| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Zentrum Anaesthesiologie, Abteilung Anaesthesiologische Forschung, and 2Zentrum Anatomie, Abteilung Elektronenmikroskopie, Universität Göttingen, Göttingen, Germany
Submitted 7 November 2005 ; accepted in final form 9 March 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
gap junctions; immunohistochemistry; nonjunctional connexin
Cell-cell interactions via connexins have been found to play a significant role in IP (6, 22) and contribute to reduction not only of infarct size but also of susceptibility to arrhythmias. The principal role of connexins in IP is well known, especially from the finding that IP cannot be elicited in heterozygous connexin43 (Cx43)-deficient mice (23). The specific function of connexins in this condition, however, is incompletely understood. It is even unclear whether the beneficial effects of IP result from a stimulated opening or closure of connexin channels. Ischemia per se initiates closure of these channels (1). It has been assumed that reduction of infarct size is the result of the impeded spread of injury from the primarily lesioned cells toward neighboring, still viable, myocytes as a result of an accelerated closure of connexins by IP (5, 7, 15). In contrast, the beneficial effects of IP on arrhythmogenesis have been explained by just the opposite mechanism: an improved intercellular communication via connexins in preconditioned hearts, which results in facilitated electrical impulse transmission and, thus, reduced occurrence of pathological rhythms (2, 10, 24).
Considering the obviously heterogeneous distribution of myocardial necrosis within the area at risk (25), the following question arose: Would spatial changes in connexin distributions explain these seemingly contradictory results? It appeared conceivable that changes in the localization of the connexins within the area at risk might explain alterations of solute exchange and electrical conductivity. To test this hypothesis, we performed experiments in anesthetized rats subjected to regional myocardial ischemia and studied the localization of Cx43 in the myocardium by immunofluorescence. We detected changes in Cx43 distribution that might offer a new and different understanding of the role of connexins in IP-induced changes in infarct size and cellular excitability.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Male Wistar rats (2 mo of age, 200–220 g body wt) were anesthetized with thiobutabarbital (120 mg/kg ip) and continuous inhalation of 70% N2O-30% O2. Catheters were placed in the left femoral artery for blood pressure (Statham P23Db and SP 1200) and heart rate (beat-to-beat counter) monitoring and in the left femoral vein for drug administration. After additional intravenous application of barbiturate under control of arterial blood pressure, the trachea was carefully accessed, and a tube was inserted for artificial ventilation with a respirator (model 681, Harvard; tidal volume 3.5 ml, respiration rate 40/min). The thoracic cavity was opened, and a catheter (Statham P23Db, SP1200) was inserted into the left atrium via the left atrial appendage for left atrial pressure monitoring.
For coronary occlusion, a 7-0 thread was placed around the descending branch of the left coronary artery, and the free ends of the thread were passed into a polyethylene tube. The thread was carefully drawn along the tube (with the artery protected from lesions by a small plastic rod) to cause a reversible occlusion of the coronary artery. At regular intervals, the blood gases were controlled (AVL 990 automatic blood gas system) and the ventilation was readjusted, if necessary.
For IP, coronary occlusion was induced three times for 5-min periods. Each of these occlusions was followed by 10 min of reperfusion and, with the exception of group 1, a subsequent 15- or 45-min extended occlusion. In part of the experiments, the prolonged ischemia was followed by 15 min of reperfusion. The following experimental conditions were studied: IP alone (group 1), 15 min of coronary occlusion (group 2), IP followed by 15 min of coronary occlusion (group 3), 45 min of coronary occlusion (group 4), IP followed by 45 min of coronary occlusion (group 5), 45 min of coronary occlusion followed by 15 min of reperfusion (group 6), and IP followed by 45 min of coronary occlusion and 15 min of reperfusion (group 7).
