibody (1:5,000 dilution; Sigma). Immunoreactive bands were visualized with enhanced chemiluminescence (Amersham Pharmacia Biotech, Little Chalfont, United Kingdom). Relevant cell signaling pathways were confirmed by the use of appropriate pathway-specific inhibitors. To analyze inhibition of the PI3K/Akt and MAPK pathways, cells were treated with LY294002 (Calbiochem, Darmstadt, Germany) and U0126 (Calbiochem) for 30 min or 6 h, respectively.

Transient transfection of β2M small interfering RNA. SN12C cells were transiently transfected either with β2M small interfering RNA (siRNA) duplex (si-β2M; final concentration of 40 nmol/L) or a control siRNA (a random scrambled sequence: si-Scr; final concentration of 40 nmol/L) using LipofectAMINE 2000 according to the manufacturer's instructions. Sequences of the siRNA against β2M and control scramble synthesized by Invitrogen are 5′-UUGCUAUGUGUCUGGGUUU(dT)(dT)-3′ and 5′-UUCAUGUGUCUGUGGUGUU(dT)(dT)-3′, respectively. Following transfection, cells were subjected to MTS assay, and invasion and migration assays or immunoblot analysis with the antibodies against β2M, caspase-3, caspase-9, and poly(ADP-ribose) polymerase.

In vivo animal experiment. The effect of β2M on osteoclastogenesis and tumorigenicity in vivo was assessed in tumor-bearing nude mice. Four-week-old male athymic nude mice (19-21 g; BALB/c nu/nu mice, National Cancer Institute, Frederick, MD) were housed in accordance with the NIH guidelines using an animal protocol approved by the Institutional Animal Care and Use Committee of Emory University. Cells were trypsinized, and single-cell suspensions (1 × 10⁶ cells) of each subclone were then injected into both tibias (n = 8). The estimated volume of bone tumors was calculated by three axes (X, Y, and Z) measured from a radiograph using the formula of π/6XYZ (21). Tumor size was also quantified by measuring hind limb diameter every 5 days. For s.c. injection, 2 × 10⁶ cells were mixed 1:1 with Matrigel and then injected into the flank (n = 8). S.c. tumor mass was measured in two dimensions with calipers, and the tumor volume was calculated every 5 days according to the equation (l × w²) / 2, where l = length and w = width (22). Animals were sacrificed by CO₂ asphyxiation if the tumor volumes exceeded the allowable size, and tumor weights were measured. At the time of sacrificing the mice, both hind limbs and tumor tissues were harvested for immunohistochemistry and H&E staining. Selected specimens of tibia were subjected to micro-computed tomography analysis using established procedures (23).

Immunohistochemical protocol. After sacrifice, s.c. tumor tissues were fixed with 10% buffered formalin and embedded in paraffin. The tibias were removed and fixed with 10% buffered formalin and decalcified in 10% EDTA solution. Sections from the formalin-fixed, paraffin-embedded tissues were cut to 4 μm and deparaffinized in xylene followed by treatment with a graded series of ethanol and rehydration in PBS. For antigen retrieval, the slides were treated with a Target Retrieval Solution (S1699; DAKO, Carpinteria, CA) in a pressure cooker at 125°C and 20 p.s.i. for 30 s. Next, sections were immersed in 3% hydrogen peroxide and then incubated in 2% goat serum. The sections were incubated with the following antibodies overnight at 4°C: CREB (Cell Signaling), phospho-CREB (Ser¹³³; Cell Signaling), MIB-1 (DAKO) for Ki-67, anti-human VEGF (PeproTech, London, England), platelet/endothelial cell adhesion molecule 1 (M-20; Santa Cruz Biotechnology) for CD31, and β₂-microglobulin (BBM.1; Santa Cruz Biotechnology). They were diluted 100×, 40×, 1×, 100×, 600×, and 500×, respectively, with PBS containing 1% bovine serum albumin. After washing with PBS, sections were incubated with secondary antibodies, which were conjugated with peroxidase-labeled amino acid polymer (DAKO). The immune complex was visualized using a 3,3′-diaminobenzidine peroxytrichloride substrate solution (DAKO). Slides were then counterstained with hematoxylin and mounted. For bone samples, tartrate-resistant acid phosphate staining was also done for detecting osteoclasts. Osteoclasts were determined as tartrate-resistant acid phosphate–positive staining multinuclear cells using light microscopy. The number of osteoclasts per millimeter of bone was determined by examination of tartrate-resistant acid phosphate–stained sections at ×10 magnification.

In parallel, human renal nonmalignant and carcinoma tissue specimens were also stained with antibody against β2M (Santa Cruz Biotechnology) to confirm the expression of β2M. These tissue specimens were obtained from the Kidney Satellite Tissue Bank of Emory University under institutionally approved protocols.

