Cell Line Specific Modulation of Connexin43 Expression after Exposure to Ionizing Radiation

Abstract

Gap junctional intercellular communication plays a significant role in mediating radiation-induced bystander effects. However, the level of Cx43 itself is influenced by ionizing radiation, which could modify the bystander effect. Here we have investigated several cell lines for the modulation of Cx43 expression 24 h after irradiation with 5 Gy X-rays. The mouse endothelial cell line bEnd3 revealed a significantly elevated level of Cx43 already 15 min after exposure to X-rays, whereas human hybrid endothelial cells (EA.hy926) exhibited a transient downregulation of Cx43 mRNA. No obvious changes in the communication properties of the different cell lines could be observed after irradiation. The communication-deficient malignant human trophoblast cell line Jeg3 stably transfected with Cx43 did not reveal any induction of endogenous nor alteration in the exogenous Cx43 transcript level upon exposure to 5 Gy. Taken together, our data show a cell line specific modulation of Cx43 expression after exposure to X-rays.

Keywords:

• Connexin43
• gap junctions
• irradiation
• Jeg3
• bEnd3
• trophoblast

INTRODUCTION

Recent evidence suggests that multicellular organisms coordinate their response to ionizing radiation by intercellular communication via gap junction channels (1, 2, 3, 4, 5, 6). However, it has been controversially discussed whether gap junctional intercellular communication (GJIC) is involved in mediating bystander effects.

Gap junction channels may mediate bystander effects by allowing direct intercellular exchange of small molecules, ions, and second messengers (cAMP, IP₃, Ca^{2 +}) between irradiated and nonirradiated bystander cells which enable the channels to control and coordinate cell growth. The connexin gene family consists of 20 members in the murine and 21 members in the human genome which are expressed in a tissue specific manner (7).

Radiation-induced bystander effects occur in cells that are not directly irradiated but that receive signals from neighboring irradiated cells (8). Biological effects of low doses of ionizing radiation are currently intensively debated. It has been demonstrated that the lethal effects of ionizing radiation at very low doses are underestimated (9, 10).

Bystander effects mediate the killing of unirradiated cells, the induction of sister chromatid exchanges, chromosomal aberrations, the induction of gene mutations, micronuclei formation, the transformation to the neoplastic state, changes in gene expression, an increase in reactive oxygen species (ROS), and a decrease or an enhancement of cell clonogenicity as well as radioresistance (reviewed by 11, 12). Although the molecular mechanisms of bystander effects are not well understood, it has been suggested that direct cell-to-cell communication via gap junctions and/or diffusible cellular factors excreted from irradiated cells in the growth medium are involved (1, 2, 13, 14, 15, 16, 17, 18, 19, 20).

As connexins have been shown to affect cellular sensitivity and response to ionizing radiation, it seems likely that ionizing radiation could directly influence the level of gap junction protein expression and thereby the extent of bystander effects. Previous reports have already shown that the level of connexin proteins is changed after irradiation, but no information is available on its mechanisms (1, 2, 11, 21, 22, 23, 24). It is known that low dose radiation-induced alteration of Cx43 expression is tightly controlled by various transcription factors binding to the Cx43 promoter, such as nuclear factor of activated T-cells (NFAT) and activator protein (AP-1) (24).

Here we focus on irradiation effects in modulating cell-cell communication properties. Therefore we examined Cx43 expression after X-ray irradiation of mouse endothelial cells (bEnd3), human hybrid endothelial cells (EA.hy926) and malignant human trophoblast cells (Jeg3) stably transfected with Cx43. Both endothelial cells as well as the epithelial Jeg3 cell line are well known to respond to environmental stress such as ionizing radiation by induction of p53 (25, 26). Furthermore, the communication-deficient Jeg3 cell line can be transfected with inducible connexin proteins and thus serves as an ideal cell system to investigate the role of gap junction channels in mediating the radiation-induced bystander effect.

