Identification of NF-kappaB-dependent gene networks in respiratory syncytial virus-infected cells - PubMed (original) (raw)

Identification of NF-kappaB-dependent gene networks in respiratory syncytial virus-infected cells

Bing Tian et al. J Virol. 2002 Jul.

Abstract

Respiratory syncytial virus (RSV) is a mucosa-restricted virus that is a leading cause of epidemic respiratory tract infections in children. In epithelial cells, RSV replication activates nuclear translocation of the inducible transcription factor nuclear factor kappaB (NF-kappaB) through proteolysis of its cytoplasmic inhibitor, IkappaB. In spite of a putative role in mediating virus-inducible gene expression, the spectrum of NF-kappaB-dependent genes induced by RSV infection has not yet been determined. To address this, we developed a tightly regulated cell system expressing a nondegradable, epitope-tagged IkappaBalpha isoform (Flag-IkappaBalpha Mut) whose expression could be controlled by exogenous addition of nontoxic concentrations of doxycycline. Flag-IkappaBalpha Mut expression potently inhibited IkappaBalpha proteolysis, NF-kappaB binding, and NF-kappaB-dependent gene transcription in cells stimulated with the prototypical NF-kappaB-activating cytokine tumor necrosis factor alpha (TNF-alpha) and in response to RSV infection. High-density oligonucleotide microarrays were then used to profile constitutive and RSV-induced gene expression in the absence or presence of Flag-IkappaBalpha Mut. Comparison of these profiles revealed 380 genes whose expression was significantly changed by the dominant-negative NF-kappaB. Of these, 236 genes were constitutive (not RSV regulated), and surprisingly, only 144 genes were RSV regulated, representing numerically approximately 10% of the total population of RSV-inducible genes at this time point. Hierarchical clustering of the 144 RSV- and Flag-IkappaBalpha Mut-regulated genes identified two discrete gene clusters. The first group had high constitutive expression, and its expression levels fell in response to RSV infection. In this group, constitutive mRNA expression was increased by Flag-IkappaBalpha Mut expression, and the RSV-induced decrease in expression was partly inhibited. In the second group, constitutive expression was very low (or undetectable) and, after RSV infection, expression levels strongly increased. In this group, NF-kappaB was required for RSV-inducible expression because Flag-IkappaBalpha Mut expression blocked their induction by RSV. This latter cluster includes chemokines, transcriptional regulators, intracellular proteins regulating translation and proteolysis, and secreted proteins (complement components and growth factor regulators). These data suggest that NF-kappaB action induces global cellular responses after viral infection.

PubMed Disclaimer

Figures

FIG. 1.

Time-course of Flag-IκBα Mut expression after doxycycline deprivation. IκBα Mut-expressing cells isolated in the presence of doxycycline (2 μg/ml in growth medium) were transferred to doxycycline-free medium. At the indicated times, cytoplasmic extracts were prepared, and 150 μg was fractionated on SDS-10% PAGE and subjected to Western immunoblot by using anti-Flag M2 monoclonal antibody-peroxidase conjugate. After Flag detection, the immunoblot was reprobed for β-actin as a loading control.

FIG. 2.