Evaluation of β2M, pCREB/CREB, Ki-67, VEGF, and microvessel density. To assess β2M and VEGF expression, whole sections were scored semiquantitatively using a visual grading system based on the intensity of staining (−, negative; ±, equivocal; +, weak; ++, moderate; +++, strong), according to the intensity of chromogen deposition in the majority of cancer cells evaluated independently by two observers. Two groups were considered for the statistical analysis: intensity levels −, ±, and + and strong focal staining pattern or diffuse light staining were considered low-intensity staining, whereas ++, +++, and strong diffuse staining pattern were considered high-intensity staining. The evaluation of pCREB/CREB and Ki-67 expression was based on the proportion of positive-stained cells among a total of 1,000 cells that were counted. For the quantification of microvessel density, 10 random 0.739 mm² per field at ×200 magnification were captured, and microvessels were counted.

Statistical analysis. Values were expressed as means ± SE. Statistical analysis was done using the Student's t test or one-way ANOVA. Relationships between qualitative variables were determined using the χ² test. Ps < 0.05 were considered statistically significant.

Results

Expression of β2M in human renal tissue specimens. The expression of β2M was assessed by immunohistochemistry in 12 human renal tissue specimens containing both normal and renal cell carcinoma (two, five, and five patients of G₁, G₂, and G₃ clear cell carcinoma, respectively) within each specimen. The expression of β2M was detected in all renal cell carcinoma tissues (Fig. 1A, right ), but β2M staining was sparse and only found in the luminal border of normal renal tubules adjacent to renal cell carcinoma tissues (Fig. 1A, left). β2M was largely membrane bound in all cancer cells, and in some cases, it was also expressed focally in the cell cytoplasm.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Expression of β2M in tissue specimens of renal cell carcinoma and effects of β2M on cancer cell growth and behavior. A, immunohistochemical staining of β2M in human renal nonmalignant and carcinoma tissue specimens. Negative or weak β2M stains were observed in normal renal tissue (left), whereas renal cell carcinoma tissue (clear cell carcinoma, G₂) stained positively and strongly with anti-β2M antibody (right). Magnification, ×100. Inset, magnified image. Magnification, ×200. B, RT-PCR and immunoblot analyses of cell lysates derived from SN12C subclones. Relative expression values of β2M mRNA were normalized by the amounts of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA expression. C, β2M concentration was measured by ELISA in conditioned medium. ng β2M protein/μg total protein. The value of the control (SN12C/P) was taken as 1, and other values were calculated from this. Columns, mean from three independent experiments; bars, SE. *, P < 0.01 compared with SN12C/P and SN12C/Neo cells. D, the effects of β2M on cell proliferation in each subclone were determined by MTS assay. Points, mean from three independent experiments; bars, SE. *, P < 0.01 compared with SN12C/P and SN12C/Neo cells. E, the effects of β2M on anchorage-independent growth of SN12C subclones. After 14 d of incubation, visible colonies (>200 μm) were photographed (top) and counted (bottom). Columns, mean from three independent experiments collected from each experimental group; bars, SE. *, P < 0.01 compared with SN12C/P and SN12C/Neo cells. F, the effects of β2M on invasion and migration of SN12C subclones. A_{590 nm} values correspond to cells that migrated to the lower side of the filter. Columns, mean from three independent experiments; bars, SE. *, P < 0.05 compared with SN12C/P and SN12C/Neo cells.

Expression levels of β2M in human renal cell carcinoma cell lines and in parental and β2M-transfected SN12C subclones. All human renal cell carcinoma cell lines examined in this study expressed β2M protein as evaluated by immunoblot analysis. Compared with ACHN cells, SN12C and Caki-1 cells expressed higher levels of β2M by 2.0- and 1.8-fold, respectively. To understand the roles of β2M in cell proliferation and apoptosis, we chose to study SN12C cells because of its high levels of β2M expression, which can be subjected to overexpression or knockdown manipulation of β2M protein. Parental SN12C cells (SN12C/P) were stably transfected with the human β2M cDNA expression vector pcDNA3.1-β2M or pcDNA3.1-Neo vector alone as a control. We selected three clones labeled as having low (L), medium (M), or high (H) levels of β2M expression for further characterization. These clones were subjected to semiquantitative RT-PCR and immunoblot analyses to confirm the level of β2M expression (Fig. 1B). In the following experiments, we used two clones (SN12C/β2M-H expressing high β2M levels and SN12C/β2M-M expressing medium β2M levels) for molecular analysis. SN12C/P and the derivative SN12C/P transfected with empty pcDNA3.1 vector (SN12C/Neo) served as controls.

β2M protein secretion by parental and derivative SN12C subclones. We first examined the β2M protein level in culture medium as determined by ELISA in each subclone. β2M protein was undetectable in cell free MEM culture medium. The concentrations of β2M in the supernatants from β2M-overexpressing transfectants were significantly higher than those of SN12C/P and SN12C/Neo (P < 0.01; Fig. 1C). Moreover, β2M secretion in the conditioned medium correlated with the protein expression levels of β2M observed by immunoblot (Fig. 1B) in derivative SN12C subclones. These results confirm that β2M is a soluble factor secreted by SN12C cells.

