Gap junctions regulate vessel diameter in chick chorioallantoic membrane vasculature by both tone‐dependent and structural mechanisms

In this study, we examined the impact of gap junction blockade on chick chorioallantoic membrane microvessels.


| INTRODUC TI ON
Gap junctions mediate intercellular communication. Composed of two hexamer transmembrane spanning hemichannels, called connexons, they allow the exchange of small molecules (<1 kDa) and ions by bridging the extracellular space of neighboring cells thereby connecting their cytoplasm. 1 Each connexon consists of six Cx. Among the more than 20 Cx isoforms Cx37 (GJA4), Cx40 (GJA5), and Cx43 (GJA1) are the ones predominantly expressed in blood vessels. [2][3][4] Several studies have shown the importance of Cxs for vascular function and development. Knockout mice lacking GJA4 and GJA5 show severe vascular dysmorphogenesis with distended blood vessels and hemorrhages. 5,6 In vitro, siRNA knockdown of GJA4, GJA5, or GJA1 in human umbilical vein endothelial cells resulted in decreased branch formation in tube formation assays. 7 The group of Duling showed that knockout of endothelial Cx43 resulted in hypotension, bradycardia, and elevated plasma levels of NO and angiotensin I and II in mice. 8 Furthermore, GJ-based communication is considered essential for the propagation of vasomotor responses along arterial vessels. [9][10][11] However, our current understanding of vascular GJs and their In previous work, we proposed a possible role for GJs-based communication in structural adaptation of vascular beds. Based on mathematical models for adaptation of vessel diameters, we predicted that microvascular networks regulated by local hemodynamic (pressure, shear stress) and metabolic stimuli cannot generate functionally adequate flow distributions unless information transfer along the wall of blood vessels (conducted response) also acts as a stimulus for structural adaptation. [17][18][19] Conversely, absence of conducted stimuli is expected to lead to functional shunting and malperfusion. 20,21 To test this hypothesis, we investigated the role of vascular Cxs in the regulation of vessel diameters in functional vascular networks of the CAM model of fertilized chicken eggs. Since conducted responses may contribute to both acute (tone-dependent) and structural regulation of diameters, we also observed vessel diameters when exposed to a vasodilator cocktail. Changes in diameters of fully dilated vessels were considered to represent structural remodeling.

| Ex ovo CAM model
The ex ovo CAM model was used as previously described. 22

| Intravital microscopy
For intravital microscopy, culture dishes were placed on a custommade, temperature and humidity-controlled microscopic stage. The microscope (Axiotech) was equipped with two objectives (2.5×, NA 0.085; 5×, NA 0.16; Zeiss, transillumination) and CAM vascular networks were visualized for up to 24 hours. Microscopic images were taken using a CCD camera (CX9000, MBF Bioscience) and processed by appropriate image analysis software (Neurolucida, Vs8; MBF Bioscience). Video sequences were recorded using a CMOS camera (Sony ICE600). Vessel diameters were measured using ImageJ Vs 1.51n software. 23,24 After treatment with Cx blockers, a fraction of blood vessels could not be visualized anymore due to diameter reduction and impaired blood flow (see Figure 2A). Their post-treatment diameter was arbitrarily set to 5 µm.

| Reverse transcription polymerase chain reaction
Total RNA was extracted from CAM samples at development day 14

| Western blotting
Western blotting was performed as described. 25 CAM tissue lysates were prepared by excision of 5 mm 2 of CAM tissue, which was washed two times with ice-cold PBS, lysed in ice-cold RIPA buffer supplemented with protease inhibitors (Roche cOmplete, Mini Protease Inhibitor Cocktail) and subsequently homogenized with a homogenizer (Minilys, Bertin Instruments). Homogenates were centrifuged at 20 000 g, and supernatant was collected for protein quantification analysis using BCA Thermo Scientific Pierce™ Protein Assay. Fifteen-twenty microgram of protein was loaded on an SDS-PAGE gel. Proteins were transferred to nitrocellulose membrane using a wet blot tank system (Bio-Rad) for 2 hours. Membranes were blocked for 1 hour at room temperature with 5% skim milk in TBS-T before incubation with primary (overnight, 4°C) and secondary (1 hour, room temperature) antibodies in TBS-T. In between and after incubation, membranes were washed three times with TBS-T at room temperature for 10 minutes. Antibody-stained proteins were visualized using the Fusion SL system from Vilber and ECL Select Western Blotting Detection Reagent (GE Healthcare).

| Antibodies
Given the high conservation of amino acid sequences of Cx isoforms in different species, we were able to use commercially available, rabbit polyclonal, primary antibodies directed against murine (Cx37), or human Cxs (Cx40, Cx43) to test for the expression of chick Cxs. 3,26 GJA5 encodes for Cx42 in chicken and Cx40 in mammals and we used two different Cx40 antibodies to analyze its expression patterns. Table 2 shows the primary antibodies used for immunoblotting and immunofluorescence and sequence identity between their respective binding regions in mice/humans and chick
One hundred microliter of each sample was incubated with 100 µL LDH assay reagent for 10 minutes at room temperature in the dark.
Release of LDH was determined by a colorimetric kit as described previously. 27 Extinction (E) was measured at 490 nm. Cytotoxicity of a specific compound was calculated as where E low control is the extinction of PBS-treated cells that were used as negative control. E high control is the extinction after treatment with 100 µL 2% (v/v) Triton X-100 which served as positive control.

