Collagen XVII (BP180) modulates keratinocyte expression of the proinflammatory chemokine, IL-8

Abstract

Collagen XVII (COL17), a transmembrane protein expressed in epidermal keratinocytes (EK), is targeted by pathogenic autoantibodies in bullous pemphigoid. Treatment of EK with anti-COL17 autoantibodies triggers the production of proinflammatory cytokines. In this study, we test the hypothesis that COL17 is involved in the regulation of the EK proinflammatory response, using IL-8 expression as the primary readout. The absence of COL17 in EK derived from a junctional epidermolysis bullosa patient or shRNA-mediated knockdown of COL17 in normal EK resulted in a dysregulation of IL-8 responses under various conditions. The COL17-deficient cells showed an abnormally high IL-8 response after treatment with lipopolysaccharide (LPS), ultraviolet-B radiation or tumor necrosis factor, but exhibited a blunted IL-8 response to phorbol 12-myristate 13-acetate exposure. Induction of COL17 expression in COL17-negative EK led to a normalization of the LPS-induced proinflammatory response. Although α6β4 integrin was found to be up-regulated in COL17-deficient EK, siRNA-mediated knockdown of the α6 and β4 subunits revealed that COL17's effects on the LPS IL-8 response are not dependent on this integrin. In LPS-treated cells, inhibition of NF-kappa B activity in COL17-negative EK resulted in a normalization of their IL-8 response, and expression of an NF-kappa B-driven reporter was shown to be higher in COL17-deficient, compared with normal EK. These findings support the hypothesis that COL17 plays an important regulatory role in the EK proinflammatory response, acting largely via NF-kappa B. Future investigations will focus on further defining the molecular basis of this novel control network.

Figure 1. Alterations in COL17 expression in EK are associated with a dysregulation of the pro-inflammatory response

EK with normal and abnormal COL17 expression levels were exposed to UVB radiation (500 mJ/cm² over a period of 1 minute; panels a and b), 25 μg/ml LPS (c,d), or 20 ng/ml PMA (e,f). The culture supernatants were harvested 4 h after the start of treatment and analyzed by ELISA for IL-8 protein (a,c,e). At this same time point, mRNA was isolated from the cells and analyzed by qPCR for IL-8 mRNA (b,d,f). Each bar in this figure represents the average of three assay results ± S.D. The data shown are representative of a minimum of three experiments. Compared with UVB-treated NHEK and JEBEK+, UVB-treated JEBEK produced higher levels of IL-8 protein (panel a; each p<0.02) and mRNA (b; each p<0.04). Likewise, LPS-treated JEBEK produced higher levels of IL-8 protein (c) and mRNA (d) compared with LPS-treated NHEK [p(protein)<0.02; p(mRNA)p<0.01)] and JEBEK+ [p(protein)<0.05; p(mRNA)<0.01]. In contrast, after PMA treatment (e and f) NHEK produced a higher level of IL-8 protein than either JEBEK+ (p<0.02) or JEBEK (p<0.02), but there were no significant differences in the IL-8 mRNA levels produced by these 3 cell lines. Black bars = JEBEK; Dark gray bars = JEBEK+; light gray bars = NHEK.

Figure 2. shRNA-mediated knockdown of COL17 expression in NHEK leads to an enhancement of the LPS-induced IL-8 response

(a) JEBEK and NHEK were transfected with a GFP expression construct alone (“mock”) or were co-transfected with GFP and either a control shRNA (lacZ) or COL17-specific shRNA, as indicated. The bar graph shows the COL17 labeling intesities of the GFP-positive cells, as determined by flow cytometry and represented as mean fluorescence intensity (MFI). The transfection efficiencies ranged between 32-36%. The data are representative of three experiments. (b) NHEK, JEBEK (labeled “JEB”) and JEBEK+ (“JEB+”) transfected with either the COL17-specific shRNA (labeled “C17”) or the lacZ shRNA construct were exposed to 25 μg/ml LPS (black bars) or plain medium (gray bars) for 4 hours, after which the IL-8 levels in the culture supernatant fractions were measured by ELISA. After LPS treatment, the level of IL-8 produced by the COL17 knockdown NHEK was significantly higher than that of the control knockdown NHEK (indicated by an asterisk; p<0.03). In response to LPS, COL17-positive JEBEK+(control shRNA) expressed significantly less IL-8 than either JEBEK(control shRNA; p<0.03) or COL17-knock-down JEBEK+ (p<0.05). As expected, the COL17 shRNA had no effect on the IL-8 response of JEBEK. The data shown are representative of three experiments. Each bar represents the average ± S.D. of triplicate assay results.

