*2.5. Localization of JAM-A in Alveolar Epithelial Cells after BLM Treatment and the Influence of oxATP*

Confluent E10 cells were incubated with 100 mU/mL BLM to study the effect of BLM on the distribution of JAM-A. Immunofluorescence revealed a regular localization of JAM-A to TJ at sites of cell–cell contact in the control cells (Figure 4). BLM exposure for 24 h (Figure 4B) or oxATP alone (Figure 4C) did not change JAM-A immunoreactivity and its cellular localization, but the cell size increased after BLM exposure. oxATP and BLM together normalized the BLM induced cell swelling (Figure 4D).

N'D N'D

**Figure 4.** Effects of P2X7R inhibition by 150 μM oxATP, which was added 2 h prior to 100 mU/mL BLM treatment. Equal protein amounts of cell lysates were used in SDS-PAGE and analyzed by Western blot. α-Tub served as the loading control. Untreated cells were used as the control and normalized to 100%. Representative blots from three independent experiments are shown. Charts are presented as the mean ± SEM (*n* = 3) of P2X7R/ α-Tub, GSK-3β(Ser9)/ α-Tub and JAM-A/ α-Tub. P-values: P2X7R 0.0338; GSK-3β(Ser9) 0.002; and JAM-A 0.05. Immunofluorescence demonstration of JAM-A in untreated (**A**), BLM (**B**), or oxATP (**C**) treated E10 cells. Note the increased cell size after BLM exposure (**B**), which was ameliorated after oxATP (**D**). Representative images of multiple experiments (*n* = 3) are shown. Bar = 20 μm. \* *p* < 0.05, \*\* *p* < 0.01.

#### *2.6. The P2X7R Agonist BzATP A*ff*ects the JAM-A Protein Content*

Next, we explored whether stimulation of P2X7R in alveolar epithelial E10 cells with BzATP would induce a modulation in the JAM-A protein content. E10 cells were exposed to 100 μM BzATP for one and two days to stimulate P2X7R. The Western blots are shown in Figure 5.

**Figure 5.** Analysis of protein levels in alveolar epithelial cell line E10 after treatment with 150 μM BzATP. Cells were treated with BzATP for 24 h and 48 h. For SDS-PAGE, equal protein amounts of cell lysates were used and analyzed by Western blot with antibodies against P2X7R, JAM-A, and α-Tub. Untreated cells were used as the control and normalized to 100%. Protein levels were normalized to α-Tub and are shown as the mean ± SEM (*n* = 3) in relation to the control. One representative blot is pictured. P-values: P2X7R 0.1667 and JAM-A 0.1048.

The response to purinergic receptor stimulation by BzATP, which presumably involves a Ca2<sup>+</sup> channel opening, resulted in an increase in Ca2<sup>+</sup> [17]. As an indication for this, we were able to show in previous work that the CaM content was increased after BzATP treatment [17]. In the present experiments, we demonstrated that the JAM-A protein content was influenced by the BzATP treatment. After 24 h of BzATP treatment in E10 cells, a minimal increase in the JAM-A protein concentration occurred, whereas a stronger decrease was seen after 48 h (Figure 5).

#### **3. Discussion**

Maintenance of the integrity of the alveolar barrier is realized by TJ among neighboring AECs consisting of occludin, claudins, ZOs, and JAM-A.

Recently, we have demonstrated that in P2X7−/<sup>−</sup> mice, claudin-18 is upregulated and the inactive form of GSK-3β, GSK-3β(Ser9), is also upregulated in comparison to WT mice [12]. Our current data show that another TJ protein, JAM-A, in P2X7−/<sup>−</sup> mice is strongly upregulated at the protein level. P2X7−/<sup>−</sup> mice exhibited reduced lung inflammation with reduced fibrosis markers such as lung collagen [13]. Deletion of P2X7R is a protective factor in acute lung injury [14]. Through the upregulation of JAM-A and claudin-4 and -15 in a kinase-dependent manner, an improved barrier function of the oral epithelium could be demonstrated [21].

Our data suggest that increased constitutive JAM-A protein level may have a protective effect against BLM-induced lung injury in P2X7−/<sup>−</sup> mice. BLM-treated PCLS from P2X7−/<sup>−</sup> mice responded with a slighter increase in mRNA expression of JAM-A than BLM-treated PCLS from WT mice. The reduced level of JAM-A upregulation in the P2X7−/<sup>−</sup> mice in comparison to the WT mice indicates a lower sensitivity of this protein to the effects of BLM in the alveolar epithelium.

JAM-A is concentrated at epithelial and endothelial tight junctions and it has been shown that JAM-A localizes to claudin-based tight junction fibrils in epithelial cells [22]. It does not directly regulate the barrier between the cells, but rather interacts as a signaling molecule with divergent downstream target proteins [23]. Nevertheless, in various epithelial and endothelial cell lines including primary rat alveolar epithelial cells, it has been shown that siRNA mediated downregulation of JAM-A expression results in enhanced paracellular permeability, as determined by TER measurements [16]. Long-term treatment of PCLS (five days) with BLM showed a strong downregulation of JAM-A when compared to the untreated PCLS (own unpublished data).

To examine the consequences of decreased GSK-3β activity on JAM-A under BLM treatment, the GSK-3β inhibitor LiCl was used in this study. JAM-A is affected by the inhibition of GSK-3β. The initial upregulation of JAM-A under BLM treatment is prevented by the inactivation of GSK-3β. Inhibition of GSK-3β causes JAM-A expression to remain at levels comparable to that of the untreated cells.