In each group, six experiments were performed. At the end of each experiment, the coronary artery (if in a state of reperfusion) was reoccluded, and fluorescein isothiocyanate-albumin (FITC, 0.2 ml of a 5% solution) was injected intravenously to label the non-ischemia-exposed area. After 0.5 min, the heart was clamped at its base, excised, and then transferred to –120°C isopentane. Frozen 5-µm-thick sections on the same plane as the center of the infarct were prepared as pairs of adjacent slices. One section was freeze-dried for observation of FITC and NADH by fluorescence microscopy (25). The other section was used for immunohistochemical detection of Cx43. The latter was placed on a precooled slide and allowed to adhere to the glass through gentle warming. After the section was returned to room temperature, it was floated with chloroform-diethyl ether (1:1) for 
10 s and air-dried for a few minutes. After the slides were floated in PBS for 30 min, the primary antibody anti-Cx43, a monoclonal antibody developed in mice (mouse anti-Cx43; catalog no. MAB 3068, Chemicon), was applied. In addition, experiments were performed for purposes of control with polyclonal rabbit anti-Cx43 antibody (catalog no. 71-0700, Zymed Laboratories) and a monoclonal mouse anti-Cx43 antibody (catalog no. 252-270, Biotrend, Cologne, Germany). The sections were exposed to the primary antibody for 2 h at a final concentration of 20 µg/ml. After three 10-min washes in PBS, the secondary antibodies, anti-mouse IgG conjugated with tetramethylrhodamine isothiocyanate (catalog no. T 5393, Sigma) and anti-rabbit IgG conjugated with tetramethylrhodamine isothiocyanate (catalog no. T 6778, Sigma), respectively, were applied. After 15 min of incubation, the sections were washed in PBS (30 min), transferred to pure ethanol, and embedded in a synthetic resin (Roti Histokitt).
For detection of any unspecific binding of the immunoglobulins, IgG from mouse serum (20 µg/ml; catalog no. I 5381, Sigma), instead of the above-named primary antibodies, was used. As a secondary antibody, the above-mentioned anti-mouse IgG was applied.
Evaluation of Histological Sections
Determination of area at risk and infarct size.
The extent of the area at risk and the size of the infarct, i.e., zones devoid of FITC labeling and NADH fluorescence, respectively, were determined in the freeze-dried sections. The sections were systematically scanned by a computer-controlled motor drive, and at each point of observation (1,600 points per section), the presence of FITC-labeled capillaries within a 60-µm-diameter central field was determined. Then the motor drive relocated each point and it was determined whether the loss of NADH autofluorescence, which was considered a sign of irreversible damage, had occurred in the myocyte lying at the center of the viewing field. With this information, it was possible to calculate the percentage of infarction within the area at risk (25). Because the extent of damaged tissue becomes relevant only after an extended period of ischemia, infarct size was determined only in groups 4–7.
Evaluation of anti-Cx43 immunofluorescence. An eight-bit charge-coupled device camera (model CF8/1DXC, Kappa, Gleichen, Germany) attached to a microscope with a x25 lens was used to produce a set of six images that were randomly distributed across area exposed to ischemia and six images from the normally perfused area. Each image covered a natural area of 188 x 145 µm2 (752 x 582 pixels). Because of differences in immunofluorescence intensity within the different zones of the myocardium, it was necessary to adjust the duration of light exposure to the respective conditions to avoid overexposure of the structures being studied.
For evaluation purposes, the intensity of anti-Cx43 staining in each image was determined within an area of the cells' interiors, their gap junctions, and their free cellular margins (Fig. 1). The respective ratios of these data, i.e., light intensity in the free cellular wall to that in the cellular interior and light intensity in the gap junction to that in the cellular interior, were determined in 30 and 15 cells, respectively, per image. The mean values of these data were calculated for each image. From these data, mean values for the respective zones were determined for each experiment; these values were then used in the statistical evaluation of intra- and intergroup differences. To ensure that any differences in the reference parameter, i.e., intracellular fluorescence, between the normal and the ischemia-exposed zones were not overlooked, additional sets of three images each were made: one set from ischemia-exposed tissue and one set from non-ischemia-exposed tissue per experiment. Identical lighting conditions were maintained over the entire heart section, even though this led to a partial overexposure of gap junctional fluorescence. From the resulting images, it was possible to compare intracellular fluorescence intensities under the various conditions. For image analysis, the SigmaScan program was used. All quantitative measures were performed on unprocessed images, with the exception of Figs. 1–3. To demonstrate the immunofluorescence in the cellular margins, a gamma correction of the images was necessary to avoid light scattering from gap junctional fluorescence.
