Lipopolysaccharide enhances transforming growth factor β1-induced platelet-derived growth factor-B expression in bile duct epithelial cells

Abstract

Background and aim: Platelet-derived growth factor (PDGF)-B is a potent profibrogenic mediator expressed by bile duct epithelial cells (BDECs) that contributes to liver fibrosis after bile duct ligation. However, the mechanism of PDGF-B induction in BDECs during cholestasis is not known. Transforming growth factor β (TGFβ) and lipopolysaccharide (LPS) also contribute to the profibrogenic response after bile duct ligation. We tested the hypothesis that LPS and TGFβ1 synergistically induce PDGF-B expression in BDECs.

Methods: Transformed human BDECs (MMNK-1 cells) and primary rat BDECs were stimulated with LPS and/or TGFβ1, and signaling pathways through which LPS potentiates TGFβ1-induced PDGF-B mRNA expression were investigated.

Results: Stimulation of MMNK-1 cells with LPS alone did not significantly induce PDGF-B mRNA expression. However, LPS co-treatment enhanced TGFβ1 induction of PDGF-B mRNA in MMNK-1 cells and also in primary rat BDECs. Importantly, co-treatment of MMNK-1 cells with LPS and TGFβ1 also significantly increased PDGF-BB protein expression. Interestingly, LPS did not affect TGFβ1 activation of a SMAD-dependent reporter construct. Rather, stimulation of MMNK-1 cells with LPS, but not TGFβ1, increased JNK1/2 phosphorylation. Expression of dominant negative JNK2, but not dominant negative JNK1, inhibited the LPS potentiation of TGFβ1-induced PDGF-B mRNA expression in MMNK-1 cells. In addition, LPS treatment caused IκBα degradation and activation of a nuclear factor κB (NFκB)-dependent reporter construct. Expression of an IκBα super repressor inhibited activation of NFκB and attenuated LPS potentiation of TGFβ1-induced PDGF-B mRNA.

Conclusions: The results indicate that LPS activation of NFκB and JNK2 enhances TGFβ1-induced PDGF-B expression in BDECs.

Figure 1. Effect of TGFβ1 and LPS on PDGF-B mRNA and protein induction in MMNK-1 cells and rat BDECs

(A,B) MMNK-1 cells were treated with 100 ng/mL LPS and/or 5 ng/mL TGFβ1 for the indicated times, and (A) PDGF-B mRNA and (B) PDGF-BB protein levels were determined. Data are expressed as mean ± SEM and as fold change vs. control-treated cells (mRNA) or as pg/ml (protein). (C) Primary rat BDECs were stimulated with 100 ng/mL LPS for one hour then treated with 5 ng/mL TGFβ1 for the indicated times, and PDGF-B mRNA levels were determined. Data are expressed as mean ± SEM and as fold change vs. control-treated cells. n=3–5 independent experiments *Significantly different than TGFβ1 alone, p<0.05.

Figure 2. Effect of TGFβ1 and LPS on SMAD reporter activity in MMNK-1 cells

MMNK-1 cells transfected overnight with a SMAD-dependent luciferase reporter construct were treated with 100 ng/mL LPS and/or 5 ng/mL TGFβ1 for 4 hr. Firefly luciferase relative light units were determined and normalized to renilla luciferase expression. Data are expressed as mean ± SEM and as fold change vs. control-treated cells. n=3–4 independent experiments *Significantly different than the same treatment in the absence of TGFβ1, p<0.05.

Figure 3. Role of the JNK1/2 MAPK pathway in LPS potentiation of TGFβ1-induced PDGF-B mRNA expression in MMNK-1 cells

(A) MMNK-1 cells were treated with 100 ng/mL LPS and/or 5 ng/mL TGFβ1 for 1 hr. Levels of phosphorylated and total JNK1/2 were determined by western blotting. A western blot representative of three independent experiments is shown. (B–C) MMNK-1 cells were transfected overnight with pcDNA3.1 (CTL), DN-JNK1, or DN-JNK2 expression vectors. Cells were treated with 100 ng/mL LPS and/or 5 ng/mL TGFβ1, and PDGF-B mRNA levels were determined (B) 2 hr or (C) 4 hr later. Data are expressed as mean ± SEM and as fold change vs. control-treated cells. n=3–4 independent experiments *Significantly different than the same treatment in the absence of LPS, p<0.05. #Significantly different from cells transfected with pcDNA3.1 and given the same treatment, p<0.05.

Figure 4. NFκB activation in MMNK-1 cells stimulated with LPS and TGFβ1

(A) MMNK-1 cells were treated with 100 ng/mL LPS and/or 5 ng/mL TGFβ1 for 1 hr. Cytosolic levels of IκBα and GAPDH were determined by western blotting. A western blot representative of four independent experiments is shown. (B) MMNK-1 cells were transfected overnight with a NFκB-dependent luciferase reporter construct then treated with 100 ng/mL LPS and/or 5 ng/mL TGFβ1 for 4 hr. Firefly luciferase relative light units were determined and normalized to renilla luciferase expression. Data are expressed as mean ± SEM and as fold change vs. control-treated cells. n=10 independent experiments *Significantly different than the same treatment in the absence of LPS, p<0.05.

Figure 5. Role of NFκB in LPS potentiation of TGFβ1-induced PDGF-B mRNA expression in MMNK-1 cells

(A,D) MMNK-1 cells were transfected overnight with a NFκB-dependent luciferase reporter construct and pcDNA3.1 (CTL) or (A) a DN-IKK2 construct or (D) an IκBα super repressor (IκBα-SR). Cells were treated with 100 ng/mL LPS and/or 5 ng/mL TGFβ1 for 4 hr. Firefly luciferase relative light units were determined and normalized to renilla luciferase expression. Data are expressed as mean ± SEM and as fold change vs. control-treated cells. (B–C) MMNK-1 cells were transfected overnight with CTL or DN-IKK2 then treated with 100 ng/mL LPS and/or 5 ng/mL TGFβ1. PDGF-B mRNA expression was determined (B) 2 hr or (C) 4 hr later. Data are expressed as mean ± SEM and as fold change vs. control-treated cells. (E–F) MMNK-1 cells were transfected overnight with CTL or IκBα-SR then treated with 100 ng/mL LPS and/or 5 ng/mL TGFβ1. PDGF-B mRNA expression was determined (E) 2 hr or (F) 4 hr later. Data are expressed as mean ± SEM and as fold change vs. control-treated cells. n=3–6 independent experiments *Significantly different than the same treatment in the absence of LPS, p<0.05. #Significantly different from cells transfected with pcDNA3.1 and given the same treatment, p<0.05.