Figure 3. The up-regulation of α6β4 integrin expression in a COL17-deficient EK line is not essential for the enhanced IL-8 response of these cells

(a) JEBEK and NHEK were transfected with an α6 integrin-specific (si-α6) or control siRNA (si-ctrl) and then analyzed by flow cytometry for expression of α6 integrin. In the control-transfected cells (as well as in untransfected cells -- data not shown), JEBEK expressed a level of α6 integrin on their surface that was much higher than that of NHEK (MFI = 1,233 and 546 relative units, respectively). Transfection with the α6-specific siRNA reduced α6 expression by 52% in JEBEK and by 40% in NHEK. Panel b shows the results of a representative Western blot analysis (n=3) of β4 integrin expression in JEBEK and NHEK that were transfected with a β4 integrin-specific (si-β4) or a control siRNA (si-ctrl). In si-ctrltransfected cells, JEBEK expressed 62% more β4 integrin compared with that of NHEK. After transfection with the β4 siRNA, β4 integrin expression was reduced in JEBEK and NHEK by 46% and 36%, respectively. Panel c shows the analysis of LPS-induced IL-8 expression by JEBEK and NHEK transfected with the α6 / β4 integrin-specific and control siRNAs described above. No significant differences were seen in the pairwise comparisons within each cell line. The data shown in each of the three panels are representative of three experiments. In panel c, each bar represents the average ± S.D. of triplicate assay results.

Figure 4. Inhibitor and reporter gene analyses point to NFKB as a mediator of COL17’s effects on the EK IL-8 response

(a) NHEK (gray bars) and JEBEK (black bars) were treated for 12 h with Bay-11-7082 (“BAY”, an NFKB inhibitor) or SB203580 (“SB”, a p38MAPK inhibitor) prior to a 4 h incubation with 25 μg/ml LPS or vehicle control. Addition of increasing concentrations of Bay-11-7082 to the cells corresponded to decreasing differences in IL-8 levels of NHEK and JEBEK. This trend was observed throughout the range of Bay-11-7082 concentrations tested, only two of which are shown – 250nM, 500nM, 1μM, 2.5μM, 5μM, 10μM. Increasing concentrations of SB203580 (5, 10, 25, 50, 100μM; data for 10 and 50 μM are shown) did not show this same trend. (b) NHEK and JEBEK (labeled “JEB”) were treated with TNF, LPS or control in the presence or absence of Bay-11-7082. Comparing with vs. without inhibitor, significant decreases (asterisks) in IL-8 levels were seen in TNF-treated NHEK (p<0.05), as well as in TNF- and LPS-treated JEBEK (both p values < 0.05). (c) NHEK and JEBEK were transfected with either an NFKB-driven or AP-1-driven luciferase gene construct and then treated with LPS for 4h in the presence or absence of Bay-11-7082 or SB203580. Expression of the NFKB-driven reporter was much higher in LPS-treated JEBEK compared with either LPS-treated NHEK (p<0.01) or JEBEK without LPS (p<0.01). As far as the AP-1-driven reporter, expression in LPS-treated JEBEK was higher than in LPS-treated NHEK (p<0.05), but was not significantly higher than in JEBEK without LPS. It is noteworthy that both reporter genes were expressed at higher levels in JEBEK vs. NHEK in the absence of LPS. The data shown in each of the three panels are representative of three experiments. Each bar represents the average ± S.D. of triplicate assay results.