Overexpression of β2M in SN12C promotes cell proliferation in vitro. To determine the possible role of β2M in promoting renal cell carcinoma cell growth, we examined the growth rates of each subclone by MTS assay. The growth rates of β2M-overexpressing transfectants were significantly higher than those of SN12C/P and SN12C/Neo (P < 0.01; Fig. 1D). SN12C/β2M-M and SN12C/β2M-H had significantly higher growth rates than SN12C/β2M-L (P < 0.01). These results suggest that β2M is an effective soluble growth-promoting factor for SN12C cells in vitro. We found that ectopic expression of β2M increases cell number by increasing cells in the proliferative phase as assessed by flow cytometry. SN12C/β2M-H cells were observed to have a substantial decreased percentage of cells in the G₀-G₁ phase (36.8 ± 2.1%, P < 0.01), with a corresponding increased percentage of cells in S (43.9 ± 0.6%, P < 0.01) and G₂-M phase (19.4 ± 2.7%) when compared with SN12C/Neo cells (G₀-G₁: 54.0 ± 3.8%, S: 28.8 ± 2.8%, G₂-M: 17.3 ± 1.0%). These results collectively suggest that β2M increases renal cell carcinoma cell number primarily through increased cell proliferation.

Effect of β2M on anchorage-independent growth of SN12C cells in Matrigel. To determine the effect of β2M expression on anchorage-independent growth of SN12C cells, we did a colony formation assay in Matrigel. The numbers of colonies formed by β2M-overexpressing transfectants (SN12C/β2M-M: 130 ± 9 and SN12C/β2M-H: 157 ± 9) were significantly higher than those of SN12C/P (72 ± 4, P < 0.01), SN12C/Neo (68 ± 4, P < 0.01), and SN12C/β2M-L (77 ± 4, P < 0.01; Fig. 1E).

Effect of β2M on the invasion and migration of SN12C cells. We next analyzed whether the β2M expression level correlated with the in vitro invasiveness of cells as tested by their capacity to pass through a Matrigel barrier as assayed in a modified Boyden chamber. In comparison with SN12C/P and SN12C/Neo, SN12C/β2M-M and SN12C/β2M-H presented with a 1.6- and 1.9-fold higher invasiveness, respectively, at 48 h after seeding (P < 0.05; Fig. 1F). The ability of β2M-overexpressing transfectants to invade through Matrigel increased with increased β2M expression. This increased invasiveness led us to examine if β2M may also affect cell migration, a key component of the invasive process. β2M overexpression increased cell migration in SN12C/β2M-M and SN12C/β2M-H by 1.7- and 1.9-fold, respectively, in comparison with SN12C/P and SN12C/Neo (P < 0.05; Fig. 1F). These results suggest that β2M increases in vitro SN12C cell invasion and migration.

β2M activates the PKA-CREB-VEGF axis in SN12C cells. To determine whether phosphorylation of CREB may be mediated through an activation of the PKA signaling pathway induced by β2M, we assessed the effects of β2M, forskolin (a PKA signaling pathway activator), and H-89 (an inhibitor of PKA signaling) on activation of phosphorylation of CREB in SN12C/P cells. We evaluated the effects of conditioned media (0.5 μg/mL) from SN12C/β2M-H cells, recombinant β2M protein (0.5 μg/mL), or forskolin (5 μmol/L) on the activation of phosphorylation of CREB in SN12C/P cells. In parallel, we also evaluated the effects of H-89 (5 μmol/L), which inhibited phosphorylation of CREB in cells treated with conditioned media containing β2M, recombinant β2M, or forskolin with VEGF expression as the signal output. The expression of VEGF roughly paralleled the levels of phosphorylation of CREB (Fig. 2A ). Activation of cAMP-dependent PKA signaling pathway by β2M or forskolin resulted in increased phosphorylation of CREB and increased expression of VEGF, and these effects can be blocked by H-89 in SN12C/P cells (Fig. 2A). In response to β2M overexpression, we found elevated levels of phosphorylation of CREB/activating transcription factor-1 and VEGF165 and VEGF121 mRNAs and VEGF protein associated with SN12C subclones that expressed variable levels of β2M (Fig. 2B). Consistent with these results, we also found 2.5- and 3.1-fold increased VEGF secretion by SN12C/β2M-M and SN12C/β2M-H cells, respectively (P < 0.05; Fig. 2C), when compared with SN12C/P cells. Next, we investigated the expression of VEGFR-2 (Flk-1) and its co-receptor neuropilin 1 by RT-PCR and immunoblot analyses. Although there was no change in the expression level of VEGFR-2, RT-PCR and immunoblot analyses showed that neuropilin 1 mRNA and protein expression were elevated in SN12C/β2M-M and SN12C/β2M-H compared with SN12C/P and SN12C/Neo (Fig. 2D). These results suggest that activation of PKA-CREB signaling elevated VEGF and VEGFR-2 co-receptor neuropilin 1 expression in SN12C cells.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