Effect of doxycycline withdrawal on rTNF-α-induced IκBα proteolysis and NF-κB binding. (A) Time course of IκBα degradation after rTNF-α stimulation. Flag-IκBα Mut-expressing cells were cultured with or without doxycycline for 7 days. The cells were then treated without (lane 0) or with rTNF-α (25 ng/ml) for the indicated times (minutes). Top panel, cytoplasmic extracts of the cells were prepared and subjected to Western immunoblot with anti-IκBα antibody. ∗, migration of phospho-IκBα and Flag-IκBα Mut; ---, migration of endogenous IκBα. Bottom panel, similar protein loading was confirmed by stripping the blot and reprobing with anti-β-actin. (B) Analysis of IκBα isoforms in doxycycline-treated cells. Cytoplasmic lysates corresponding to lanes 1, 2, and 5 from panel A were fractionated and probed with anti-phospho-specific IκBα and anti-Flag alone and in combination with anti-IκBα. Although the Flag-IκBα Mut protein migrated slower than the native IκBα, it was clearly separable from phospho-IκBα and represents the slower migrating band observed in A. (C) Time course of NF-κB binding. Sucrose cushion-purified nuclear extracts were prepared from the same experiment and subjected to EMSA by using 32P-labeled APRE WT duplex oligonucleotide probe. The complexes were fractionated on 6% native polyacrylamide gels; an autoradiographic exposure is shown. The relative migration of the RelA/NF-κB1 (p50) heterodimer and the p50 homodimer (p50)2 complexes are indicated (34). (D) Competition assay in EMSA. Nuclear extracts from rTNF-α-stimulated cells (15 min, cultured in the presence of doxycycline) were subjected to competition analysis by the addition of a molar excess of nonradioactive double-stranded oligonucleotide competitor at the time of addition of radioactive probe. Competitors used were APRE WT and APRE M2 (Mut). (E) Antibody supershift in EMSA. Nuclear extracts from rTNF-α-stimulated cells (15 min, plus doxycycline) were mixed with 1 μl of affinity-purified polyclonal antibodies for 1 h at 4°C prior to EMSA analysis. PI, preimmune IgG. Bottom panel, brief autoradiographic exposure of the native NF-κB binding complexes. Top panel, longer exposure to detect the supershifted complexes (indicated by asterisk). Addition of either anti-RelA or anti-NF-κB1(p50) antibodies completely disrupted the RelA/NF-κB1(p50) complex.

FIG. 3.

Effect of Flag-IκBα Mut expression on RSV-induced IκBα proteolysis and NF-κB binding. (A) Western immunoblot for cytoplasmic IκBα levels. Flag-IκBα Mut-expressing cells were cultured in the presence or absence of doxycycline for 7 days. The cells were infected without (lane 0) or with purified RSV (multiplicity of infection of 1) for the indicated times (hours) prior to harvest. Cytoplasmic extracts of the cells were analyzed by Western immunoblot for IκBα by using affinity-purified anti-IκBα antibodies. ∗, migration of phospho-IκBα and Flag-IκBα Mut; ---, migration of endogenous IκBα. Bottom panel, similar protein loading was confirmed by stripping the blot and reprobing with anti-β-actin. (B) Time course of NF-κB binding. Sucrose cushion-purified nuclear extracts were subjected to EMSA as in Fig. 2B except twice as much nuclear protein was used to detect the relatively weaker induction of DNA binding. The positions of the RelA/NF-κB1 (p50) heterodimer and the p50 homodimer (p50)2 complexes are indicated (34). The RSV-inducible complexes have composition and binding specificity identical to those of the rTNF-α-inducible ones (34).

FIG. 4.

Effect of Flag-IκBα Mut expression on NF-κB-dependent transcription. (A) Flag-IκBα Mut-expressing cells were transiently transfected with (APREWT)3-p59rAT/LUC reporter vector and the alkaline phosphatase-internal control plasmid pSV2PAP in the absence or presence of doxycycline as indicated. Cells were cultured for an additional 40 h after transfection and stimulated with rTNF-α (25 ng/ml, 6 h). Normalized luciferase activity is presented for a representative transfection. The number in parentheses corresponds to the fold induction of reporter activity from the TNF stimulation relative to unstimulated cells. ∗∗, P < 0.001, Student's t test, compared to control, n = 3 independent experiments. (B) Experiment as in A in which (IL-6κBE)3-p59rAT/LUC reporter was used. (C) The cells were transfected with the native −162/+44 hIL-8/LUC. Experiment and data analysis are as in A. ∗, P < 0.05, Student's t test, compared to control.

FIG. 5.

Effect of Flag-IκBα Mut expression on RSV-inducible chemokine expression. Flag-IκBα Mut-expressing cells were cultured with or without doxycycline for 7 days. Cells were infected with purified RSV (multiplicity of infection, 1) for indicated times (in hours) prior to harvest of total cellular RNA. mRNA abundance was determined in each sample by multiprobe RNase protection assay. Shown is a representative autoradiographic exposure after denaturing gel electrophoresis. Left lane is undigested input probe (marker). Locations of protected fragments for RANTES, IP-10, IL-8, and the two housekeeping genes L32 and GAPDH are indicated at right. (B) Time course of normalized RANTES mRNA abundance (relative to GAPDH) was determined by exposure of the gel to a PhosphorImager cassette and plotted as a function of the duration of RSV infection (in hours). Errors are standard deviations from three independent experiments. ∗, P < 0.05, Student's t test, compared to no doxycycline treatment for each time point. (C) Time course of normalized IL-8 mRNA abundance (relative to GAPDH) was determined as in B.