|
|
For comparison of paired data (intraindividual comparison of immunofluorescence intensities within the ischemia- and non-ischemia-exposed zones), Wilcoxon's matched-pairs signed-rank test was used. When two independent samples were compared (comparison of infarct sizes in the different groups), the Wilcoxon-Mann-Whitney U-test was applied. Comparisons involving more than two groups were analyzed by one-way ANOVA (intergroup comparison of immunofluorescence intensities in the ischemia-exposed zones). Post hoc tests were performed utilizing Bonferroni's adjustment. It was possible to apply ANOVA, because the data to be compared did not differ significantly from normal distributions (Kolmogoroff-Smirnoff test). Values are means ± SE. SigmaStat was used to run the statistical analysis. Differences were considered significant if P < 0.05.
All experiments were approved by the responsible federal authority and conform to the National Institutes of Health (NIH) "Guide for the Care and Use of Laboratory Animals" (NIH Publication No. 85-23, revised 1996).
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In our study of the effects of IP on connexin distribution (group 1), we observed a relatively homogeneous distribution of gap junctional Cx43 immunofluorescence within the area at risk similar to that in the non-ischemia-exposed zones. The use of a higher optical resolution, however, revealed a specific difference between the areas. In addition to regular immunostaining at the intercalated disks and the lateral, spotlike intercellular contacts, labeling was discernable also along the entire margins of the myocytes (Fig. 2). The optical resolution of the microscopic images allowed us to conclude that this immunostaining resulted from labeling of the plasma membrane outside gap junctions, rather than staining of the extracellular space (see Fig. 3B). Although the intensity of fluorescence was generally lower outside than inside the gap junctions, the intensity was sufficient to allow viewing of the individual myocytes separately from one another. In non-ischemia-exposed myocardium, such a distinction is impossible (Fig. 2 and Fig. 3, A and B). The ratio of marginal (outside gap junctions) to intracellular fluorescence intensities (1.22 ± 0.04) in the area at risk was significantly different from that of the non-ischemia-exposed area (1.04 ± 0.01). Gap junctional fluorescence intensities were nearly identical in both areas (Table 1).
|
|
At the end of a continuous 15-min period of ischemia, as in group 1, Cx43 immunofluorescence could be seen in the gap junctions as well as at the cellular margins (group 2). Whereas staining was relatively uniform along the periphery of the myocytes in group 1, labeling of the cellular membranes was concentrated near the intercalated disks in group 2; however, it decreased as the distance to the disks increased. The immunofluorescence appeared to have spread out of the gap junctions without having reached the entire circumference of the cell (Fig. 3C). The mean ratio of light intensity in the free cellular margin to that in the myocyte interior was 1.16 ± 0.03 (ratio in non-ischemia-exposed areas = 1.00 ± 0.00). There was no significant difference in the intensity of gap junctional fluorescence between the ischemia- and non-ischemia-exposed zones (Table 1).
Compared with group 2, even without quantification, it became evident that Cx43 immunofluorescence within the membranes of the myocytes was more intense when IP preceded ischemia (group 3; Fig. 3D). The ratio of fluorescence in the free cellular membrane to that in the interior of the cell was significantly higher (1.31 ± 0.03) in group 3 than in group 2 (Fig. 4). In group 3, the intensity of fluorescence in the gap junctions was significantly lower in the ischemia-exposed zone than in the nonoccluded area (Table 1).
|
When the period of sustained ischemia was prolonged to 45 min without IP (group 4), the changes in Cx43 membrane fluorescence were accentuated (ratio of marginal to interior fluorescence intensity = 1.30 ± 0.03). As observed in group 2, the intensity of staining was highest near the gap junctions and subsided toward the gap junction-free sections of the cellular margins. In the gap junctions themselves, the overall fluorescence intensity was clearly lower in the ischemia- than in the non-ischemia-exposed areas (Table 1).