We demonstrate that bEnd3 cells show a significantly increased level of Cx43 transcript and protein expression after exposure to X-rays. In contrast, the EA.hy926 cells exhibit a transient downregulation of Cx43 mRNA after irradiation which returns to the basic level of expression after 24 h. Despite obvious alterations in mRNA and protein levels in the different cell lines, no obvious changes in the communication properties of the cell lines could be measured by calcein dye transfer combined with fluorescence activated cell sorting (FACS) analysis. The malignant communication-deficient Jeg3 cells stably transfected with Cx43 demonstrated no regulation or induction of Cx43 after irradiation.

We conclude that it is advisable to characterize the connexin expression levels and communication properties of the used cell lines after irradiation before examining bystander effects to exclude a change in connexin expression pattern and probably channel properties as a primary effect of radiation exposure.

MATERIAL AND METHODS

Cell Line and Culture

The human choriocarcinoma cell line Jeg3 (ATCC HTB-36) was purchased from the American Type Culture Collection (Manassas, USA). Jeg3 cells were grown in minimal essential medium (MEM, Invitrogen, Karlsruhe, Germany) supplemented with 10% fetal calf serum (certified tetracycline-free; Biochrome, Berlin, Germany).

Jeg3 Tet/Cx43 transfectants were cultivated in medium containing 500 μ g/ml G418 sulfate (PAA laboratories, Cölbe, Germany) and 0.5 μ g/ml puromycin (Sigma, Munich, Germany). Cx43 expression in these cells was induced following 48 h treatment of the cells with 1 μ g/ml doxycycline HCl (Dox) (Sigma, Munich, Germany).

The human hybrid EA.hy926 cell line was derived by fusing human umbilical vein endothelial cells with the permanent human lung epithelial cell line A549 (kindly provided by Dr. Cora-Jean S. Edgell, University of North Carolina, USA). EA.hy926 cells were grown in Dulbecco's Modified Eagles Media (DMEM, Invitrogen, Karlsruhe, Germany) with 4.5 g/l glucose and supplemented with 10% fetal calf serum.

The immortalized endothelial bEnd3 cell line, originally derived from mouse brain capillaries (27) was maintained in DMEM media supplemented with 10% fetal calf serum and 1% penicillin – streptomycin sulphate (Invitrogen, Karlsruhe, Germany).

Plasmid Construction and Transfection

The plasmid construction of the rat Cx43 expression plasmid based on the tet-on system and the transfection procedure were described previously (28). The Cx43 transfected Jeg3 cells were routinely checked for induction of Cx43 expression by immunocytochemistry following 48 h treatment with 1 μ g/ml Dox (see above).

Cell Irradiation

Confluent cell cultures were irradiated at room temperature using a Pantak X-ray machine (Pantak, East Haven, Connecticut, USA) operated at 310 kV, 10 mA, with a 2 mm Al filter (effective photon energy ∼ 90 kV), at a distance from 75 cm and a dose rate from 2.7 Gy/min. After irradiation culture flasks were returned immediately to the incubator.

Northern Blot Analysis

Total RNA was isolated from confluent cell monolayers using a Qiagen RNeasy Kit (Qiagen, Hilden, Germany). RNA samples (5 μ g or 10 μ g) were separated on a denaturing agarose gel before being transferred to a nylon membrane (Hybond-N; Amersham Biosciences, USA). Hybridization at 42°C and labeling of the specific probes with [α-³²P] dCTP by using the random primed labeling system (Amersham Biosciences, USA) were performed as described previously (28). The following probes were used: rat heart Cx43 cDNA fragment (29) and as a control glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (30). Densitometric analysis was performed using Gel Imager (Intas, Göttingen, Germany) and the GELSCAN Pro. V4.0 software (BioSciTec, Frankfurt, Germany). The expression of each signal was normalized to the according GAPDH signal.

Western Blot Analysis

Protein extracts were prepared with modified RIPA lysis buffer (50 mM Tris/HCl, 150 mM NaCl, 1% NP-40, 0.25% Na-deoxycholate, 1 mM EDTA) supplemented with EDTA free complete protease inhibitors (Roche, Penzberg, Germany).