β2M-mediated PKA-CREB signaling pathway and effects of β2M on the expression levels of CREB, VEGF, VEGFR-2, neuropilin 1, ERK1/2, and Akt and on the activation of CREB, ERK1/2, and Akt in SN12C subclones. A, SN12C/P cells were exposed to SN12C/β2M-H conditioned medium (CM; 0.5 μg/mL), recombinant human β2M protein (0.5 μg/mL) and forskolin (FSK; 5 μmol/L) for 24 h. H-89 (5 μmol/L) was added to SN12C/P cells for 2 h and then exposed to SN12C/β2M-H conditioned medium (0.5 μg/mL), recombinant human β2M protein (0.5 μg/mL), and forskolin (5 μmol/L) for an additional 24 h. Phosphorylation of PKA C and CREB was determined by immunoblot analysis using phospho-PKA C (Thr¹⁹⁷) and phospho-CREB (Ser¹³³) antibodies, and levels of expression of PKA C, CREB, and VEGF were also determined by immunoblot analyses. B, levels of expression of CREB and VEGF were determined by RT-PCR and immunoblot analyses in each subclone. Phosphorylation of CREB was determined by immunoblot analysis using phospho-CREB antibody in SN12C subclones. C, VEGF concentration was measured by ELISA in conditioned medium of each subclone. pg VEGF protein/μg total protein. The value of the control (SN12C/P) was taken as 1, and other values were calculated from this. Columns, mean from three independent experiments; bars, SE. *, P < 0.05 compared with SN12C/P and SN12C/Neo cells. D, levels of expression of VEGFR-2 (Flk-1) and its co-receptor neuropilin 1 were determined by RT-PCR and immunoblot analyses in each subclone. Immunoblot was done using VEGFR-2 (Flk-1) and neuropilin 1 antibodies. E, phosphorylation of Akt and ERK1/2 was determined by immunoblot analysis using phospho-Akt (Ser⁴⁷³) and phospho-ERK1/2 (Thr¹⁸⁵/Tyr¹⁸⁷) antibodies in each subclone. Blots were stripped and reprobed with antibodies against total Akt, total ERK, and β-actin. F, the transfectants were pretreated with U0126 (20 μmol/L) or LY294002 (LY; 25 μmol/L) for 30 min (left) or 6 h (right) and cultured for 12 h. Immunoblot analysis was done using antibodies that specifically recognize total and phospho-Akt and ERK1/2 (top). Cell survival was then determined by MTS assay (bottom). Columns, mean from three independent experiments; bars, SE. *, P < 0.05 compared with nontreatment SN12/Neo or SN12C/β2M-H cells.

Differential regulation of PI3K/Akt and MAPK pathways by β2M in SN12C cells. To determine if there is a convergence of cell signaling among β2M/PKA/CREB, PI3K/Akt, and MAPK, we examined the basal and β2M-activated status of the PI3K/Akt and MAPK signaling pathways. The status of phosphorylation of both kinases was analyzed using phospho-specific antibodies. As shown in Fig. 2E, Akt was highly phosphorylated in SN12C/β2M-M and SN12C/β2M-H compared with SN12C/P and SN12C/Neo cells. We also found moderate increase in the phosphorylation of ERK in SN12C/β2M-M and SN12C/β2M-H compared with SN12C/P and SN12C/Neo cells. These results suggest that activation of PI3K/Akt and MAPK pathways can be triggered as the consequence of ectopic expression of β2M in SN12C cells. β2M has a profound effect on the PI3K/Akt pathway but a more modest effect on the phosphorylation of ERK, although protein levels of non-phosphorylated Akt and ERK were not affected by β2M transfection. Thus, the PI3K/Akt pathway seems to play a more dominant role in conferring β2M-mediated increases in cell proliferation in SN12C cells.

Consistent with these observations, we evaluated the effect of LY294002 (a PI3K inhibitor) and U0126 (a MAPK inhibitor) on these respective pathways in SN12C cells overexpressing β2M. SN12C/Neo cells served as controls. Twenty micromolars of U0126 treatment for 30 min and 6 h significantly decreased SN12C/Neo cell numbers as determined by MTS assay (P < 0.05). In contrast, although 20 μmol/L U0126 treatment for 30 min blocked phosphorylation of ERK induced by β2M in SN12C/β2M-H, this treatment had no significant effect on cell number (Fig. 2F, left). The same dose of U0126 treatment for 6 h also completely blocked phosphorylation of ERK, whereas it did not significantly decrease cell number in SN12C/β2M-H cells (Fig. 2F, right). When cells were treated with 25 μmol/L LY294002 for 30 min, this treatment had no significant effect on Akt phosphorylation in SN12C/β2M-H but had a significant depressant effect of the SN12C/Neo cells (Fig. 2F, left). When cells were treated with 25 μmol/L LY294002 for 6 h, phosphorylation of Akt was almost completely blocked in SN12C/β2M-H and SN12C/Neo cells (Fig. 2F, right). The suppression of Akt phosphorylation by LY294002 correlated with the enhanced cell death in SN12C/β2M-H and SN12C/Neo cells (P < 0.05). These results in aggregate indicate that the PI3K/Akt signaling pathway plays a crucial role in cell proliferation in β2M-transfected cells, although cell proliferation regulated by β2M, at least in part, originated from an activated MAPK signaling activity.