When IP preceded 45 min of occlusion (group 5), a greater intensity of Cx43 fluorescence in the cellular margins became apparent. The fluorescence could be observed over nearly the entire circumference of the myocytes, with a slight accentuation in the proximity of the gap junctions. The ratio of Cx43 immunofluorescence in the free membrane to that in the interior of the cell (1.46 ± 0.03) was significantly higher in group 5 than in group 4 (Fig. 4). Again, the gap junctional fluorescence was weaker in the ischemia-exposed zone than in the nonoccluded area (Table 1).
Effects of 45 Minutes of Ischemia Without IP or Preceded by IP Followed by Reperfusion on Cx43 Distribution
When prolonged ischemia was followed by reperfusion (group 6), immunofluorescence of Cx43 in the free cellular margins could also be detected. As a whole, the effect was less intense, however, and more heterogeneously distributed across the area at risk. In some regions, myocytes appeared to be almost enclosed by Cx43 fluorescence; in others, no labeling could be detected. As a whole, the ratio of fluorescence intensity in the free cellular wall to that in the interior of the cells was 1.24 ± 0.05. The mean fluorescence in the gap junctions revealed a ratio that was significantly lower than that of the non-ischemia-exposed areas.
The gross appearance of the sections from IP-exposed myocardium (group 7) was not significantly different from that of the non-IP-exposed group. In the ischemia-exposed zone, the ratio of fluorescence in the free wall to that in the cellular interior (1.33 ± 0.05) was similar to that in the preceding experiments; the gap junctional fluorescence was also similar in both groups (Fig. 4, Table 1).
Methodological Controls
With respect to the stability of basic cardiovascular parameters, regional myocardial ischemia led to only moderate reductions in arterial blood pressure and increases in left atrial pressure. No specific changes in heart rate due to coronary occlusion were found (Table 2).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Discussion of Methods
For induction of IP, the hearts were subjected to three short periods of regional ischemia before prolonged occlusion of the coronary artery. This measure led to the expected reduction of infarct size and, thus, supports the validity of the present experimental model. The question of methodological artifacts is especially important with regard to the observed changes in connexin distribution. Faulty immunostaining is rather unlikely, because the specific pattern was restricted to the ischemic zone of the heart; the appearance of the nonoccluded areas was entirely normal (Fig. 3). The transition from a regular pattern to the pathological distribution correlated exactly with the transition from the nonoccluded to the ischemia-exposed tissue. A similar difference became evident within the area at risk when we compared the small subendocardial and subepicardial layers of the heart (constantly supplied from the left ventricular lumen and the environmental air, respectively) with the interior of the myocardial wall, the supply of which had been completely cut off (Fig. 2).
Different charges of antibodies directed against the same antigen may show some variation in tissue binding because of differences in the Fc portion of the antibody's constant region. Binding to unrelated determinants may result (21). Each of the additional antibodies displayed binding to the gap junction-free regions of the myocyte plasma membrane in the ischemia-exposed myocardium. A further argument against a role of artifacts, e.g., due to unspecific binding of immunoglobulins via their constant regions, is the finding that mouse immunoglobulin per se did not cause binding to any region.
Quantification of the changes in Cx43 immunofluorescence proved to be a difficult problem. We decided to determine the intensity within three small, well-defined measuring fields: the cellular interior, the cellular free membrane, and the gap junctions. Bias due to subjective selection of cells was minimized through the use of marker lines and the maintenance of a preferred angle from the midpoint of the cell to the measuring field in the cellular margin.
Because the latter evaluation was based on comparisons of fluorescence intensities between different cellular structures, absolute differences between ischemia- and non-ischemia-exposed tissue could have been overlooked. For this reason, we also focused on the detection of differences in light intensity within the myocyte interior in both types of tissue. We found no increase in intracellular fluorescence intensity under any ischemic condition (Table 1); however, we did find a modest, but nonsignificant, decrease in the ischemia-exposed zones.
