Modulation of SOCS protein expression influences the interferon responsiveness of human melanoma cells

Recombinant human (hu) IFN-α2b (specific activity of 2 × 10⁸ IU/mg) was purchased from Schering-Plough, Inc. (Kenilworth, NJ). The HT144, Hs294T, SK-MEL-5 and A375 human melanoma cell lines were purchased from the American Type Culture Collection (Manassass, VA). The 1106 MEL, 18105 MEL, MEL-39, 1174 MEL, FO1 and 1259 MEL human melanoma cell lines were obtained from Dr. Soldano Ferrone (University of Pittsburgh Cancer Institute, Pittsburgh, PA). The radial growth phase WM 1552c and vertical growth phase WM 793b human melanoma cell lines were provided by Dr. M. Herlyn (Wistar Institute, Philadelphia, PA) and cultured as described [26]. Human embryonic melanocytes were purchased from ScienCell Research Laboratories (Carlsbad, CA) and cultured per manufacturer's recommendations. All studies were conducted with the approval of The Ohio State University Institutional Biosafety Committee.

Expression of SOCS1 and SOCS3 proteins in melanoma cell lines was confirmed via immunoblot analysis with antibodies directed against SOCS1, SOCS3 (Abcam, Cambridge, UK), phosphorylated STAT3 (Tyr705), STAT3 (Cell Signaling Technology, Inc., Danvers, MA), or β-actin as previously described [27].

Phosphorylation of STAT1 at Tyr⁷⁰¹ was measured using an intracellular flow cytometric assay as previously described; with modifications [28]. To measure STAT1 phosphorylation in retroviral-transduced cells, a goat anti-rabbit APC-conjugated secondary Ab was employed (Molecular Probes, Eugene, OR). Amplified fluorescence signals were expressed as specific fluorescence (Fsp = F_t -F_b), where F_t represents the median value of total staining, and F_b represents the median value of background staining (obtained by staining with the isotype control Ab) [28].

Real-time reverse transcription-polymerase chain reaction (RT-PCR) was used to quantitate transcript levels of the IFN-α stimulated genes ISG-15 and IFIT2 or IFN-γ-stimulated gene, IRF1 in melanoma cell cultures as previously described [28, 29]. Briefly, melanoma cells were treated with IFN-α (10⁴ U/mL), IFN-γ (1 ng/mL) or PBS for 4 hours. Total RNA was then isolated using the RNeasy RNA Isolation Kit (Qiagen, Valencia, CA), quantitated, and converted to cDNA using random hexamers as primers for first-strand synthesis. Two microliters of the resulting cDNA was used as a template to measure the levels of mRNA for ISG expression by Real-time RT-PCR with pre-designed primer/probe sets (Assays On Demand; Applied Biosystems, Foster City, CA) and 2× TaqMan Universal PCR Master Mix (Applied Biosystems) per the manufacturer's recommendations. Pre-designed primer/probe sets for human β-actin were used as an internal control in each reaction well (Applied Biosystems) and data was analyzed using Sequence Detector software (version 1.6; Applied Biosystems).

The PINCO retroviral transfer plasmid, originally from the laboratory of Dr. P.G. Pelicci (Grignani et al., 1998), was obtained through the courtesy of Dr. Martin Sattler (Dana Farber Cancer Institute, Boston, MA). This retroviral vector permits the expression of a gene of interest from the 5' long terminal repeat (LTR) as well as the enhanced green fluorescent protein (EGFP) from an internal cytomegalovirus (CMV) immediate early promoter. The development of PINCO-based retroviral constructs that over-express SOCS1 and SOCS3 proteins has been previously described by our group [30, 31]. SOCS1-, and SOCS3-overexpressing retroviral supernatants were generated by transient transfection of the Phoneix-Ampho packaging cell line with appropriate PINCO-SOCS expression plasmid DNA and 0.9 μg of a plasmid DNA expressing the VSV-G protein (pVSV-G) using the ProFection^® Mammalian Transfection System-Calcium Phosphate kit according to manufacturer's instructions (Promega, Madison, WI) as previously described [31]. Human melanoma cell lines (HT144, 1259 MEL) were transduced with the resulting SOCS-over-expressing retroviral constructs as previously described [31].

Silencing of SOCS1 and SOCS3 expression was accomplished using pSilencer siRNA expression vectors (Invitrogen). siRNA oligonucleotides that target SOCS1 (5'-CTGGTTGTTGTAGCAGCTTAA-3') and SOCS3 (5'-TCGGGAGTTCCTGGACCAGTA-3') sequences were cloned into the pSilencer vector per manufacturer's recommendations. Briefly, 1259 MEL melanoma cells were transfected with 2 μg siRNA encoding plasmids or pSilencer vector alone (negative control) via electroporation using the Amaxa Nucelofector device and cell-specific nucleofector reagent (Amaxa, Inc., Gaithersburg, MD) according to the manufacturer's recommendations.

Mixed effects regression models were used to analyze the experimental outcomes, including a random effect for each experiment. The outcomes of specific fluorescence and fold change in gene expression were both log-transformed to meet the necessary model assumptions of normality and constant variance. A Tukey-Kramer adjustment for multiple comparisons was used to calculate the p-values when comparing groups within the same experiment [32]. Adjusted p-values less than 0.05 were considered statistically significant.

