Intracellular expression of toll-like receptor 4 in neuroblastoma cells and their unresponsiveness to lipopolysaccharide

Human neuroblastoma cell lines, NB-1 and SK-N-SH, and human monocytic cell line, U-937, were obtained from Riken Cell bank (Tsukuba, Japan). NB-1 and U-937 cells were maintained in RPMI containing 5% heat inactivated fetal calf serum (GIBCO-BRL, Gaithersburg, MD, USA) and SK-N-SH cells was maintained in MEM with 10% heat-inactivated fetal calf serum. Cells were cultured in the presence of antibiotics at 37°C under 5% CO2.

NB-1 cells were seeded (1 × 10⁶/ml) in 35 mm dish for overnight, trypsinized and washed with phosphate-buffed saline (PBS). To detect cell surface expression of TLR4, the cells were incubated with PE-conjugated anti-TLR4 antibody for 20 min in dark on ice, washed and resuspended in PBS. To detect intracellular expression of TLR4, the cells were fixed with a fixation buffer (eBioscience) in dark and treated with a permeabilization buffer (eBioscience) according to the instruction of the manufacturer. The cells were stained with PE-conjugated anti-TLR-4 antibody. Cell surface or intracellular TLR-4-positive cells were analyzed by a fluorescence activated cells sorter (BD FACS Calibur, San Jose, CA, USA). PE-conjugated anti-rat isotype IgG was used as a negative control for the experiments.

RT-PCR was performed as described previously [7]. Briefly, NB-1 and SK-N-SH cells were seeded (1 × 10⁶/ml) in 35 mm dish for overnight, trypsinized and washed with PBS. RNA was extracted from the cells with RNeasy mini kit (Qiagen, Valencia, CA, USA). Semi-quantative RT-PCR was carried out by using access quick RT-PCR system (Promega, Madison, WI, USA). Primer Sequence for TLR4, Forward 5 TGGATACGTTTCCTTATAAG 3 and Reverse 5 GAAATGGAGGCACCCCTTC 3 [8], CD14, forward 5 CTGCAACTTCTCCGAACCTC 3 and reverse 5 TAGGTCCTCGAGCGTCAGTT 5, MD2, forward 5 TTCCACCCTGTTTTCTTCCA 3 and reverse 5 AATCGTCATCAGATCCTCGC 3, MyD88, forward 5 GCACATGGGCACATACAGAC 3 and reverse 5 TGGGTCCTTTCCAGAGTTTG 3 [9]. GAPDH for TLR4 and G3PDH for CD14, MD2, MyD88 was used as an equal loading control. Optimized reverse transcription and PCR conditions were as follows, TLR4 (48°C for 45 min followed by 95°C for 2 min and 30 cycles at 95°C 30 sec, 56°C for 30 sec, 72°C for 60 sec), CD14, MD2 and MyD88 (48°C for 45 min followed by 95°C for 2 min and 28 cycles at 95°C 30 sec, 60°C 30 sec, 72°C for 1 min). For each experiment RNA extracted from human monocyte U937 was used as a positive control for TLR4, CD14, MD2 and MyD88. PCR products were analyzed by electrophoresis on 2% agarose gel. The gels were stained with CYBR safe DNA gel stain (Molecular probe, Eugene, OR, USA) and visualized under an ultraviolet transilluminator. The 100 bp DNA size marker (Invitrogen, Carlsbad, CA, USA) was also run to determine approximate size of the product.

Immunoblotting was performed as described previously [7]. To differentiate monocyte to machrophages, U937 cells were treated with PMA for 24 hours, washed and incubated for another 24 hours before treated with LPS as indicated below. NB-1 cells were directly seeded in 35 mm culture dish, treated with LPS (100 ng/ml) for various time. The cell lysates were extracted by lysis buffer containing 0.5 M Tris-HCL, 4% SDS and 2 mercaptoethanol, and boiled at 80°C for 5 min. The protein concentration of each sample was determined by BCA protein assay reagent (Pierce, Rockford, IL, USA). An equal amount of protein (40 μg) was analyzed by SDS-polyacrylamide gel electrophoresis under reducing conditions and transferred to a membrane filter. The membranes were treated with an appropriately diluted antibody for overnight. The immune complexes were detected with a 1:5000 dilution of horseradish peroxidase-conjugated protein G for 1 h and the bands were visualized with a chemiluminesent reagent (Pierce, Rockford, IL). The chemiluminescence was analyzed by a light capture system (AE6955, Atto Corp., Tokyo, Japan) with CS grab, version 1.0. Densimetric analysis of each band was analyzed with CS analyzer software, version 1.0 (Atto Corp., Tokyo, Japan).

The luciferase reporter gene assay was performed as described previously [10]. Briefly, NB-1 cells (3 × 10⁵/ml) were plated in a 35 mm plastic dish. On the following day, the cells were transfected with 0.5 mg/ml of pNF-κB-TA-luc luciferase reporter gene (Invitrogen, Carlsbad, CA, USA) and 0.05 mg of pRL-TK plasmid (Promega, Madison, WI, USA) by lipofectamine 2000 transfection reagent (Gibco-BRL). The transfected cells were incubated for 48 h, stimulated with LPS (100 ng/ml) and IL-1β (100 ng/ml) for 6 h, lysed with a lysis reagent from Promega, and the luciferase activity was determined by the dual luciferase assay kit (Promega). To confirm the activation of NF-κB, we used another plasmid pNiFty2-Luc from Invivogen, San-Diego, CA, USA.

