Induction of an inflammatory loop by interleukin-1β and tumor necrosis factor-α involves NF-kB and STAT-1 in differentiated human neuroprogenitor cells

Abstract

Proinflammatory cytokines secreted from microglia are known to induce a secondary immune response in astrocytes leading to an inflammatory loop. Cytokines also interfere with neurogenesis during aging and in neurodegenerative diseases. The present study examined the mechanism of induction of inflammatory mediators at the transcriptional level in human differentiated neuroprogenitor cells (NPCs). Interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) induced the expression of cytokines and chemokines in differentiated human NPCs as shown by an immune pathway-specific array. Network motif (NM) analysis of these genes revealed 118 three-node NMs, suggesting complex interactions between inflammatory mediators and transcription factors. Immunofluorescent staining showed increases in the levels of IL-8 and CXCL10 proteins in neurons and glial cells. Findings from Taqman low density array suggested the synergistic actions of IL-1β and TNF-α in the induction of a majority of inflammatory genes by a mechanism involving NF-kB and STAT-1. Nuclear localization of these transcription factors in differentiated NPCs was observed following exposure to IL-1α and TNF-α. Further studies on CXCL10, a chemokine known to be elevated in the Alzheimer's brain, showed that TNF-α is a stronger inducer of CXCL10 promoter when compared to IL-1β. The synergy between these cytokines was lost when ISRE or kB elements in CXCL10 promoter were mutated. Our findings suggest that the activation of inflammatory pathways in neurons and astrocytes through transcription factors including NF-kB and STAT-1 play important roles in neuroglial interactions and in sustaining the vicious cycle of inflammatory response.

Figure 1. Human NPC-derived neuronal model.

A. NPCs, derived from human fetal brain, were cultured and expanded as neurospheres in suspension. B. Four day-old neurospheres were seeded in dishes coated with poly-L-lysine and mouse laminin and cultured in the presence of 20 ng/ml NGF, 10 ng/ml BDNF, 100 µM dibutryl cyclic AMP and 1 µM retinoic acid for two wks. Differentiated NPCs were fixed and immunostained for neuronal markers MAP2a with FITC (B green) and β tubulin with cy3 (C; red) and for the glial marker, GFAP with FITC (D; green). Nuclei were stained blue with DAPI. The neuronal markers were observed in ∼90% of differentiated NPCs.

Figure 2. Network motifs (NMs) involved during the induction of inflammatory mediators.

From the list of inflammatory genes induced in differentiated human NPCs exposed to IL-1β and TNF-α, protein-protein and protein-DNA interactions were extracted from appropriate databases. NM analysis was performed using FANMODE tool. Representative NMs (A-L) from 118 identified are presented. Multi-input motifs (Fig. 2A–Fig. 2C) and feed-forward loops (Fig. 2D–Fig. 2G) were observed.

Figure 3. Cytokine-mediated elevation of IL-8 and CXCL10 proteins.

Human NPCs, differentiated for 2 weeks, were preincubated with Golgi plug (1 µl/ml), a secretion inhibitor and then exposed to a combination of 5 ng/ml each of IL-1β and TNF-α for 18 h. Treated cells were fixed and immunostained for MAP2a with FITC (green) and for IL-8 (A) or CXCL10 (B) with cy3 (red). Cytokine-mediated increases in the levels of IL-8 and CXCL10 were observed in MAP2a-positive cells. C. Differentiated NPCs exposed to IL-1β and TNF-α (Cyt) were processed for the Western blot analysis of IL-8, CXCL10 and β actin. Representative images are presented. The band intensities of IL-8 and CXCL10 were determined by scanning and corrected for the levels of β actin. The results are M±SE of three independent observations. **P<0.001 compared to untreated control.

Figure 4. Cytokine-mediated activation of signaling to NF-kB and STAT-1.

Differentiated NPCs were exposed to 5 ng of IL-1β and TNF-α, alone or in combination, for 2 h and processed for the Western blot analysis of the upstream kinases of NF-kB (A and B) and STAT-1 (C and D) pathways. Band intensities were quantitated by scanning. Increased phosphorylation of IKK (PIKK) and p65 (Pp65) was observed, especially in TNF-α-treated cells with IL-1β showing modest actions. Both cytokines showed additive actions in the phosphorylation of JAK2 (PJAK2) and STAT-1 (PSTAT-1). The results are M±SE of three independent observations. #P<0.05, *P<0.01 and **P<0.001 versus untreated control.

Figure 5. Cytokine-mediated nuclear localization of NF-kB and STAT-1.

Differentiated NPCs were treated with TNF-α (A; 5 ng/ml) or a combination of IL-1β and TNF-α (B; 5 ng/ml each) for 6 h, fixed and immunostained for the phosphorylated form of p65 (Pp65; A), a member of NF-kB family or STAT-1 (B) with cy3 (red) and for MAP2a with FITC (green). Nuclear localization of phospho p65 and STAT-1 was observed following exposure to TNF-α and a combination of IL-1β and TNF-α respectively. Representative images from three independent experiments are presented.

Figure 6. Induction of CXCL10 promoter by IL-1β and TNF-α.

A. Truncated wild type, ISRE mutant and kB mutant CXCL-10 promoter constructs linked to a firefly luciferase reporter gene were used in the transient transfection assay. B. Human NPC-derived neurons were transfected with wild type CXCL10 promoter along with a constitutively active renilla luciferase reporter and exposed to increasing concentrations of IL-1β and TNF-α. The cytokines were also added to transfected neurons in the absence and presence of 2 µM of Aβ oligomers or Aβ fibrils (C). After 18 h, the treated neurons were processed for the assay of luciferases. The ratios of firefly and renilla luciferase were taken as measures of promoter activity. TNF-α was a stronger inducer of CXCL10 compared to IL-1β. *P<0.01; **P<0.001 compared to untreated control. #P<0.01 vs cells treated with cytokines in the absence of Aβ aggregates.

Figure 7. Cytokine-mediated induction of CXCL10 promoter requires NF-kB and STAT-1.

A. Differentiated NPCs transfected with wild type CXCL10 promoter were preincubated in the presence of 10 µM of Bay 11-7085, NF-kB inhibitor or 1 µM of JAK inhibitor to block STAT-1 activation, followed by exposure to 2 ng/ml each of IL-1β and TNF-α for 18 h. B. Cells were transfected with wild type, and mutant promoter constructs and cultured in the presence of 2 ng/ml each of IL-1β, TNF-α or both for 18 h. Treated cells (A and B) were processed for the assay of luciferases. Inhibition of NF-kB or STAT-1 resulted in significant decreases in cytokine-mediated promoter activation. Mutation of ISRE or kB elements led to loss of synergy between IL-1β and TNF-α. *P<0.01; **P<0.001 compared to untreated control. #P<0.01 vs cells treated with cytokines in the absence of inhibitors (A) or cells transfected with wild type promoter (B).