Protein content was determined using the BCA protein assay (Perbio Science, Bonn, Germany). Protein samples (50 μ g) were separated on a 15% polyacrylamide gel and electrophoretically transferred to polyvinylidene difluoride membrane (Amersham Biosciences, USA). Membranes were blocked with 5% nonfat dried milk in Tris-buffered saline (TBS) with 0.15% Tween-20 and incubated with the primary antibody. The following primary antibodies were used: rabbit anti-Cx43, raised against residues 367–382 (1:1000; kindly provided by E. Kardami, Manitoba, Canada) (31) and mouse anti-human GAPDH antibody (1:1000, Chemicon, Hampshire, UK) for normalization of protein expression. Primary antibody binding was detected using the following secondary antibodies: anti-rabbit and anti-mouse IgG antibody conjugated to horseradish peroxidase (1:2000; Santa Cruz Biotechnologies, Heidelberg, Germany). Detection was achieved with the ECL chemiluminescence kit (Amersham Biosciences, USA) according to the protocol and densitometric analysis was performed using Gel Imager (Intas, Göttingen, Germany).

Immunofluorescence and Microscopy

Indirect immunocytochemistry on cells was performed as described previously (32). The following primary antibody was used: anti-Cx43 rabbit polyclonal antibody (1:100) (33). Donkey anti-rabbit Alexa Fluor® 488 (1:300, MoBiTech, Göttingen, Germany) was used as a secondary antibody. Finally, cells were mounted with Mowiol (Sigma, Munich, Germany) to prevent photobleaching. Photomicrographs were obtained using a confocal laser scanning microscope (LSM 510, Zeiss, Oberkochen, Germany).

Dye Transfer Assay and Analysis of Cell Coupling by Flow Cytometry

This assay is based on transfer of calcein from preloaded, donor cells to recipient, DiI-stained acceptor cells (34). To analyze the communication properties of the different cell lines, 48 h before the experiment, 1.5 million cells were seeded onto Petri dishes to obtain confluent cultures. For loading the donor cells with the green membrane-permeable dye Calcein AM (C-3100; Molecular Probes, USA) the cells were trypsinized, resuspended in 5 ml of staining solution (1 μ l Calcein AM stock solution (50 μ g Calcein AM/50 μ l DMSO) in MEM medium with 10% FCS), and agitated for 30 min at 37°C in an incubator. The cells were washed three times with medium and resuspended in culture medium. The acceptor cell population was loaded with the permanent red membrane dye DiI (V-22885, Molecular Probes, USA) by directly adding a solution of 10 μ l DiI stock solution per 1× 10⁶ cells in 4 ml growth medium (5 μ M final concentration) to the adherent cultures for 1 h. In each experiment, the ratio of acceptor and donor cells was 1:2. After an incubation time of 4 h at 37°C, these cocultures were trypsinized, resuspended in culture medium, and analyzed by flow cytometry on a FACScan flow cytometer (Beckman Coulter, Krefeld, Germany) using Cell Quest software. The ratio of communicating cells (coupling degree) was calculated as the number of DiI-stained acceptor cells with enhanced calcein fluorescence (overlayed together as yellow) in percent of the total number of acceptor cells.

The fluorescence intensity of the cells was recorded with an argon laser set up for an excitation wavelength at 488 nm. At least 10,000 events were collected for each sample. Up to three independent measurements were performed from each sample. Samples were examined by gating on the acceptor cell population, characterized by an intermediate side scatter and bright staining for DiI. For the analysis of the coupling degree of the cells after irradiation, the DiI-stained adherent acceptor cells were irradiated with 5 Gy and 15 min after irradiation the nonirradiated donor cells were plated on the acceptor cells, cocultivated for 4 h and analyzed by FACS analysis.

Statistics

Results are reported as the mean ± standard deviation of the mean (SD). Levels of significance were determined at the 0.05 level by the Student's t-test.