| Pharmacological modulation of gap junction function
Effects of the following GJ blockers (all purchased from Sigma Aldrich) on arterial and venous vessel diameters were determined: • Carbenoxolone-the glycyrrhizin acid metabolite CBX is a non- Diameter decline upon PA treatment was more linear with about one-third of diameter reduction occurring at 1, 3, and 6 h, respectively. Due to shrinking and impaired blood flow, some blood vessels could not be visualized after Cx blockade. For each substance, at least N = 122 ART (VEN) were analyzed in n = 3 CAMs. For ART (VEN), initial diameters ranged from 11 to 247 µm (11-193 µm). Data are given as means ± SEM, * P < .05 vs PBS control was considered statistically significant sequence is completely conserved in chick Cx43, except that isoleucine (position 214 in human Cx43) is replaced by valine, which is also a nonpolar amino acid like isoleucine. GAP27 concentration used for experiments was 1000 µmol/L. Combination of GJ blocking peptides may increase their efficacy. 34 Here, however, only one peptide was used.
All GJ blockers were dissolved in PBS and applied every 2 hours (t = 0, 2, 4 hours) in drops of 25 µL to warrant a permanent coverage of the investigated CAM areas.
Additionally, we studied the effects of the GJ function stimulating agent DIP. DIP increases the concentration of cytosolic cAMP which is believed to increase GJ coupling via different pathways such as elevated Cx trafficking, changes in phosphorylation status and increased Cx expression. 35 Chick chorioallantoic membrane vasculature was exposed for

| Vascular tone of CAM vessels
In order to determine VT of arteriolar and venous CAM vessels, where D max is the maximal diameter after application of the dilatory cocktail and D is the vessel diameter before dilatation (resting diameter or diameter after CBX treatment, respectively). Data are given as mean ± SEM. Diameter changes were analyzed using Student's t test; a P-value < .05 was considered significant.
Western blotting for protein expression of the respective mRNA revealed the presence of Cx40/42 and Cx43 ( Figure 1B). Among the two antibodies used to analyze Cx40/42 expression (Table 2), only the SAB1304973 antibody targeting the N-terminal region of Cx40/42 consistently detected Cx40/42 so that it was used for further experiments.
To determine spatial distribution of Cx40/42 and Cx43 in CAM tissues, CFM was carried out and only Cx43 was detected. Cx43 was ubiquitously expressed in arterial ( Figure 1C, detail 2, lower panel) but not venous vessel walls ( Figure 1C, detail 1, lower panel). Staining with the pericyte/ smooth muscle cell marker desmin allowed tracking of arterial and venous vessels ( Figure 1C, upper panel).

| Connexin blockers cause arterial and venous vessel diameter decrease
To test whether the observed protein expression had functional relevance in intact CAM vasculature, we treated CAM networks with the non-specific Cx blockers CBX and PA as well as the Cx43 specific blocker GAP27 (Figure 2A). All inhibitors decreased vessel diameters in both ART and VEN ( Figure 2B).

| Connexin blockers do not cause unspecific effects
We wanted to exclude the possibility that the effects occurring under the application of CBX and PA are non-specific. Firstly, we carried out LDH assays to test for eventual cytotoxicity of the nonspecific Cx blockers CBX and PA. As shown in Figure 4, neither CBX nor PA had relevant cytotoxic effects on EA.hy926 endothelial cells.
Secondly, we investigated whether DIP, which stimulates Cx function, limits the effects of CBX on CAM vasculature. Figure 5A shows intravital micrographs of CAM sections treated with PBS (control), CBX, DIP, or a combination of CBX and DIP. In accordance with the data given in Figure 2, vessel diameters decreased significantly upon CBX treatment ( Figure 5B). In the presence of DIP, however, CBX-induced vessel diameter reduction was abolished suggesting that the observed effects are mediated by changes in Cx function rather than being non-specific.

| CBX-induced vessel diameter decrease is mediated by tone-dependent and toneindependent mechanisms
In order to evaluate the role of VT before and during Cx blockade, we applied a vasodilator cocktail consisting of Ach, ADO, PAP, and SNP on untreated and CBX-treated CAM networks ( Figure 6).
Vascular tone of arterial (10 ± 1%) and venous vessels (6 ± 4%) was low under resting conditions. Initially, increase of VT contributed significantly to the CBX-induced vessel diameter reduction: more than half of the diameter decrease of ART and VEN that occurred within the first 3 hours after CBX treatment could be reversed by maximal vasodilatation. After 6 hours of CBX treatment, however, vasodilatory responses were strongly reduced, indicating that diameter reduction had become structurally fixed.

| D ISCUSS I ON
It has been extensively shown that vascular GJs provide the basis We found that Cx43 is ubiquitously expressed in arterial but not venous CAM vessel walls ( Figure 1C) and that Cx blockade with the non-specific inhibitors CBX, PA as well as the specific inhibitor GAP27 led to diameter decrease in both arterial and venous vessels ( Figure 2). This effect is likely to be specific to Cx blockade since Blood vessels of vascular beds responsible for oxygen uptake (eg, pulmonary or fetoplacental beds like the CAM) constrict in response to hypoxia. [51][52][53][54] In this context, we have reported an essential role for Cxs in the mediation of hypoxic pulmonary vasoconstriction. 55 Our present observations suggest that the role of GJs in fetoplacental beds such as the CAM expands beyond mediating vasoconstrictive responses.
Firstly, GJs are responsible for transferring vasodilatory signals from capillaries to arterioles in order to maintain low-resting tone of these vessels (ART 10 ± 1%, VEN 6 ± 4%, Figure 6). This

PER S PEC TIVE
In summary, GJs are important for network homeostasis by acutely

D I SCLOS U R E
The authors declare no conflict of interest.