FIG. 6.

Effect of Flag-IκBα Mut expression on RSV transcription. Flag-IκBα Mut-expressing cells were cultured with or without doxycycline for 7 days and infected with purified RSV (multiplicity of infection, 1) for the indicated times (in hours). Top, total cellular RNA was extracted, and abundance of RSV N transcript was determined by Northern blot (shown is a representative autoradiographic exposure). Bottom, the blot was reprobed with 18S RNA as an RNA recovery marker.

FIG. 7.

NF-κB-dependent gene networks identified by high-density microarrays. (A) Analysis of RSV- and doxycycline-regulated genes. Two-way ANOVA was used to analyze the data set from three independent array experiments corresponding to control (no doxycycline treatment), control (with doxycycline treatment), RSV 12-h infection (no doxycycline treatment), and RSV 12-h infection (with doxycycline treatment) and genes with a P value [Pr(F)] of <0.05 as a result of either treatment were compared. Shown is a Venn diagram of genes common to both datasets. (B) Clustering and heat map analysis of the RSV- and Flag-IκBα Mut-regulated data set. Agglomerative hierarchical clustering was performed by using the unweighted pair-group method with arithmetic mean technique (see text). A heat map for each gene for the three independent experimental data points is shown at right. For each experiment, the treatment conditions are given as control (C), doxycycline treated (D), and RSV infected (R). Green represents the minimum value of 5 scaled fluorescence intensity units, black represents the middle value of 5,000 scaled units, and red represents the maximum value of 10,000 scaled units. Top left is the number of nodes at each point on the horizontal scale. Single vertical bar is a subnode further analyzed in C; the double vertical is further analyzed in D. (C) Clustering and heat map analysis of the genes whose constitutive expression is influenced by Flag-IκBα Mut. Left, an enlargement of the dendrogram and heat map from B is shown; right, data from identically treated cells expressing empty vector. Gene 1 is growth hormone (GenBank accession no. J03071), 2 is protein tyrosine phosphatase (X54131), 3 is unknown (AL050002), 4 is serine/threonine kinase (AB018324), 5 is RAB4 (M28211), 6 is pp21 (M99701), 7 is frizzled (L37882), 8 is RNP A0 (U23803), 9 is unknown (AF035292), 10 is enolase (X56832), 11 is α1 type XI collagen (J04177), 12 is selenium donor (U34044), 13 is α1 type XVI collagen (M92642), 14 is Rar (U05227), 15 is microophthalmia-associated transcription factor (AB006909), 16 is UFD2 (AF043117), and 17 is FIP-1 (U41654). ∗∗ indicates genes influenced by doxycycline treatment alone. (D) Clustering and heat map analysis of genes upregulated by RSV and inhibited by Flag-IκBα Mut. Left, an enlargement of the dendrogram and heat map from B is shown; right are array data from identically treated cells expressing empty vector. Details of the genes numbered to the right of the Flag-IκB Mut heat map are provided in Table 1.

FIG. 7.

References

1. Aherne, W. T., T. Bird, S. D. B. Court, P. S. Gardner, and J. McQuillin. 1970. Pathological changes in virus infections of the lower respiratory tract in children. J. Clin. Pathol. 23:7-18. - PMC - PubMed
1. Arsura, M., F. Mercurio, A. L. Oliver, S. S. Thorgeirsson, and G. E. Sonenshein. 2000. Role of the IκB kinase complex in oncogenic Ras- and Raf-mediated transformation of rat liver epithelial cells. Mol. Cell. Biol. 20:5381-5391. - PMC - PubMed
1. Baldwin, A. S. J. 2001. The transcription factor NF-κB and human disease. J. Clin. Investig. 107:3-6. - PMC - PubMed
1. Barkett, M., and T. Gilmore. 1999. Control of apoptosis by Rel/NF-κB transcription factors. Oncogene 18:6910-6924. - PubMed
1. Barnes, P. J., and M. Karin. 1997. Nuclear factor-κB: a pivotal transcription factor in chronic inflammatory diseases. N. Engl. J. Med. 336:1066-1071. - PubMed

Identification of NF-kappaB-dependent gene networks in respiratory syncytial virus-infected cells - PubMed (original) (raw)