RESULTS

Modulation of Cx43 Expression in Mouse bEnd3 Endothelial Cells after X-ray Irradiation

The mouse endothelial cell line bEnd3 expresses high amounts of Cx43 (35). To investigate radiation-induced modulation of Cx43 transcript expression in bEnd3 cells, Northern blot analysis was performed by harvesting RNA samples from cells 15 min up to 24 h radiation with 5 Gy X-rays (Figure 1A). Already 15 min after irradiation bEnd3 cells showed a significantly elevated level of Cx43 transcripts up to 24 h (3-fold upregulation), as compared to the nonirradiated controls. Cx43 protein levels revealed a 1.8-fold upregulation in bEnd3 cells already 15 min up to 24 h after exposure to X-rays analyzed by Western blotting (Figure 1B). The Cx43 mRNA levels in bEnd3 cells after irradiation correlated to the amount of Cx43 protein. Furthermore, radiation exposure resulted in posttranslational modification of Cx43. In addition to the unphosphorylated Cx43 protein, the phosphorylated form of Cx43—indicating by a slower migrating band at 45 kDa—was upregulated (see Figure 1B). The upregulation of Cx43 protein could be confirmed by immunocytochemistry 2 h after irradiation (Figure 1C). The results showed a stronger immunolabeling of Cx43 in bEnd3 cells 2 h after irradiation (Figure 1C, right image) compared to the nonirradiated cells (Figure 1C, left image). However, in addition to the labeling of Cx43 at the cell membranes (arrows in Figure 1C), a high amount of Cx43 protein was also observed in the cytoplasm (see Figure 1C, right image).

Figure 1. Upregulation of Cx43 in bEnd3 mouse endothelial cells after X-ray irradiation. Northern blot analysis of Cx43 mRNA (A) and Western blot analysis of Cx43 protein expression (B) in bEnd3 cells exposed to 5 Gy X-rays. RNA and protein were harvested 15 min up to 24 h after irradiation. Mouse heart RNA and protein is used as a control. Densitometric analyses of three independent experiments are shown. Already 15 min after irradiation, Cx43 mRNA and unphosphorylated and phosphorylated Cx43 protein were significantly upregulated in bEnd3 cells. (C) Cx43 immunocytochemistry in bEnd3 cells before and 2 h after irradiation with 5 Gy. 2 h after irradiation bEnd3 cells showed an upregulation of Cx43 protein at the cell membranes as well as intracellularly compared to the nonirradiated cells. (D) Cell coupling in bEnd3 cells before and after irradiation measured by calcein transfer and FACS analysis. BEnd3 cells exhibited a strong mean coupling efficiency (98%) which did not significantly change after irradiation of the acceptor cells with 5 Gy X-rays and cocultured with nonirradiated donor cells (97%). Data represent means ± SD (n = 3). *P ≤ 0.05, relative expression is significantly enhanced in reference to the nonirradiated control (Co). Scale bar in (C) represents: 40 μ m.

For the measurement of gap junctional intercellular communication (GJIC) before and after irradiation two differentially stained cell populations, Calcein-labeled donor and DiI-stained acceptor cells were cocultured for 4 h. The coupling time of 4 h is needed for the formation of cell-cell membrane contacts between the two cell populations and is defined as the time leading to the strongest coupling efficiency. Shorter coculture times (30 min, 1 h, 2 h) led to markedly decreased coupling properties (data not shown). Measurements of acceptor cells with enhanced calcein fluorescence (coupled acceptor cells) were performed by flow cytometry and the ratio of communicating cells (coupling efficiency) was calculated in percent of the total number of acceptor cells.

Nonirradiated bEnd3 cells already revealed a high mean coupling degree of 98% which is close to saturation of possible cell coupling (Figure 1D). The quantification of cell coupling after irradiation between irradiated bEnd3 acceptor cells and nonirradiated donor cells did not show significant differences in coupling efficiency compared to nonirradiated cell populations. We also found no change in coupling properties after shorter coculture periods compared to the nonirradiated control cells.