β2M overexpression induces accelerated SN12C tumor growth in nude mice. To determine whether β2M overexpression may enhance SN12C tumor growth in mice, we compared the growth of the SN12C/β2M-H and SN12C/Neo cells injected s.c. in athymic nude mice. Figure 3A shows photographs of s.c. xenografted tumors in nude mice. Mice were sacrificed on day 90 after s.c. injection. Figure 3B shows that the mice injected with SN12C/β2M-H formed significantly bigger tumors compared with SN12C/Neo at the time the mice were sacrificed (mean volume = 318 ± 116 mm³ versus 1,306 ± 371 mm³; SN12C/Neo versus SN12C/β2M-H, n = 8 for each group; P < 0.01). Figure 3C also shows that tumor weight in SN12C/β2M-H group was significantly elevated by 2.9-fold compared with SN12C/Neo group (P < 0.01). Histologic evaluation showed that SN12C/Neo tumors were composed of numerous necrotic areas compared with SN12C/β2M-H tumors grown as xenografts s.c. (Fig. 3G, arrows). Immunohistochemical staining using anti-β2M antibody confirmed β2M expression in SN12C/β2M-H tumors with strong and heterogeneous membrane staining of β2M in comparison with SN12C/Neo tumors (Fig. 3G; Table 1 ).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

S.c. and intratibial tumor growth in nude mice. A, SN12C/Neo or SN12C/β2M-H cells were injected bilaterally s.c. into nude mice, forming four tumors per mouse. Representative macroscopic appearance of nude mice injected with SN12C/Neo clone or SN12C/β2M-H s.c. Arrows indicate s.c. tumor formation. B, tumor volumes were measured every 5 d (only 10-d tumor volumes were plotted). Points, mean of eight tumors for each group; bars, SE. *, P < 0.01 compared with SN12C/Neo. C, animals were euthanized and tumors were excised and weighed. Columns, mean of tumor weights for each group; bars, SE. *, P < 0.01 compared with SN12C/Neo. D, SN12C/Neo or SN12C/β2M-H cells were injected intratibially into nude mice. Representative radiographs of nude mice tibia at autopsy. Radiographs revealed marked osteolytic lesions with occasional foci of strongly osteoblastic lesions in the SN12C/β2M-H–implanted tibias compared with SN12C/Neo–implanted tibias. E, micro-computed tomography images revealed massive osteoblastic and osteolytic reactions in tibias implanted with SN12C/β2M-H compared with SN12C/Neo–implanted tibias. F, tumor size was quantified every 5 d by measuring hind limb diameter (only 10-d tumor volumes were plotted). Points, mean of eight tibias for each group; bars, SE. *, P < 0.01 compared with SN12C/Neo. G, H&E, immunohistochemical staining of β2M, and tartrate-resistant acid phosphate (TRAP) staining in SN12C/Neo–implanted (top) and SN12C/β2M-H–implanted (bottom) tumors. H&E subcutis: magnification, ×100. Tibia: magnification, ×40. β2M staining: magnification, ×200. Multinuclear cells that stained red after tartrate-resistant acid phosphate staining determined osteoclasts. Magnification, ×200.

We also evaluated the behavior of the β2M-overexpressing clone in nude mouse bone. Radiographs showed bigger lesions for tibias implanted with SN12C/β2M-H compared with tibias implanted with SN12C/Neo (Fig. 3D). As shown in Fig. 3E, massive osteoblastic and osteolytic changes in tibias implanted with SN12C/β2M-H were also observed by micro-computed tomography. The tumor size was significantly greater in tibias implanted with SN12C/β2M-H compared with tibias implanted with SN12C/Neo (n = 8 for each group; P < 0.01; Fig. 3F). The estimated volume of bone tumors was also measured by radiograph at the end point of the animal experiment. End point tumor volume in SN12C/β2M-H tumors was significantly increased by 2.7-fold compared with SN12C/Neo tumors [median volume (range): 78.2 (14.7-257.9) mm³ versus 214.8 (101.1-800.1) mm³; SN12C/Neo versus SN12C/β2M-H, n = 8 for each group; P < 0.001]. Histology revealed that unlike SN12C/Neo tumor cells, SN12C/β2M-H tumor cells seem to be highly aggressive replacing bone marrow and destroying much of the cortical shafts (Fig. 3G). Immunohistochemical staining using anti-β2M antibody confirmed that SN12C/β2M-H tumors showed strong cytoplasmic staining with variable degrees of membrane staining compared with SN12C/Neo tumors harvested from mice tibias (Fig. 3G; Table 1). We observed a large number of tartrate-resistant acid phosphate–positive osteoclasts at the bone/tumor interface in SN12C/β2M-H tumors, indicating that osteoclasts, directly adjacent to tumor cells, were with increased activity (Fig. 3G). Histomorphometric analysis of SN12C/β2M-H tumors showed a 3.0-fold increase of osteoclasts compared with SN12C/Neo tumor (4.5 ± 0.7 osteoclasts/mm versus 13.6 ± 2.1 osteoclasts/mm; SN12C/Neo versus SN12C/β2M-H; n = 8 for each group; P < 0.01). These results suggest that β2M-overexpressing renal cell carcinoma cells enhanced osteolysis through accelerated osteoclastogenesis.

