Neonatal mammals are known to possess a higher tolerance to hypoxia and ischemia than adult animals of the same species (Fazekas et al., 1941; Mortola, 1999; Ostadal et al., 1999). Physiological adaptations known to increase hypoxia tolerance in newborn mammals include polcythaemia, a leftward shift of the oxygen dissociation curve, reduction of body temperature and heart rate, and a deviation from body size allometry (Singer, 1999. Body size allometry, also known as Kleiber's rule (Kleiber, 1961, refers to the fact that mass-specific basal metabolic rate increases with decreasing body size, i.e., the smaller the animal, the higher its specific metabolic rate. In a recent study, we have demonstrated that deviation from body size allometry in the newborn rat is not restricted to the refers to the fact that mass-specific basal metabolic rate increases with decreasing body size, i.e., the smaller the animal, the higher its specific metabolic rate. In a recent study, we have demonstrated that deviation from body size allometry in the newborn rat is not restricted to the (Mühlfeld et al., 2005).
To approach the mechanisms that are involved in the increased neonatal tolerance to ischemia, we hypothesized that neonatal mammals are capable of using strategies known from other physiological states of increased tolerance to ischemia. Ischemic preconditioning (Schultz et al., 1997) and natural hibernation (Su, 2000) are related to the effects of endogenous opioid peptides and the opening of mitochondrial ATP-regulated potassium [K+ (ATP)] channels. Most authors have provided evidence for a cardioprotective effect of opioid receptor stimulation and opening of K+ (ATP) channels during ischemia of the adult mammalian heart (Bolling et al., 1997; Huh et al., 2001), although differences exist between species and opioid receptor subtypes (Romano et al., 2004).
Northern blot analysis and in situ hybridization studies revealed the expression of opioid peptides in the developing and adult rat heart in both cell culture and myocardial tissue (Springhorn and Claycomb, 1992; McLaughlin and Wu, 1998; McLaughlin and Allar, 1998). Binding assay studies proved the existence of κ- and δ-opioid receptors in different stages of postnatal development of rats and mice (Zimlichman et al., 1996). The existence of mitochondrial K+ (ATP) channels in neonatal cardiomyocytes has also been confirmed (Kicinska and Szewczyk, 2003).
We hypothesized that both the endogenous opioid system and mitochondrial potassium channels influence the neonatal rat heart's response to ischemia and that blockade of these systems by naloxone (nonselective opioid receptor antagonist) or 5-hydroxydecanoate [blockade of mitochondrial K+ (ATP) channels], respectively, reduces the ischemia tolerance of newborn rat hearts. Support for our hypothesis came from two studies on simulated ischemia of neonatal cardiomyocyte cell culture. Diazoxide, an opener of mitochondrial K+ (ATP) channels, protected isolated neonatal cardiomyocytes against ischemia and this protection was blocked by 5-hydroxydecanoate (Kicinska and Szewczyk, 2003). A metabolic downregulation of noncontracting neonatal mammalian myocytes subjected to hypoxia was demonstrated (Casey and Arthur, 2000).
The present study, therefore, investigated ischemic myocardium in untreated, sodium chloride-, naloxone-, and 5-hydroxydecanoate-treated neonatal rats to evaluate the degree of ischemic injury by electron microscopy. Special emphasis was placed on a sound quantification of the ischemic swelling of mitochondria and cardiomyocytes using design-based stereological methods.
MATERIALS AND METHODS
All experiments performed in this study comply with the current German laws. The investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH publications 85-23, revised 1996). The experiments were approved by the Bioethical Committee of the District of Braunschweig, Germany.
A total of 36 newborn rats (body mass, 6.85 ± 1.02 g; heart mass, 48.43 ± 11.39 mg) less than 24 hr old were assigned to a control (Con; n = 5), a sham (sham; n = 8), a placebo- (Pla; n = 7), a naloxone- (Nal; n = 8), or a 5-hydroxydecanoate-treated (HD; n = 8) group. Thirty minutes prior to sacrifice, Pla animals received an isotonic NaCl injection (10 μl/g body weight, i.p.); animals of Nal and HD were injected naloxone (10 ng/g body weight, i.p.) or 5-hydroxydecanoate (50 ng/g body weight, i.p.), respectively (needle size, 25 G). Thus, Pla, Nal, and HD animals received the same volume of injection. Animals of sham and Con did not receive any treatment but were also taken from their littermates and kept alone for 30 min. All animals were sacrificed by cervical dislocation. After a parasternal thoracotomy, cardiac arrest was instituted by perfusion of HTK solution (25 μl/min for 4 min at room temperature; Custodiol, Köhler Chemie, Germany) via the left ventricle in sham, Pla, Nal, and HD. The hearts were kept in situ and were subjected to global ischemia at room temperature. Drying was prevented by regular drops of NaCl solution. The hearts were fixed by perfusion with aldehyde fixative (1.5% glutaraldehyde, 1.5% paraformaldehyde in 0.15 M HEPES buffer) 90 min after cardiac arrest. Control hearts were immediately fixed by perfusion with aldehyde fixative after sacrifice (Table 1).
