Inhibition of STAT 1 Phosphorylation by Human Parainfluenza Virus Type 3 C Protein

Abstract

The P mRNA of the viruses belonging to the subfamily Paramyxovirinae possesses a unique property of giving rise to several accessory proteins by a process that involves the utilization of overlapping open reading frames (the C proteins) and by an “RNA-editing” mechanism (the V proteins). Although these proteins are considered accessory, numerous studies have highlighted the importance of these proteins in virus transcription and interferon signaling, including our previous observation on the role of human parainfluenza virus type 3 (HPIV 3) C protein in the transcription of viral genome (Malur et al., Virus Res. 99:199-204, 2004). In this report, we have addressed its role in interferon signaling by generating a stable cell line, L-C6, by using the lentiviral expression system which expresses HPIV 3 C protein. The L-C6 cells were efficient in abrogating both alpha and gamma interferon-induced antiviral states and demonstrated a drastic reduction in the formation of gamma-activated factor complexes in the cell extracts. Western blot analysis subsequently revealed a defect in the phosphorylation of STAT 1 in these cells. Taken together, our results indicate that HPIV 3 C protein is capable of counteracting the interferon signaling pathway by specifically inhibiting the activation of STAT 1.

The subfamily Paramyxovirinae consists of three genera, namely, Paramyxovirus, Morbillivirus, and Rubulavirus, and all viruses belonging to this subfamily are characterized by the presence of a versatile P mRNA, which not only gives rise to the phosphoprotein (P), an essential component of the polymerase complex, but also C proteins from separate overlapping open reading frames (ORFs). Moreover, by a unique “RNA-editing” reaction involving site-specific insertion of pseudotemplated residues within the P gene, additional proteins termed D, V, and W are synthesized (2, 16). The numbers of C proteins expressed from each of the P mRNA of the members of Paramyxovirinae are different; e.g., four C proteins, C′, C, Y1, and Y2, are found in Sendai virus (SeV), two are found in human parainfluenza virus type 1 (HPIV 1), and one C protein is found in human parainfluenza virus type 3 (HPIV 3) and measles virus, while rubulaviruses do not express any C protein. All three genera express a V protein from an edited RNA, with the exception of HPIV 1 (2, 16).

Several studies pertaining to the expression of these proteins or development of mutant viruses have demonstrated that the Paramyxovirinae C and V proteins are involved in viral replication (3, 4, 5, 6, 9, 10, 11, 13, 15, 17, 20, 21, 23, 25, 27). Moreover, in the cases of SeV, measles virus, simian virus 5, and HPIV 2, these proteins have been shown to be capable of counteracting the interferon (IFN) signaling pathway by using a variety of mechanisms (for reviews, see references 1 and 8). Apart from the importance of these proteins in viral replication (4), the role of the HPIV 3 C, D, and V proteins in interferon signaling is currently unknown.

In the case of HPIV 3, a single C protein 199 amino acids in length, is synthesized by the P mRNA from an alternate ORF, in addition to another protein, P-D, that is synthesized as a result of the RNA-editing mechanism, while the predicted synthesis of the V protein remains unconfirmed (4). A recombinant HPIV 3 virus devoid of C ORF (rC-KO), isolated by using a reverse genetic approach, displayed attenuated properties both in vitro and in vivo. On the other hand, in vitro and in vivo replication of two other recombinant viruses individually lacking D and V ORFs (rD-KO and rV-KO, respectively) remained unaffected. However, a double mutant virus (rDV-KO) was attenuated in vivo (4).

Recent studies from our laboratory demonstrated that the C protein was capable of inhibiting HPIV 3 minigenome transcription in a dose-dependent manner, and a similar inhibitory effect by using the heterologous SeV C protein was observed. By computational analysis, we uncovered the presence of a coiled-coil motif within the HPIV 3 C protein, and the presence of such a motif in other paramyxovirus C proteins was confirmed. Subsequently, the role of this motif in HPIV 3 minigenome transcription was verified when a mutant abrogated the inhibitory effect of C protein (19).