Immunohistochemical analysis.Table 1 summarizes the immunohistochemical findings. For s.c. tumors, although equivalent anti-CREB antibody staining was observed in SN12C/Neo and SN12C/β2M-H tumors, substantial higher anti-pCREB antibody stained cell nuclei were observed in SN12C/β2M-H than those of SN12C/Neo tumors (Fig. 4 ). The mean percentage of pCREB⁺/CREB⁺ tumor cells in SN12C/Neo tumors and SN12C/β2M-H tumors was 52 ± 5% and 83 ± 11%, respectively (P < 0.005). The mean percentage of Ki-67–positive tumor cells in SN12C/β2M-H tumors was also significantly increased compared with SN12C/Neo tumors (31 ± 8% versus 82 ± 8%, SN12C/Neo versus SN12C/β2M-H, P < 0.005). We analyzed angiogenesis by using anti-VEGF and anti-CD31 antibodies. All the tumor cells stained positively for VEGF; however, tumor cells in SN12C/β2M-H tumors stained strongly and uniformly for VEGF compared with SN12C/Neo tumors (Fig. 4). Microvessel density correlated with expression of VEGF in both groups. Both VEGF expression and microvessel density were found to be significantly higher in SN12C/β2M-H tumors compared with SN12C/Neo tumors (Table 1). Comparison of immunohistochemistry of bone versus s.c. tumors showed parallelism of expression of all these proteins with comparable level of differences (Fig. 4; Table 1).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Immunohistochemical determination of CREB, phosphorylated CREB (pCREB), Ki-67, VEGF, and CD31 in s.c. and intratibial tumors in nude mice. S.c. and intratibial tumors from SN12C/Neo or SN12C/β2M-H were harvested and processed for immunohistochemical analysis. The sections were stained for CREB, pCREB, Ki-67, VEGF (magnification, ×200), and CD31 (magnification, ×100).

Effect of β2M siRNA on cell proliferation, apoptosis, invasion, and migration. Because β2M conferred increased SN12C cell growth in vitro and in vivo, we tested the possibility that β2M-induced intracellular signaling may be a therapeutic target. SN12C cells were transiently transfected with either si-β2M or si-Scr. After 72 h, cells were harvested and subjected to immunoblot analysis, showing that si-β2M effectively down-regulated the expression of β2M (Fig. 5A, left ). This inhibition persisted in cells 7 days after transfection (Fig. 5A, right). Next, we examined the growth rates of each cell by MTS assay. si-β2M significantly inhibited the growth of SN12C cells compared with nontransfected parental (−), Lipo-Control, and si-Scr–transfected cells (P < 0.01; Fig. 5B). To investigate the molecular mechanism of this cell growth inhibition by si-β2M, activation of caspase-9 and caspase-3 and subsequent cleavage of poly(ADP-ribose) polymerase were investigated by immunoblot analysis. si-β2M but not si-Scr induced the activation and processing of caspase-9 and caspase-3 and cleavage of poly(ADP-ribose) polymerase (Fig. 5C). These results suggest that β2M siRNA reduced cell proliferation via induction of apoptosis in SN12C cells.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Effects of β2M siRNA on cell proliferation, apoptosis, invasion, and migration in SN12C cells. A, β2M siRNA (si-β2M), LipofectAMINE 2000 control (Lipo-Control), and control scramble siRNA (si-Scr) were added to the media using lipophilic transfection-enhancing reagent (LipofectAMINE 2000). Cells were harvested after 72 h, and immunoblot analysis was done using anti-β2M antibody (top). For duration of β2M inhibition in SN12C cells, SN12C cells were transfected with si-Scr or si-β2M, and cells were lysed 3, 5, 7, and 9 d after transfection as indicated (bottom). B, effects of si-β2M on cell proliferation of SN12C cells were determined by MTS assay. Points, mean from three independent experiments; bars, SE. *, P < 0.01 compared with nontransfected parental (−), LipofectAMINE 2000 control, and si-Scr–transfected cells. C, si-β2M but not si-Scr induced activation and processing of caspase-9 and caspase-3 and cleavage of poly(ADP-ribose) polymerase (PARP) in SN12C cells. D, effects of si-β2M on invasion and migration of SN12C cells. A_590nm values correspond to cells that migrated to the lower side of the filter. Columns, mean from three independent experiments; bars, SE. *, P < 0.05 compared with nontransfected parental (−) and si-Scr–transfected cells. E, proposed molecular mechanism, whereby β2M can affect cancer progression in human renal cell carcinoma. β2M can activate cAMP-dependent PKA activity through an unknown receptor, and then this activation can induce activation of CREB, which increases angiogenesis, cell proliferation, and survival via activation of VEGF signaling and its downstream cell survival pathways PI3K/Akt and MAPK.