Discussion of Results
The results of these experiments indicate an altered distribution of Cx43, which is displayed as a spreading of its natural positions in gap junctions. Labeling in the free cellular margin was highest near the gap junctions and lowest at greatest distances from the gap junctions. In the literature, such a process has seldom been described. Although it is well known that the concentration of connexins in ischemic myocardium generally decreases (9), there is only marginal information concerning Cx43 in the free plasma membrane. It has been observed that newly synthesized connexons are first delivered to nonjunctional areas of the plasma membrane and then recruited by the gap junctional plaques (4, 12). To our knowledge, the inverse process involving the spread of connexins from gap junctions to the free plasma membrane has only been described by de Mazière and Scheuermann (3) in experiments on isolated hearts subjected to hypoxic conditions. In their electron-microscopic images, they found evidence of a dissemination of connexons into the sarcolemma, supposedly due to disintegration of gap junctional plaques.
The significance of the present observations is the correlation to the effects of IP. Although it is principally impossible to deduce a causal link between two events that develop in parallel, i.e., IP-induced changes in connexin distribution, on the one hand, and reduction in ischemic damage, on the other, several arguments favor such a link. The existence of connexins as a prerequisite for IP-induced protection has clearly been demonstrated through the use of heterozygous Cx43-deficient mice (23). It is, however, difficult to understand the role of Cx43 in this process, because infarct-restricting and antiarrhythmic effects of IP cannot be explained through general changes in intercellular communication (see the introduction). It is interesting to note that effects of connexins unrelated to cell-cell interactions, e.g., changes in volume homeostasis, have been increasingly considered (19, 22).
The present observation of Cx43 outside the direct intercellular contacts also implies that effects of Cx43 other than those induced by changes in gap junctional communication must play a role in cellular survival, especially in IP. Some data indeed point to a protective function of connexins or connexons, the uncoupled membrane proteins. In isolated, noncommunicating myocytes under stressing conditions, a state of IP could be simulated only, however, when the cells were able to express Cx43, not when the cells had been isolated from heterozygous Cx43-deficient animals (13). Also, in nonmyocyte cells or tissues, it has been found that connexins are involved in improving the survival rates, as in the antiapoptotic action of biphosphonates in bone tissue (11, 20) and in the protection of ischemia-exposed neuronal tissues (17, 18). When cultured astrocytes were forced to express and incorporate increased amounts of Cx43 in their plasma membrane, they were significantly more resistant than untreated control cells to the effects of metabolic inhibition and other forms of cellular injury. Even when growing nonconfluently or when expressing mutated, nonpermeant channels, the protective effect of the connexons in their cellular membranes was identifiable (14). Although such observations cannot be explained by a specific function of extrajunctional connexons, they do underscore the fact that these proteins are involved in processes that contribute to survival of cells under critical conditions.
We propose the following working hypothesis: IP initiates a process of expanded occurrence of Cx43 in the free plasma membrane of the myocytes, outside their natural position in gap junctions. When the phase of prolonged ischemia begins, in preconditioned myocytes, in contrast to nonpreconditioned cells, Cx43 has already accumulated in the plasma membrane, and it appears to be this condition that improves the cell's chance of surviving the following period of prolonged ischemia.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. H.-C. Lin, N. Lou, N. Kang, T. Takano, F. Hu, X. Han, Q. Xu, D. Lovatt, A. Torres, K. Willecke, et al. A Central Role of Connexin 43 in Hypoxic Preconditioning J. Neurosci., January 16, 2008; 28(3): 681 - 695. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Shen, M.-J. Lin, A. Yaradanakul, V. Lariccia, J. A. Hill, and D. W. Hilgemann Dual control of cardiac Na+ Ca2+ exchange by PIP2: analysis of the surface membrane fraction by extracellular cysteine PEGylation J. Physiol., August 1, 2007; 582(3): 1011 - 1026. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