Statistical significance was determined by Student's t test. Experimental result is expressed as the mean of triplicates ± SD in at least 3 independent experiments.

By using RT-PCR, we investigated whether human neuroblastoma cell line NB-1 expresses TLR4 or not. NB-1 cells were also treated with or without LPS (100 ng/ml) for 4 hr. The expression of TLR4 was examined with RT-PCR using extacted RNA. As shown in Fig. 1, NB-1 cells definitely express TLR4 mRNA (Fig 1A, lane 2). LPS treatment did not affect the expression of TLR4 (Fig 1A, Lane 3). To exclude the exceptional expression of TLR4 in NB-1 cells, we examined TLR4 expression in another human neuroblastoma cell line SK-N-SH. RT-PCR analysis demonstrated that SK-N-SH cell line as well as NB-1 cell line expressed TLR4 mRNA (Fig. 1B), suggesting that two neuroblastoma cell lines might commonly express TLR4.

In mammalian cells, TLRs are expressed either on cell surface or in the cytoplasm i.e golgi apparatus [11]. For example, TLR1, TLR2 and TLR4 are located on cell surface and are recruited to phagosomes after activation by their respective ligands whereas TLR3, TLR7 and TLR9 which recognize nucleic acid-like structures are located in subcellular compartment [12‐16]. First, we investigated the cell surface expression of TLR4 on NB-1 cells with flow cytometric analysis. NB-1 cells were stained with either PE-conjugated anti-human TLR4 antibody or PE-conjugated isotype matched IgG. Positively stained cells were not detected in NB-1 cells (Fig 2A), suggesting no cell surface expression of TLR4. In addition, U937 cells as positive control were strongly positive for TLR4. Second, we examined the intracellular expression of TLR4 using permeabilized NB-1 cells because RT-PCR analysis demonstrated the presence of TLR4 mRNA. The histogram of permeabilized NB-1 cells stained by anti-TLR4 antibody shifted to the right side (Fig. 2B), suggesting the intracellular localization of TLR4.

There are several key molecules leading to TLR4-dependent cellular activation by LPS. Besides TLR4, a series of molecules, such as CD14, MD2 and MyD88 are essential for LPS response [17, 18]. First, we studied the expression of CD14, MD2 and MyD88 mRNA on NB-1 cells. RNA was extracted from NB-1 cells treated with or without LPS (100 ng/ml) for 4 hr and used to RT-PCR for CD14, MD2 and MyD88. RT-PCR analysis demonstrated that mRNAs of the three molecules are expressed in NB-1 cells (Fig. 3). In addition, LPS treatment did not alter their expression in NB-1 cells. Human monocytic cell line U937 as a positive control was also positive for CD14, MD2 and MyD88 mRNA. Laser flow cytometric analysis demonstrated cell surface expression of CD14 in NB-1 cells (Data not shown).

It has been well documented that NF-κB associated with IκB resides in cytoplasm as an inactive form. On the process of NF-κB activation, IκBα is rapidly phosphorylated and degraded and then free NF-κB is translocated to nucleus [19]. In order to investigate whether LPS is able to activate NF-κB via intracellular TLR4, we examined degradation of IκB-α in NB-1 cells treated with LPS (100 ng/ml) for various time. As shown in Fig. 4, the expression of IκB-α partially decreased up to 30 min after LPS treatment and it returned to the original state 45 and 60 min after LPS treatment. The possibility that a small amount of contaminants in the LPS preparation might induce the partial activation still remained.

Based on partial degradation of IκB-α upon LPS treatment, we studied the activation of NF-κB by NF-κB-dependent luciferase assay. NB-1 cells transfected with the reporter gene were treated with LPS at various concentrations for 6 h. Human IL-1β as a positive inducer [20] was used at 100 ng/ml for 6 hr and the NF-κB activity was determined with a relative luciferase activity. LPS at 5 μg/ml failed to activate NF-κB activity in NB-1 cells whereas IL-1β markedly enhanced it (Fig. 5). Further, we tried another NF-κB-inducible reporter plasmid pNiFty2-Luc to detect NF-κB activation. However, LPS failed to activate NF-κB although IL-1β activated it significantly. Further, treatment with LPS for 24 h did not induce NF-kB-dependent TNF-α production in NB-1 cells (data not shown).

Among all TLRs, only TLR3 and TLR4 can induce cellular activation in both MyD88 dependent and independent pathway [21]. In the MyD88-independent pathway, LPS treatment leads to activation of transcription factor IRF-3, followed by production of IFN-β. Subsequently, IFN-β activates STAT1, leading to the induction of several IFN-inducible genes [22‐24]. We examined a possibility that NB-1 cells might have a defect in MyD88-independent pathway, especially IRF-3. Immunoblotting analysis demonstrated that LPS induces phosphorylation of IRF-3 in U937 but not in NB-1 cells. (Fig. 6A). Further, we examined the expression of IRF-3 in NB-1 cells and U937 cells. Immunoblotting analysis demonstrated that NB-1 cells did not express IRF-3 (Fig. 6B) although it was definitely expressed in U937 cells.

LPS is known to activate JNK1/2 and p38 MAP kinases through TAK1, a downstream molecule of TLR4-mediated signaling pathway [25]. We examined the activation of JNK1/2 and p38 MAP kinases in LPS-treated NB-1 cells. LPS did not activate JNK1/2 and p38 MAP kinases in NB-1 cells until 60 min (Fig. 7) whereas LPS activated them at 30 min in U937 cells. In addition, the treatment with LPS for 2 h also failed to activate p38 and JNK1/2 in NB-1 cells (data not shown).