We then tested whether the inhibition of β2M expression in SN12C cells was sufficient to decrease cell invasion and migration. As shown in Fig. 5D, in vitro invasion and migration of cells transfected with si-β2M was reduced by 60.2% and 46.5% of the control, respectively (P < 0.05). These results suggest that β2M siRNA negatively affected the behavior of SN12C cells.

Discussion

The association between β2M protein and cancer has received much attention in the past decades because of its participation in host immune defense mechanisms (2–4). Because β2M forms a complex and presents the MHC class I molecule to the cell surface, it is not surprising to note that malignant cancer cells lose MHC class I antigen, decreases steady-state level of β2M, and become immune evasive (5–8). However, data accumulated in the literature suggested otherwise. That is, most of the studies showed increased expression of β2M levels in renal cancer tissues, cells, and serum of cancer patients (10–18), but none of these studies defined the possible directive roles β2M in cell growth and signaling. To verify whether β2M has oncogenic activity and is a signaling molecule seems beneficial for establishing a new therapeutic target. The goals of this study are as follows: (a) to define the biological effects of β2M in human renal carcinoma cells in culture and in experimental animal models, (b) to characterize the downstream signaling pathways of β2M in a cell model of human renal cell carcinoma, and (c) to explore β2M as a novel target of therapy for human renal cell carcinoma. Results of this study have allowed us to expand the growth and cell signaling roles of β2M beyond stabilization and presentation of MHC class I molecule in cells. Our results also raise the question that β2M must have a far-reaching function than the common belief of a housekeeping gene in cells.

In the present study, we showed that (a) β2M promoted the proliferation, invasion, and migration of human renal cell carcinoma cells in vitro; (b) β2M increased renal cell carcinoma tumor growth in mouse s.c. space and in the skeleton; (c) β2M activated the phosphorylation of CREB through cAMP-dependent PKA signaling and up-regulated its target genes, which include the mRNAs and proteins of VEGF and neuropilin 1; (d) β2M activated not only β2M/PKA/CREB signaling but also activated its convergent signaling network PI3K/Akt and MAPK; and (e) down-regulation of β2M signaling by siRNA significantly inhibited renal cell carcinoma cell growth via induction of apoptosis and inhibition of cancer cell invasion and migration. Although the dose-dependent growth and signaling roles of β2M was shown only in a human renal cell carcinoma cell line, these effects are likely to be applicable to human renal cell carcinoma in general because we and others have shown that β2M expression is up-regulated in human renal cancer tissues (24–26), and that the growth and signaling roles of β2M were observed in other independent studies using human prostate (27), breast, and lung cancer cells.⁷ To our knowledge, this is the first report to show the possible directive growth and signaling roles of β2M in human renal cell carcinoma.

β2M, a well-known housekeeping gene, is a 12-kDa non-glycosylated polypeptide composed of 100 amino acids. It is one of the components of MHC class I molecules on the cell surface of all nucleated cells. Its major-characterized function is to interact with and stabilize the tertiary structure of the MHC class I α-chain (1). In addition, MHC molecules, including β2M, play an important role in regulating tumor immunity as well as cellular and humoral immunities (2, 3). The complex roles of β2M in cancer and bone metastasis, however, are descriptive and largely undefined. Increased synthesis and release of β2M, as indicated by an elevated serum or urine β2M concentration, occurs in several malignant diseases, including prostate cancer, lung cancer, renal cell carcinoma, myeloma, and lymphocytic malignancies as well as autoimmune and infectious diseases (10–17). In these malignancies, serum β2M level is a significant prognostic factor. Therefore, we hypothesized that β2M is not only a surrogate biomarker for tumor burden but may also regulate cancer cell growth and survival. To test this hypothesis, we established clonal cell lines of a human renal cell carcinoma (SN12C) that overexpresses β2M and examined whether β2M might have direct effects on cell growth, invasion, and migration in vitro in a dose-dependent manner and tumor growth in vivo. We also evaluated the possible intracellular signaling roles of β2M. In addition, we investigated whether β2M may be a new therapeutic target.