After perfusion fixation, the hearts were stored in aldehyde fixative for 12–24 hr at 4°C. The atria were removed, the ventricles weighed. Afterward, the ventricles were cut into 7–9 slices. Slices for electron microscopy were chosen by systematic random sampling and cut according to the orientator principle (Mattfeldt et al., 1990) to obtain isotropic uniform random (IUR-) tissue blocks. These were subsequently rinsed in 0.15 M HEPES buffer and 0.1 M sodium cacodylate buffer, osmicated, washed repeatedly in distilled water, and dehydrated in an ascending acetone series. Finally, the specimens were embedded in araldite (SERVA Electrophoresis, Heidelberg, Germany).
Three araldite blocks of each animal were chosen randomly and cut according to a systematic random method (Weibel, 1979). Semithin sections were stained with methylene blue for qualitative light microscopic analysis. Ultrathin sections for quantitative investigations were mounted on copper grids and stained with lead citrate and uranyl acetate.
EM analysis was performed with an EM 900 (Zeiss, Oberkochen, Germany) equipped with a digital camera and a computer image analysis system (Analysis® 3.1, Soft Imaging System, Münster, Germany) at a final magnification of 36,000×. Stereological test fields were gained systematically randomly. A test system with 18 points was projected onto each test field. Volume densities of cardiomyocytic compartments were estimated by point counting (Weibel, 1979; Schmiedl et al., 1990). The volume densities were calculated according to Equation 1 and referred to the cardiomyocyte (fi) as the reference volume:
where VV(C/R) is the volume density of a specific compartment within the reference volume, P(C) is the number of points hitting the particular compartment, and P(R) is the number of points in the reference volume. Thus, volume densities of myofibrils, VV(mf/fi), sarcoplasm, VV(sp/fi), mitochondria, VV(mi/fi), and the myocyte nucleus, VV(nuc/fi), were assessed.
From the volume densities of mitochondria, sarcoplasm, and myofibrils, the cellular edema index (CEI) was calculated according to Equation 2:
The volume-weighted mean volume of mitochondria, VV(mi), was estimated using the point sampled intercepts method (Gundersen and Jensen, 1985). The formula for this parameter is given by
where l0 is the edge-to-edge chord length of a mitochondrial profile along a line intercept passing through a sampling point, and P(mi) is the number of points of a rectangular point grid hitting mitochondrial profiles.
Body and ventricle mass are given as mean ± standard deviation. Stereological results are presented as mean (CV), with CV being the coefficient of variation, calculated by standard deviation divided by the mean. For evaluation of the ischemic injury, the results of each animal were normalized by referring them to the corresponding mean of control hearts, e.g., CEI (Pla)Norm (%) = CEI (Pla)/CEIMean(Con) × 100. Data were analyzed with the two-sided nonparametric Mann-Whitney U-test. Differences between corresponding data were considered significant at P < 0.05 and highly significant at P < 0.01.
Qualitative light and electron microscopy confirmed that all hearts except controls were arrested in diastole by HTK solution (relaxed myofibrils) and perfused properly (wide-open capillaries, only a small amount of erythrocytes within blood vessels). In control group, myocardium was well preserved with electron-dense mitochondria, high sarcoplasmic glycogen content, evenly dispersed nuclear chromatin, and well-preserved mitochondrial and sarcolemmal membranes. As a typical feature of neonatal myocardium, the content of free sarcoplasm was relatively high. The contraction state in control hearts was slightly higher than in the experimental groups due to the lack of cardioplegia.
Most of the ischemic hearts showed typical signs of ischemic damage such as large, less electron-dense mitochondria with loss of cristae, increase in sarcoplasm, loss of sarcoplasmic glycogen granules, clumping and margination of chromatin. The myofibrillar contraction state usually increased in ischemic myocardium, was widely relaxed due to the cardiac arrest. Small areas of increased contraction state were observed in all experimental groups.
The most pronounced alterations indicating ischemic injury were observed in sham animals and less pronounced in the naloxone and 5-hydroxydecanoate groups. In the placebo group, however, only small ischemic changes were found. Mitochondrial volume was highly increased in sham, HD, and Nal. Mitochondria of placebo group were only slightly swollen (Fig. 1).
Swelling of cardiomyocytes was evaluated using the cellular edema index. In sham, Nal, and HD, a significant increase in CEI was observed in comparison to control myocardium, whereas CEI was not increased in placebo group (Table 2). Compared with control, CEI was increased by 244% ± 39% in sham, 173% ± 28% in Nal, 142% ± 25% in HD, and 101% ± 24% in Pla (P < 0.05 between groups; Fig. 2).
Table 2. Summary of stereological data
Results are given as mean (CV) with CV being the coefficient of variation.
Statistically significant results (p < 0.05) are indicated in the rows below the values: a, vs. sham; b, vs. Nal, c, vs. HD; d, vs. Pla; e, vs. Con.