In a study aimed toward understanding the mechanism by which HPIV 3 counteracted the interferon signaling pathway, Young et al. (28), using a reporter assay, demonstrated that HPIV 3 blocked both alpha IFN (IFN-α) signaling and IFN-γ signaling. However, HPIV 3 inhibited induction of alpha IFN-stimulated gene factor 3 complex, whereas gamma-activated factor (GAF) complexes mediated by gamma IFN were detected in HPIV 3-infected cells. Furthermore, no changes in the overall levels of STAT 1 were observed, although a reduction in the levels of phosphoserine forms of STAT 1 was seen to be consistent with the idea that HPIV 3 blocked interferon signaling by possibly interfering with some STAT 1-specific function (28).

In order to understand the molecular mechanism leading to the inhibition of interferon signaling by HPIV 3 and ascertain the possible involvement of the C protein in these processes, we generated a cell line, L-C6, that stably expresses the HPIV 3 C protein. Here, we demonstrate directly that the C protein is capable of abrogating the interferon-induced antiviral state by inhibiting activation of STAT 1.

Stable expression of the C protein.

A previously modified lentiviral expression vector (LRV), with a blasticidin resistance marker (Fig. 1A), was obtained from the Virus Core Facility. The BamHI-ApaI fragment containing the C ORF with a FLAG epitope towards the 3′ end was subcloned into the corresponding sites of the lentiviral vector to obtain the recombinant C plasmid, which was then transfected into 293T cells, along with the plasmids possessing the gag, pol, and rev genes and a plasmid encoding the vesicular stomatitis virus (VSV) glycoprotein. The recombinant C lentivirus thus obtained was subsequently used to transduce HeLa cells, and 3 days after infection, blasticidin (10 μg/ml) was added to the culture medium. Colonies grown in the blasticidin selection medium were pooled together and expanded. The presence of the C protein in these cells was verified by the labeling of a portion of these cells with 50 μCi of [³⁵S]methionine followed by immunoprecipitation with the FLAG antibody conjugated to agarose beads (18), and bound proteins were separated on a 12% sodium dodecyl sulfate (SDS)-polyacrylamide gel, followed by autoradiography. As can be seen from Fig. 1B, lane 2, a band corresponding to the size of the C protein of about 24 kDa was clearly visible compared to LRV used as empty vector (Fig. 1B, lane 1). Subsequently, the C-expressing cells were plated onto a 96-well plate by use of the terminal dilution technique and grown in the selection medium. Cell extracts from individual clones were subjected to SDS-polyacrylamide gel electrophoresis (PAGE) and Western blot analysis using anti-FLAG antibody and polyclonal anti-C antibody, and a representative Western blot of cell line L-C6, expressing the C protein by using the anti-FLAG antibody, is shown in Fig. 1C. The molecular mass of the C protein (24 kDa) was in accordance with a band of similar size from another C-expressing plasmid that was used as a positive control. The anti-FLAG antibody was found to be superior to the polyclonal anti-C antibody in Western blot experiments (data not shown). The intracellular localization of the C protein in L-C6 cells was confirmed by immunofluorescence studies, following fixation with 3% paraformaldehyde and permeabilization with phosphate-buffered saline (PBS) containing 0.2% Triton X-100. After being blocked, cells were incubated in blocking buffer (3% normal goat serum, 0.1% Triton X-100 in PBS) containing either polyclonal anti-C or anti-FLAG antibodies, followed by incubation with corresponding fluorescein isothiocyanate-conjugated secondary antibodies. Cells were viewed under a fluorescence microscope. As seen in Fig. 1D, the staining of L-C6 cells with polyclonal anti-C antibody revealed an even distribution of the C protein in the cytoplasm, while a pronounced punctate pattern restricted mainly near the perinuclear area of the nucleus as seen with the anti-FLAG antibody (Fig. 1D). Thus, the L-C6 cells expressed the C protein predominantly in the cytoplasmic fraction.

FIG. 1.