Modulation of Cx43 Expression in Human EA.hy926 Hybrid Endothelial Cells after X-ray Irradiation

The human hybrid endothelial cell line EA.hy926 is characterized by high amounts of Cx43 transcript and protein, as demonstrated by Northern and Western blot analysis (Figure 2). In contrast to bEnd3 cells, EA.hy926 cells significantly downregulated Cx43 mRNA (2.4-fold) 1 h after irradiation compared to nonirradiated cells (Figure 2A). However, after 4 h, the Cx43 transcript recovered to the initial level. No obvious changes in Cx43 protein expression up to 24 h after irradiation could be observed in EA.hy926 cells after analysis of the Cx43 protein level by Western blot (Figure 2B). Cx43 immunocytochemistry 2 h after irradiation confirmed these results (Figure 2C). Nonirradiated cells as well as irradiated cells showed the same strong punctate staining pattern of Cx43 at the cell membranes and some Cx43 protein could be found in the cytoplasm, mostly in perinuclear regions. Analysis of the communication properties of the EA.hy926 cells by calcein dye transfer analysis revealed a strong mean coupling efficiency of 91% 4 h after coculturing (Figure 2D). Irradiation did not cause an obvious change in the coupling efficiency. The results show that the downregulation of Cx43 expression in EA.hy926 cells at the transcript level 1 h–4 h after irradiation could not be confirmed at the protein level and did not change coupling properties after 4 h coculture or after shorter coculture periods.

Figure 2. Modulation of Cx43 expression in human hybrid endothelial EA.hy926 cells after X-ray irradiation. Northern blot analysis of Cx43 mRNA (A) and Western blot analysis of Cx43 protein expression (B) in EA.hy926 cells exposed to 5 Gy X-rays. RNA and protein were harvested 15 min up to 24 h after irradiation. Densitometric analyses of three independent experiments are shown. Whereas the Cx43 transcript was downregulated 1 h after irradiation, the Cx43 protein level did not change. (C) Cx43 immunostaining of EA.hy926 cells before and 2 h after irradiation. The nonirradiated EA.hy926 cells showed the characteristic punctate Cx43 staining pattern at the cell membranes. 2 h after irradiation no obvious changes in Cx43 expression in EA.hy926 cells could be observed. (D) The EA.hy926 cells revealed strong cell coupling analyzed by calcein dye transfer and FACS analysis with a mean coupling efficiency of 91% which did not alter after irradiation. Data represent means ± SD (n = 3). * P ≤ 0.05, relative expression is significantly decreased in reference to the nonirradiated control (Co). Scale bar in (C) represents: 40 μ m.

Cx43 Expression in Cx43 Transfected Jeg3 Malignant Trophoblast Cells after X-ray Irradiation

The parental malignant human trophoblast Jeg3 cell line is characterized as communication-deficient and is shown to respond to X-ray exposure by an increase in activated p53, an indicator for radiation-induced cell damage (36; unpublished observations).

Jeg3 cells were stably transfected with exogenous Cx43, which can be induced by doxycycline treatment via a tetracycline-responsive inducible promoter system (28).

Induction of Cx43 mRNA and protein was tested in noncommunicating as well as communicating Jeg3 cells following 48 h induction with Dox. High amounts of Cx43 mRNA and strong immunolabeling of Cx43 protein at the cell membranes of Jeg3 Cx43 transfected cells were revealed upon induction of Cx43 with Dox compared to the uninduced cells (Figure 3A and Figure 3B). Furthermore, the induced cells revealed a significant mean coupling efficiency of 56% compared to the uninduced cells (4%) (Figure 3C).

Figure 3. Inducible Cx43 expression in Jeg3 Cx43 transfectants. (A) Northern blot analysis. The Cx43 transfected cells displayed strong induction of the exogenous Cx43 mRNA upon 48 h treatment with Dox compared to the uninduced state. (B) Cx43 immunostaining of Cx43 transfected cells revealed strong immunoreactivity in the induced culture (right, + Cx43) in contrast to the very weak staining in the uninduced cells (left, - Cx43). (C) Calcein transfer measured by FACS analysis showed a strong mean coupling efficiency in Cx43 induced Jeg3 Cx43 transfected cells (56%) in contrast to uninduced cells (4%). Data represent means ± SD (n = 3), * P ≤ 0.05, cell coupling efficiency (%) is significantly enhanced in Cx43 expressing Jeg3 cells compared to uninduced cells. Scale bar in (B) represents: 40 μ m.