The baseline expression of SOCS1 and SOCS3 protein was measured in a panel of n = 10 human metastatic melanoma cell lines by immunoblot analysis. These data revealed that all lines expressed both SOCS1 and SOCS3 proteins (Figure 1a). Of note, pre-incubation of primary antibody with commercially available SOCS-specific peptide competitors (Abcam, Inc.) eliminated SOCS3 specific immunoreactivity (Data not shown). Immunoblot analysis in representative cell lines was subsequently used to evaluate the effects of IFN-α and IFN-γ stimulation on SOCS1 and SOCS3 protein expression in vitro (Figure 1b). Immunoblot analysis revealed an induction of SOCS1 and SOCS3 proteins in some melanoma cell lines following a four hour stimulation with IFN-α (10⁴ U/mL) or IFN-γ (10 ng/mL). The IFN-induced expression of SOCS1 and SOCS3 proteins within individual cell lines was variable. For example, IFN-α and IFN-γ induced SOCS3 protein expression in the A375 melanoma cell line while both SOCS1 and SOCS3 proteins were upregulated in the HT144 cell line (Figure 1b). In contrast, neither SOCS1 nor SOCS3 proteins were further upregulated in the 1259 MEL cell line following IFN stimulation. To determine whether SOCS protein expression might be restricted to melanoma cells at a particular stage of development, we conducted similar experiments in human embryonic melanocytes (HEM) and in the WM793b and WM1552c human melanoma cell lines. These two cell lines are derived from vertical and radial growth phase melanoma cells, respectively. SOCS1 and SOCS3 were expressed at basal levels in each of these cell types (Figure 1c). Treatment with IFN-α or IFN-γ did not upregulate SOCS1 expression in any of these cell lines, while SOCS3 was upregulated by IFN-γ only in the WM1552c radial growth phase cell line.

The STAT3 transcription factor promotes a metastatic phenotype and has been shown to be constitutively phosphorylated in human melanoma cell lines. Importantly, STAT3 is also sensitive to inhibition by SOCS proteins [33] and prior studies in melanoma brain metastases have suggested that loss of SOCS1 expression could promote metastasis via elevated STAT3 signaling. We therefore analyzed pSTAT3 levels in order to determine whether SOCS1 protein expression and basal STAT3 phosphorylation were associated in the panel of melanoma cell lines. These data indicated that basal STAT3 phosphorylation (at Tyr⁷⁰⁵) and SOCS1 expression were present concurrently (Figure 1d). These data suggest that the presence of SOCS1 does not alter the level of basal pSTAT3.

To further validate the role of SOCS protein expression in regulating IFN-responsiveness of representative human melanoma cell lines (HT144, 1259 MEL), we employed retroviral constructs that expressed SOCS1 and SOCS3. Non-transduced cells and cells transduced with the empty PINCO retroviral vector were used as controls in these assays. Twenty-four hours following transduction of human melanoma cell lines with the SOCS1 and SOCS3-expressing retroviral constructs or empty PINCO vector, the expression of EGFP was confirmed in each cell line by flow cytometry (Figure 2a). EGFP⁺ cells were sorted by flow cytometry and cultured for use in subsequent experiments. Transduced melanoma cell lines over-expressed SOCS1 and SOCS3 proteins as determined by immunoblot analysis (Figure 2b-c). No increase in SOCS1 or SOCS3 expression was observed in any melanoma cell line transduced with the empty PINCO vector. Analysis of STAT1 phosphorylation by flow cytometry indicated that IFN-α-induced signal transduction was significantly inhibited in HT144 and 1259 MEL cells that over-expressed SOCS1 as compared to either parental cells or those transduced with the empty PINCO vector (all p-values = 0.019). Over-expression of SOCS3 led to a significant reduction in IFN-α-induced signal transduction in the 1259 MEL cell line (p = 0.0052), but this did not reach statistical significance in the HT144 cell line (Figure 3a-b; p = 0.087). Similar results were obtained for IFN-γ-induced activation of STAT1 (Figure 3c-d). IFN-α and IFN-γ responsiveness was most inhibited by over-expression of SOCS1 as compared to SOCS3. As expected, SOCS1 over-expression significantly reduced the transcription of IFN-α regulated genes (ISG-15 and OAS1) and the IFN-γ induced gene IRF1 in these cell lines (Figure 4a-d; all p-values ≤ 0.01). Similar to our signal transduction data, over-expression of SOCS3 in the 1259 MEL cell line produced a statistically significant inhibition of IFN-stimulated gene expression (all p-values ≤ 0.0049) but results for HT144 were mixed. A significant difference was observed for expression of ISG-15 in the HT144 cell line (p = 0.0144), but not for OAS-1 (p = 0.0511) or IRF1 (p = 0.49).

The interferon-responsiveness of SOCS1 and SOCS3-positive 1259 MEL melanoma cells transfected with a vector expressing siRNA constructs targeting SOCS1 or SOCS3 was next investigated. Cells transfected with an empty vector served as negative controls in these studies. Transfection efficiency was typically > 90% as determined by transfection of cells with a GFP-expressing plasmid in parallel (data not shown). Reduced levels of SOCS1 and SOCS3 transcript were confirmed by Real Time PCR (Figure 5a-b). We proceeded to test whether the reduced level of SOCS expression would affect the IFN-responsiveness of melanoma cells. Due to high basal expression, complete knockdown of SOCS1 and SOCS3 in these cells was difficult to achieve. However, their reduced expression following transfection of 1259 MEL cells with SOCS1 or SOCS3 siRNA led to consistent increases in IFN-α or IFN-γ-induced P-STAT1 as compared to cells transfected with control siRNA (Figure 5c). Transcription of IFN-α-induced genes (IFIT2, ISG-15) was tested by Real Time PCR and found to be significantly increased in cells transfected with SOCS1siRNA or SOCS3siRNA (Figure 5d; all p-values < 0.005).