Expression of C protein. (A) Schematic representation of lentiviral expression vector LRV with the C ORF cloned between the BamHI and ApaI restriction sites. SIN-LTR, self-inactivating long terminal repeat; RRE, Rev-responsive element; CMV, cytomegalovirus promoter; MCS, multiple cloning site; SV40, simian virus 40 promoter; SD, splice donor; SA, splice acceptor; BLSD, blasticidin. (B) Immunoprecipitation of C protein from [³⁵S]methionine-labeled cell lysates was performed using an anti-FLAG antibody bound to agarose beads followed by SDS-PAGE and autoradiography. The arrow denotes the band corresponding to the C protein. (C) Western blot analysis of L-C6 cells expressing the C protein. HeLa and L-C6 cell lysates were subjected to SDS-PAGE and Western blot analysis, and C protein was detected by use of an anti-FLAG monoclonal antibody followed by a secondary antibody conjugated to horseradish peroxidase. The blot was developed by an ECL detection system. A band corresponding to the C protein is shown by an arrow. (D) Immunofluorescence analysis of C protein. HeLa and L-C6 cells grown on coverslips were fixed and permeabilized. Following the incubation of cells with either polyclonal anti-C or anti-FLAG antibody and the corresponding fluorescein isothiocyanate-conjugated secondary antibodies, cells were viewed under a fluorescence microscope, and images were processed using the ImagePro software.

FIG. 2.

Interferon-induced antiviral states of parental and C-expressing cell lines. (A) HeLa and L-C6 cells were not treated or treated with various concentrations of IFN-γ (U/ml) for 16 h followed by mock infection (mock) or infection with VSV (0.025 MOI). Cells surviving VSV challenge were subsequently fixed and stained with crystal violet. (B) The replication of VSV from panel A was analyzed by SDS-PAGE and Western blot analysis using a VSV anti-N polyclonal antibody. The arrow denotes the position of the N protein. (C) HeLa and L-C6 cells were treated with various concentrations of IFN-α (U/ml) for 16 h, followed by infection with VSV (0.025 MOI). Cells surviving VSV challenge were subsequently fixed and stained with crystal violet.

HPIV 3 C protein abrogates IFN-induced antiviral state.

In order to determine the effect of the C protein on the IFN-induced antiviral state, HeLa and L-C6 cells were either not treated or treated with various concentrations of IFN-γ or IFN-α for 16 h in serum-free media. Cells were washed with PBS and mock infected or infected with VSV (multiplicity of infection [MOI], 0.025) for 1 h, followed by incubation in serum-free medium for 16 h. The cells were subsequently fixed and stained with crystal violet and observed for the cytopathic effect. As shown in Fig. 2A, treatment of HeLa and L-C6 cells with IFN-γ alone did not cause cytopathic effect. All cells that were not treated were detached from the plates after infection with VSV. HeLa cells treated with 100 U and 1,000 U/ml of IFN-γ were protected from the cytopathic effect induced by VSV. In contrast, L-C6 cells infected with VSV were totally detached from the plates after pretreatment with 1,000 U/ml of IFN-γ (Fig. 2A). The replication efficiency of VSV in these cells, whether nontreated or pretreated with IFN-γ, was measured by monitoring the intracellular synthesis of VSV N protein in an SDS-PAGE/Western blot analysis using a polyclonal anti-VSV N antibody. As seen in Fig. 2B, intracellular synthesis of N protein in the nontreated HeLa cells (lane 2) and that in L-C6 cells (lane 7) were similar but partially suppressed upon pretreatment of HeLa cells with 1,000 U/ml of IFN-γ (compare lanes 2 and 5). In contrast, synthesis of N protein, i.e., VSV replication in L-C6 cells, remained unchanged after pretreatment with 1,000 U/ml of IFN-γ (Fig. 2B, compare lanes 7 and 10). A similar result was obtained upon treatment of HeLa and L-C6 cells with IFN-α and subsequent infection with VSV (Fig. 2C). The data obtained from the cytopathological experiments with cultured cells together with Western blot analysis clearly suggested that the C protein was capable of counteracting the IFN-induced antiviral state. Further experiments involving electrophoretic mobility shift assay (EMSA) were undertaken in order to ascertain the role of C protein in interferon signaling.

Reduction in GAF complex formation is mediated by C protein.