Radiation neither modulates the expression of exogenous Cx43 nor induces the expression of endogenous Cx43 in the Cx43 transfected Jeg3 cells (Figure 4). Figure 4A shows the Northern blot analysis of the induced Jeg3 Cx43 transfectants 15 min up to 24 h after exposure to 5 Gy. No modulation of exogenous Cx43 was observed in Jeg3 Cx43 induced cells, but in addition no induction of endogenous Cx43 could be seen. The Cx43 nonexpressing Jeg3 cells revealed no Cx43 mRNA expression before and after irradiation (Figure 4B). Moreover, Cx43 expressing Jeg3 cells exhibited strong Cx43 immunolabeling at the membranes without alteration in Cx43 protein expression and localization 2 h after irradiation (Figure 4C).

Figure 4. Cx43 expression in Jeg3 Cx43 transfectants after X-ray irradiation. Northern blot analysis of Cx43 induced Jeg3 cells (A) and Cx43 uninduced cells (B) 15 min up to 24 h after X-ray irradiation. The Cx43 expressing cells displayed strong induction of the exogenous Cx43 mRNA upon 48 h treatment with Dox. After radiation with 5 Gy for 24 h the cells showed no modulation in Cx43 expression. (B) The uninduced Cx43 transfectants exhibited no induction of endogenous Cx43 after irradiation with 5 Gy for 24 h. (C) Cx43 immunostaining 2 h after irradiation showed no changes in Cx43 protein expression. Co = nonirradiated control. Scale bar in (C) represents: 40 μ m.

DISCUSSION

Intercellular contact has long been considered of importance in the cellular response to ionizing radiation and gap junctions have been postulated to play a critical role in radiation-induced bystander effects (37, 38, 39). It is known that Cx43-mediated gap junctional intercellular communication is involved in the bystander response observed in fibroblasts exposed to low fluences of α -particles (1, 2). Furthermore, previous reports could show that the expression of Cx43 is modulated by ionizing radiation (23, 24).

The current study shows that it is necessary to test the modulation of Cx43 expression upon irradiation before studying radiation-induced bystander effects. We found in this study that the expression of the gap junction protein Cx43 is differently modulated in the mouse endothelial cell line bEnd3 and the human hybrid endothelial cell line EA.hy926 after X-ray irradiation with 5Gy. In bEnd3 cells, Cx43 mRNA and protein are significantly upregulated after 15 min while in EA.hy926 cells only the Cx43 transcript level is downregulated after exposure to X-rays. In contrast the inducible Cx43 transfected human epithelial Jeg3 cell line showed neither an induction of the endogenous Cx43 nor a modulation of the exogenous Cx43 expression upon irradiation.

Our results are in accordance with the recent observation that Cx43 expression is highly sensitive to ionizing radiation such as α -particles and γ -rays (23, 24). It could be demonstrated that the expression of Cx43 mRNA is upregulated in human fibroblasts as well as in liver epithelial cells 4 h after exposure to α -particle doses between 1 to 24 cGy or after 6 Gy γ -rays (23). Since bEnd3 endothelial cells upregulated in addition to the nonphosphorylated Cx43 the phosphorylated Cx43 protein, the increase of Cx43 could be a result of protein de novo synthesis or a decelerated turnover of the Cx43 protein by stabilization through phosphorylation. Posttranslational modification, such as phosphorylation of connexins, is a major cellular response to environmental stress and has been implicated in the regulation of stabilization, trafficking, assembly and gating of gap junction channels (40, 41).

