Inflammatory conditions induce gap junctional communication between rat Kupffer cells both in vivo and in vitro

Abstract

Connexin43 (Cx43), a gap junction protein subunit, has been previously detected in Kupffer cells (KCs) during liver inflammation, however, KCs phagocytose cell debris that may include Cx43 protein, which could explain the detection of Cx43 in KCs. We determined that KCs express Cx43 and form gap junctions (GJs) both in vivo and in vitro. In liver sections of animals treated with LPS, Cx43 was detected at ED2+ cells interfaces, indicating formation of GJs between KCs in vivo. In vitro, unstimulated KCs cultures did not form functional GJs, and expressed low levels of Cx43 that showed a diffuse intracellular distribution. In contrast, KCs treated with LPS plus IFN-gamma, expressed a greater amount of Cx43 at both, protein and mRNA levels, and showed Cx43 at cell-cell contacts associated with higher dye coupling. In conclusion, activation of KCs in vivo or in vitro resulted in enhanced Cx43 expression levels and formation of GJ that might play relevant roles during liver inflammation.

Figure 1. LPS administration induces formation of gap junction plaques between KCs in the liver

Immunohistochemistry and confocal microscopy of liver sections from normal and LPS injected rats was performed to analyze distribution and levels of Cx43 immunoreactivity in ED2 positive cells (KCs). Panels A, E and I, correspond to DAPI nuclear stain. (A–D) In normal liver sections few KCs (ED2+ cells) were positive to Cx43. (E–H) Liver sections obtained from rats injected with LPS 24 h before sacrifice show aggregation of KCs with strong signal for Cx43 at cell-cell appositions. I to L, are enlargements of the same area shown in E–H located within the dotted square illustrated in H, to show colocalization (yellow signal) of both Cx43 and ED2. PP venule: periportal venule are indicated by arrows. Bar: 150 μm for A–H and 35 μm for I–L.

Figure 2. Treatment with LPS plus IFN-γ induces gap junctional communication and redistribution of Cx43 to cell-cell contacts between rat KCs

(A) The incidence of dye coupling (LY) was evaluated in cultures of rat KCs under control conditions and after LPS plus IFN-γ treatment. The incidence of coupling in control conditions was ~5 % (▪). In one set of experiments, treatment with 1 μg/ml LPS plus 100 U/ml IFN-γ induced a progressive increase in incidence of dye coupling (●). In a second set of experiments, the dye coupling induced by LPS plus IFN-γ was sensitive to the acute application of 18-α-glycyrrhetinic acid (AGA, 35 μM) or octanol (500 μm, Graph B) in a reversible way after washout (W/O) of the blocker (▴, see doted line). Each point corresponds to the average of 9 (AGA, Graph A) or 5 (Octanol, Graph B) experiments, in which a minimum of 20 cells were scored. (C) Dye coupling induced by LPS plus IFN-γ between KCs, was not altered by treatment with probenecid (50 μM), an organic anion transporter blocker, or oxidized ATP (oATP, 100 μM), a P2X channel blocker, suggesting that the dye coupling observed was due to gap junctions channels and no other channels permeable to LY, such as organic anion transporters or P2X ATP receptors (n=4, * p<0.005). Each time point corresponds to the average ± SD.

Figure 3

Immunofluorescence for Cx43 (with anti-Cx43 F_ab antibody) of rat KCs in culture in control and after 2 h of treatment with LPS plus IFN-γ. A, C and E, are phase contrast views of the fluorescent fields shown in B, D and F, respectively. Under control conditions the distribution Cx43 was mostly perinuclear (B). However, after LPS plus IFN-γ treatment strong Cx43 reactivity was detected at cell-cell appositions (D and F). E and F, are enlargements of the area indicated in panel D (dotted square). No immunostaining with preimmune serum was detected (data not shown), Bar: 70 μm for A–D and 8 μm for E and F.

Figure 4. Treatment with LPS plus IFN-γ increases the levels of Cx43 protein and mRNA in KC cultures

(A) The relative levels of Cx43 mRNA were measured by semiquantitative RT-PCRs. Ethidium bromide staining shown as a band of 291 bp (arrow in the upper gel) corresponds to the amplification product obtained using Cx43 specific primers and total RNA extracted from KC cultures treated with LPS plus IFN-γ for 4 and 9 h. β-actin was used a loading control and their amplification resulted in a amplification product of 281 pb. MW corresponds to 100 bp DNA ladder (n=3). (B) Cx43 protein levels were determined by Western blot analysis from KC lysates (150 μg of protein per lane) using a rabbit anti-Cx43 antibody. Connexin43 levels were analyzed at zero time or after 1 μg/ml LPS plus 100 U/ml IFN-γ treatment for 1, 2, 4 and 9 h. An aliquot of rat heart (30 μg of proteins) was used as positive control for Cx43 (lane S). The unphosphorylated (NP) and two phosphorylated forms of Cx43 (P2 and P3) are indicated on the left side of the immunoblot.