Using EMSA studies, Young et al. (28) previously demonstrated the presence of GAF complexes upon induction with IFN-γ. Moreover, the presence of STAT 1 in the interferon-induced GAF complexes was confirmed by using a STAT 1 antibody in a supershift analysis. These observations prompted us to examine the role of C protein on the formation of interferon-inducible GAF complexes. Accordingly, cytoplasmic extracts were prepared from nontreated or IFN-γ-treated HeLa, and L-C6 cells as described earlier. Cell extracts were incubated with 50,000 cpm of [γ-³²P]ATP-labeled hSIEm67 probe (26) in EMSA buffer (10 mM Tris, pH 7.5, 5% glycerol, 1 mM dithiothreitol, 0.5 mM EDTA) containing 2 μg of poly(dI-dC) for 20 min at room temperature, and the complexes were resolved on 4% polyacrylamide gel. As shown in Fig. 3A, a band corresponding to the slower migrating product predicted to be GAF in parental HeLa cells stimulated with IFN-γ was apparent (lane 2). This band was absent in nontreated HeLa cells (Fig. 3A, lane 1) or when the cell extracts were incubated with an excess of the same nonlabeled hSIEm67 oligonucleotide (Fig. 3A, lane 3). In contrast, a significant reduction in GAF complexes in L-C6 cells pretreated with IFN-γ was seen (Fig. 3A, lane 5). PhosphorImager analysis of band intensities demonstrated a 56% reduction in GAF complexes in L-C6 cells (Fig. 3A, lane 5) compared to those in HeLa cells (Fig. 3A, lane 2). These results indicated that the C protein was capable of inhibiting the formation of GAF complexes. The specific identities of the protein constituents in the GAF complex were further studied by supershift analysis using specific antibodies. Accordingly, cytoplasmic extracts from nontreated and IFN-γ-treated HeLa and L-C6 cells were incubated with antibodies specific to STAT 1 and STAT 3 prior to the addition of [γ-³²P]ATP-labeled hSIEm67 probe as mentioned above, and complexes were resolved on 4% polyacrylamide gel. As shown in Fig. 3B, IFN-γ treatment of HeLa cells led to the formation of a major GAF-specific complex as expected (lane 2), and a similar product was also formed in L-C6 cells (lane 8), albeit at significantly reduced levels (Fig. 3A, lane 5). Incubation of HeLa and L-C6 cell extracts with anti-STAT 1 antibody (Fig. 3B, lanes 5 and 11) revealed the presence of a supershifted band, indicating that the GAF complex was primarily composed of a STAT 1 homodimer. However, in the case of L-C6 cells, formation of this STAT 1 homodimer (Fig. 3B, lane 5) was significantly reduced compared with that of HeLa cells (Fig. 3B, lane 11). Anti-STAT 3 antibody, on the other hand, had no effect on the GAF complex in either HeLa (Fig. 3B, lane 6) or L-C6 cells (Fig. 3B, lane 12), indicating that STAT 3 was not involved in GAF complex formation.

FIG. 3.

(A) Formation of GAF complexes in IFN-stimulated parental and L-C6 cell lines. Results for HeLa and L-C6 cells not treated (−) or treated (+) with IFN-γ are shown. Total cell lysates were analyzed by EMSA using a γ-³²P-labeled hSIEm67 probe. The formation of GAF-specific transcription complexes was verified by incubating the IFN-treated cells with an excess nonradiolabeled hSIEm67 oligonucleotide. The arrow indicates the position of GAF complex. (B) Supershift analysis of GAF complexes, using STAT antibodies. IFN-γ-treated (+) or nontreated (−) HeLa and L-C6 cell extracts were incubated with STAT 1 and STAT 3 antibodies prior to incubation with γ-³²P-labeled hSIEm67 probe and GAF complexes were analyzed by EMSA as mentioned for Fig. 3A. The arrow indicates the presence of the supershifted (S.S) GAF complexes. IgG, immunoglobulin G.

Loss of pY-STAT 1 in L-C6 cells.