Our results clearly showed that β2M is an effective growth-promoting soluble factor for SN12C cells in vitro. This observation supports the conclusion that β2M plays an important role in regulating the growth and survival of cancer cells, and it seems compatible with the clinical data that high levels of serum β2M are associated with high tumor burden and poor prognosis (14, 28, 29). In addition, β2M-mediated cell growth was associated with cell cycle progression. It has been reported that β2M stimulates cell proliferation accompanied by a significant decrease of population doubling time in prostate cancer cell lines (9). There seems to be conflicting reports that either β2M (30, 31) or β2M antibody (32) suppresses the proliferation of myeloma cell lines. Take together, β2M is likely to have a direct mitogenic activity and play a role in modulating cell proliferation in solid tumors, probably in a cell context–dependent manner.

In this study, β2M expression levels were positively correlated with cancer cell growth, invasion, and migration in β2M-overexpressing transfectants. Although multiple mechanisms of β2M-mediated signaling can contribute to enhanced cancer cell growth and altered cell behaviors, little is known about the exact roles of β2M in signaling pathways at present. We showed that β2M activates CREB through cAMP-dependent PKA signaling and up-regulates the mRNA and protein levels of VEGF, resulting in increased cell proliferation. CREB is a bZIP transcription factor that transcriptionally activates a large number of downstream target genes, such as growth factors, angiogenesis factors, cell signaling–mediated genes, and genes involved in proliferation and survival, through cAMP response elements (33–36). Activation of CREB consequently could contribute to increased cancer cell growth and angiogenesis. VEGF, known as an important mediator of angiogenesis, is a highly specific mitogen for endothelial cells and cancer cells through receptors VEGFR-1 (Flt-1) and VEGFR-2 (Flk-1/KDR; ref. 37). Moreover, some reports have shown that PI3K and/or MAPK are important as a necessary signaling component of VEGF-mediated cell progression via ligand-receptor interaction (38, 39). In this study, we also found that increased secretion of VEGF and activation of both PI3K/Akt and MAPK signaling pathways occurred in β2M-overexpressing transfectants. Together with the results, activation of phosphorylation of CREB may induce the transcriptional activation of VEGF, which plays a critical role in human renal cell carcinoma.⁸ It is possible that signaling convergence may occur between VEGF-VEGFR and PI3K/Akt and MAPK signaling. Activation of CREB has also been reported to be activated by PI3K/Akt and MAPK (40–42). These results in sum suggest that β2M-mediated VEGF and VEGFR-2 co-receptor neuropilin 1 expression and signaling may be central to β2M-induced renal cell carcinoma cell growth, migration, and invasion in vitro and tumor growth in mouse s.c. space and skeleton.

Additionally, immunohistochemical study confirmed the relationship between β2M expression and cell proliferation and angiogenesis in vivo. β2M overexpression was associated with an increase in tumor cell proliferation (Ki-67) and angiogenesis (VEGF expression level and microvessel density) that correlated with a significant increase in phosphorylation of CREB in both s.c. and bone tumors. VEGF has been shown to be an important factor involved in the development of tumor blood supply in the progression of solid tumors (43). Numerous published reports describe the association of microvessel density with VEGF expression (44–46). VEGF can also substitute for pro-osteoclastogenic cytokine macrophage colony-stimulating factor and up-regulate receptor activator of nuclear factor-κB expression in osteoclast precursors, thus promoting osteoclastogenesis (47, 48). In the present study, we observed a large number of osteoclasts in β2M-overexpressed bone tumor. The tumors are largely osteolytic, although osteoblastic foci can be detected by both histopathology and X-ray of the skeleton. Our findings agree with previous reports that β2M promotes osteoclast formation (31), and that osteoclasts are essential for cell survival and proliferation in myeloma (49). It has been also reported that VEGF acts as an osteolytic factor in breast cancer bone metastasis (50). β2M may enhance tumor growth in bone via β2M-induced osteoclastogenesis and VEGF induction. Our results, therefore, support the current concept that β2M enhances tumor progression through enhanced angiogenesis, cell survival, and osteoclastogenesis when metastasized to bone. In this study, we showed that cancer cells may require intracellular β2M for their survival. For example, by interrupting β2M expression using its sequence-specific β2M siRNA, we observed activated apoptosis in SN12C cells. Thus, β2M signaling may be an attractive new therapeutic target for the treatment of human renal cell carcinoma.

In summary, the present study provides evidence that β2M enhances the growth and survival of human renal cell carcinoma cells via PKA/CREB activation, VEGF signaling, and cell survival signaling, including the PI3K/Akt and MAPK pathways (Fig. 5E). It is tempting to speculate that aberrant β2M expression facilitates tumor progression and bone metastasis, and that β2M signaling may be an attractive new therapeutic target. Further studies are required to determine the precise molecular mechanisms by which β2M regulates cancer cell growth, invasion, and migration, and, conversely, interruption of β2M signaling induced cancer cell apoptosis.