Our results showing strong functional communication of bEnd3 cells after irradiation correlated well with the Cx43 levels in the exposed cultures. Thus upregulation of Cx43 protein in response to ionizing radiation could lead to the formation of functional gap junction channels and to facilitation of intercellular communication. However, we found no significant increase in coupling efficiency in the X-ray exposed cultures, probably because of saturation of communication already before radiation. Another reason could be the high intracellular Cx43 protein accumulation with the consequence that the induced Cx43 proteins did not form functional channels.

In other cell systems such as the mouse epidermal cells, an increased immunolabeling of Cx43 was observed 7 days after daily irradiation with 3 Gy X-rays (21). The same results could be detected analyzing the Cx43 immunoreactivity in rat alveolar epithelial cells after X-ray irradiation (42).

In contrast, the human hybrid endothelial cell line EA.hy926 expressing high amounts of Cx43, downregulated the Cx43 transcript level 1 h after irradiation without obvious alteration in the Cx43 protein levels or the communication properties. In contrast to other observations—reporting about an increase of Cx43 expression after radiation (23)—we could demonstrate an early downregulation of Cx43 transcript in EA.hy926 endothelial cells between 1 and 4 h after irradiation which seemed to be too short to be monitored on the protein profile. We do not know whether this decrease of the Cx43 mRNA in EA.hy926 cells after irradiation is associated with a cell physiological impairment.

Thus modulation of Cx43 expression after X-ray irradiation seems to be cell specific. Our results showing a cell line specific modulation of Cx43 after exposure to 5 Gy are in agreement with other observations that the response upon UV-radiation was cell type-dependent (23, 43).

A reason for such modulation of Cx43 expression after irradiation could be the induction of transcription factors binding to the Cx43 promoter. Analysis of the promoter region of Cx43 has revealed binding sites for the redox-sensitive transcription factors nuclear κ B (44), activator protein 1 (AP-1) (45) and ATF2 (activating transcription factor 2) (46). Azzam et al. reported that the nuclear factor κ B, AP-1 and ATF2 are significantly activated in human fibroblasts exposed to low doses of α -particles (22). Glover et al. performed Cx43 promoter studies in human fibroblasts and in HeLa cells and showed that ionizing radiation activated the human Cx43 promoter in a time- and dose-dependent manner with a maximal induction 6 h after 0.5 Gy; higher doses led to a less marked increase over the same time period (24). This promoter activation was associated with an increase in Cx43 transcript and protein. Furthermore, Gloyer et al. demonstrated that low doses of radiation induce the transcriptional upregulation of Cx43 expression employing the NFAT (nuclear factor of activated T-cells) and AP-1 sites (24).

Interestingly, the Cx43 transfected malignant trophoblast Jeg3 cells revealed no regulation of exogenous Cx43. This is what we have expected because in our tet-on vector transfection system we used an artificial promoter, the tetracycline-responsive eukaryotic human minimal cytomegalovirus promoter (P_hCMV), without the above described response regulatory elements found in the native Cx43 promoter. In contrast to our observations in Jeg3 cells, Azzam et al. reported an induction of endogenous Cx43 after exposure to α -particles in a gap junction-deficient liver epithelial cell line (23).

In conclusion, our data showed a significant upregulation of Cx43 in mouse bEnd3 endothelial cells after exposure to X-rays that contrasted to a downregulation of Cx43 mRNA in the human hybrid endothelial EA.hy926 cells. The X-ray irradiation-induced modulation of Cx43 in Cx43 expressing cell lines seems thus to be cell type-dependent. The Cx43 transfected Jeg3 cells did not reveal any alteration in Cx43 expression after X-ray irradiation; therefore, they represent a suitable system for the analysis of the role of gap junctions in mediating radiation-induced bystander effects.

ACKNOWLEDGeMENTS

We would like to thank Eva Kusch and Tamara Mußfeld for excellent technical assistance and Dr. Wilfried Böcker for expert advice. We are also indebted to E. Kardami and O. Traub for providing us with Cx43 antibodies. This work was supported by grants from the European Union (EU) (Interstander, FIGH CT-2002-00218) to E.W. and G.I. and from the German Research Foundation (DFG: Wi 774/21-2) to E.W.