To gain direct evidence for the mechanism leading to the reduction in GAF complex formation, we determined the levels of phosphorylation of STAT 1 in L-C6 cells. Accordingly, cell extracts from HeLa and L-C6 cells nontreated or treated with IFN-γ (1,000 U/ml) for the indicated times were probed with antibodies to phosphorylated STAT 1 (pY-STAT 1) and STAT 1. As can be seen from Fig. 4A (top panel), the treatment of HeLa cells with IFN-γ led to a rapid and robust phosphorylation of STAT 1 starting at 15 min (lane 3) and continuing until 60 min (lane 7). In contrast, phosphorylation of STAT 1 in L-C6 cells was totally abrogated (lanes 4 and 6), and no pY-STAT 1 band could be detected, even after 60 min of treatment with IFN-γ (lane 8). Moreover, the pY-STAT 1 bands from IFN-γ-treated HeLa cell extracts were specific as no bands could be detected from nontreated HeLa cell extracts (lane 1). However, the unphosphorylated STAT 1 levels (as assayed with STAT 1 antibody) for both HeLa and L-C6 cells remained unaffected (Fig. 4A, middle panel). These results clearly demonstrated that the C protein targeted phosphorylation of STAT 1, an initial and critical step of the interferon signaling pathway.

FIG. 4.

Inhibition of pY-STAT 1 levels in L-C6 cells. (A) Total cell extracts from HeLa and L-C6 cells not treated (−) or treated with IFN-γ (1,000 U/ml) at the indicated time intervals were subjected to SDS-PAGE followed by Western blot analysis with anti-pY-STAT 1 antibody (top), anti-STAT 1 antibody (middle), or antiactin antibody (bottom). (B) A similar experiment was performed as described for panel A, except that HeLa and L-C6 cells were treated with IFN-α.

Since STAT 1 is also a component of the IFN-α signaling pathway, we next determined whether the C protein was also capable of inhibiting the phosphorylation/activation processes. Accordingly, HeLa and L-C6 cells were either not treated or treated with IFN-α (1,000 U/ml) for the indicated times mentioned earlier and probed with pY-STAT 1 and STAT 1 antibodies. Western blot analysis demonstrated an efficient formation of pY-STAT 1 in HeLa cells at all time intervals (15, 30, and 60 min) (Fig. 4B, top panel). Moreover, the pY-STAT 1 levels were similar to the observed levels of pY-STAT 1 in HeLa cells treated with IFN-γ (Fig. 4A, top panel). Interestingly, similar results leading to the abrogation of STAT 1 phosphorylation in L-C6 cells treated with IFN-α were observed at all time intervals, although a weak band was discernible upon prolonged exposure to the film (Fig. 4B, top panel, lane 8). On the other hand, the levels of STAT 1 in all instances remained virtually the same (Fig. 4B, middle panel).

The mechanism by which HPIV 3 C protein inhibits phosphorylation of STAT 1 is currently under investigation. One possible mechanism would involve the formation of an aberrant C-STAT 1 complex, resulting in the disruption of phosphorylation/dephosphorylation processes of STAT 1, thereby preventing pY-STAT 1 from binding to gamma-activated sequence (GAS) elements, a process similar to the one observed for the SeV C protein (14, 22, 24). Alternatively, the C protein may interact with cytoplasmic protein tyrosine kinases (JAK 1, JAK 2, or TYK 2), leading to the impairment in the phosphorylation of STAT 1. Thus, it would be important to study whether the C protein directly interacts with the STAT 1 protein or the JAK kinases. Identification of this intermediary step would help in elucidating the role of C protein in interferon signaling. Furthermore, it would be of interest to investigate the critical domain(s) within the C protein that regulate STAT 1 function. One approach to address this issue would involve the generation of stable cell lines expressing mutant C proteins, as in the case of SeV C protein, where it was demonstrated that specific residues within the carboxy terminus of the smallest of the C proteins, Y2, were involved in IFN antagonism (12). Garcin et al. (7), in contrast, demonstrated that 23 amino acids from the amino terminus of the full-length SeV C protein were directly involved in inducing STAT 1 instability.

In our previous report, we highlighted the presence of a coiled-coil motif within the C protein and demonstrated its importance in minigenome transcription (19). Our efforts are currently directed towards the obtaining of a stable cell line for this mutant, in addition to several other mutant C proteins, with an aim to study the precise role of the C protein in interferon signaling.