This result indicates that inhibition of GSK-3β has a positive effect on the deregulatory changes in JAM-A expression under BLM treatment. Several studies have shown that the inhibition of GSK-3β reduces the development of acute lung injury and inflammation and has a protective effect on lung fibrosis induced by BLM [24,25]. The P2X7−/<sup>−</sup> mice, which showed no fibrotic changes under BLM treatment, had a high constitutive expression of GSK-3β(Ser9) [12], which is an indication of the protective effect of GSK-3β inhibition.

While the data on GSK-3β inhibition in acute lung injury are fairly clear, there are little data on its effect on TJ proteins. Severson et al. [20] have shown that endogenous GSK-3β activity is required for maintenance of the AJC, and therefore for epithelial barrier function by regulating the expression of transmembrane proteins claudin-1 and occludin. They also observed a differential decrease in the labeling of key AJC proteins following GSK-3β inhibition, a decrease in occludin, claudin-1, and E-cadherin protein levels, but they could not show any effect on JAM-A expression and localization. They reported that in both human intestinal (SK-CO15) and kidney (MDCK) epithelial cells, a decrease in GSK-3β activity interfered with epithelial cell–cell transitions, thereby increasing paracellular permeability.

It was previously shown by Bazzoni et al. [26] that the absence of JAM-A enhanced cell motility, increased membrane protrusions, affected microtubule stability, and reduced focal adhesions in endothelial cells. The consequences of JAM-A absence were reversed on treatment with GSK-3β inhibitors.

In this study, treatment with oxATP alone reduced the expression of P2X7R and increased JAM-A when compared to the untreated WT cells. After the addition of oxATP to BLM-treated cells, the P2X7R was downregulated, but a higher amount of JAM-A protein was still measured than in the untreated cells. In this case, JAM-A was also upregulated. Furthermore, inactivation of P2X7R by oxATP led to a substantial reduction in the constitutively present level of inactive GSK-3β in untreated and BLM-treated cells.

The inhibition of P2X7R under BLM treatment resulted in the opposite effect on GSK-3β(Ser9) when compared to the effect of the GSK-3β inhibitor LiCl. The inactivated form of GSK-3β, the GSK-3β(Ser9) was even further downregulated, which was expressed downstream in the upregulation of JAM-A. Figure 6 summarizes the data:

Interestingly, the inhibition of P2X7R under BLM did not lead to the upregulation of GSK-3β as found in P2X7−/<sup>−</sup> mice. The siRNA-mediated downregulation of P2X7R led in turn to the upregulation of GSK-3β in untreated alveolar epithelial cells [12]. The different effects on GSK-3β still have to be clarified. However, knockout or inhibition of P2X7R always leads to an increase in the JAM-A protein level, indicating a repressive effect of P2X7R on the expression of JAM-A.

Furthermore, we have shown that activation of P2X7R by BzATP resulted first in a very slight upregulation of JAM-A after 24 h, and then in a strong downregulation of the protein after 48 h in alveolar epithelial cells E10. Guo et al. [27] showed in E10 cells that activation of P2X7R by BzATP increased the phosphorylation of Y216 and decreased the phosphorylation at S9 in GSK-3β without affecting the total GSK-3β expression. This result indicates that BzATP stimulates GSK-3β activity.

It has been demonstrated that JAM-A has the ability to promote the assembly and remodeling of alveolar epithelial tight junctions in response to acute lung injury and plays a protective role in preventing lung damage and promoting fluid clearance [28]. Identifying pathways that increase the expression and function of JAM-A in acute lung injury may identify new approaches to promote barrier function in response to inflammation and injury.

We were able to show for the first time that P2X7R plays an important role in the regulation of JAM-A in the alveolar epithelium. Downregulation of P2X7R or the absence of the protein leads to upregulation of JAM-A, possibly resulting in an increase in barrier function. Conversely, stimulation of P2X7R leads to downregulation of the protein. A modulating effect on JAM-A has been demonstrated under BLM treatment for GSK-3β.

**Figure 6.** BLM treatment results in increased protein levels of P2X7R and JAM-A as well as in a reduced content of the inactive form of GSK-3β GSK-3β(Ser9). After inhibition of P2X7R by oxATP under BLM exposure, the effect on both proteins is further enhanced. Inactivating of the GSK-3β by LiCl under BLM exposure directly leads to a reduction of JAM-A. The influence of P2X7R on JAM-A is rather indirect.

#### **4. Materials and Methods**

#### *4.1. Ethics Statement*

All animal experiments were approved by the Ethics Committee of the Dresden University of Technology and the license for the removal of organs was provided by Landesdirektion Dresden (file no. 24-9168.24-1/2007-26; file no. 24-9168.24-1/2010-11).

#### *4.2. Experimental Animals*

WT mice were purchased from Charles River (C57BL/6; Charles River, Wilmington, USA) and the P2X7−/<sup>−</sup> mice were obtained from Pfizer (B6.129P2-P2rx7tm1Gab/J; Pfizer, New York, NY, USA) [29]. Our animals were housed at the Animal Care Facility at the Medical Faculty "Carl Gustav Carus" of Dresden University of Technology and had steady free access to standard chow and water. All performed procedures were in accordance with the Technical University of Dresden Animal Care and Use Committee Guidelines. For our experiments, we examined the lung tissue of male and female mice with an age of 8 to 16 weeks.