Mitochondrial volume and size distribution were evaluated using the volume-weighted mean volume of mitochondria. In all experimental groups, a significant increase in mitochondrial volume was observed in comparison to control myocardium (Table 2). Volume-weighted mean mitochondrial volume was increased by 514% ± 235% in sham, 341% ± 110% in Nal, 458% ± 149% in HD, and 175% ± 70% in Pla (Fig. 3). In placebo group, this parameter was remarkably smaller than in the other groups (P < 0.01). No differences were found between sham, Nal, or HD.
The major results of the present study are as follows. One, the ultrastructure of arrested newborn myocardium (sham) is highly sensitive to ischemia. The observed features of ischemic newborn myocardial damage are similar to alterations known from adult ischemic myocardium (Jennings and Reimer, 1991). Two, the ischemic swelling of mitochondria was comparably high in sham and in naloxone- and 5-hydroxydecanoate-treated animals and was very small in placebo group. Three, ischemic swelling of cardiomyocytes was most pronounced in sham animals, whereas in naloxone- and 5-hydroxydecanoate-treated groups, lower CEI values were observed. In placebo animals, hardly any cardiomyocyte swelling was present.
In neonatal rats, the intraperitoneal injection itself induced a cardioprotective effect that could partly be blocked by naloxone or 5-hydroxydecanoate, suggesting a mechanism similar to that of preconditioning. Naloxone had a similarly high effect on mitochondrial and cardiomyocyte swelling, whereas 5-hydroxydecanoate mainly affected mitochondrial volume regulation.
Evaluation of myocardial ultrastructure by quantifying different degrees of ischemic injury was applied by several authors (DiBona and Powell, 1980; Schmiedl et al., 1995). It was shown that the swelling of mitochondria and cardiomyocytes, respectively, is closely correlated with the functional status of the ischemic heart (Murry et al., 1990; Gorge et al., 1991). Although the present study does not present data on the functional properties of the tissues investigated, it presents strong parameters indicative of the functional status of the myocardium. The volume-weighted mean volume of mitochondria is an unbiased parameter determined by using the point-sampled intercepts method (Gundersen and Jensen, 1985). It combines information on mitochondrial size and size distribution and gives an estimation of mean mitochondrial volume in terms of μm3 obtained from single sections. In contrast to the previously used mitochondrial volume-to-surface ratio, it does not depend on changes in mitochondrial surface area. However, two limitations of the point-sampled intercepts method have to be taken into account with respect to the application to mitochondria. First, it is impossible to say to what extent the variation in mitochondrial size contributes to the volume-weighted mean volume unless the number-weighted volume is also estimated. The latter requires the use of the physical disector (Sterio, 1984) at the EM level, which was considered to be not efficient in the context of the present study. Second, it cannot be excluded from a single ultrathin section that different mitochondrial profiles belong to the same mitochondrion. In each group, a few physical disector sections were analyzed in a pilot study. Less than 2% of mitochondrial profiles were observed that did not represent discrete mitochondria but represented different profiles that belong to the same mitochondrion. Therefore, we assumed that each mitochondrial profile represents a single mitochondrion accepting an error of about 2%.
The CEI as a parameter for myocyte swelling has been introduced by DiBona and Powell (1980). It is based on the understanding that the total volume of myofibrils is not altered during ischemia and therefore volume increases in free sarcoplasm and mitochondria will increase the index strongly.
In the present study, all hearts (except controls) were arrested cardioplegically using HTK solution as a cardioplegic solution widely used in cardiac surgery. This approach was used because of two reasons. First, ischemia in the newborn human mainly occurs when newborns undergo surgical treatment of congenital heart diseases and often hearts are arrested during surgery. Second, naloxone was shown to decrease the frequency of arrhythmias during cardiac ischemia (Hung et al., 1998). Since arrhythmias such as fibrillation may significantly influence myocardial metabolism and thus the susceptibility of the heart to ischemia, we excluded this factor by arresting the hearts. Our data are therefore based on cardiomyocyte energy-consuming processes apart from contraction.
Controversial reports exist about the effect of ischemic preconditioning on neonatal hearts and the importance of mitochondrial K+ (ATP) channel in this event. In neonatal rats, less than 7-day-old ischemic preconditioning failed to protect the functional status (contractility, left ventricular pressure) of the isolated heart (Langendorff mode) during subsequent ischemia (Awad et al., 1998; Ostadalova et al., 2002). No evidence was found for an involvement of mitochondrial K+ (ATP) channels in ischemic preconditioning in neonatal rats (Ostadalova et al., 2002). However, immature rabbit hearts could be preconditioned and evidence was provided for a close correlation of ischemic preconditioning and activation of mitochondrial K+ (ATP) channels in neonatal rabbit hearts since the preconditioning-induced cardioprotection was completely abolished by 5-hydroxydecanoate (Baker et al., 1999). It has been speculated that ischemic pre-condtioning may be a reactivation of newborn ischemia tolerance in adult hearts (Doenst et al., 2003; Mühlfeld et al., 2005). The present study reports a reaction of the newborn rat to mild stress induced by an intraperitoneal injection that clearly shows characteristics of preconditioning. It has not been possible yet to speculate about the adequate stimulus for this reaction, but it may involve mild pain, stress, or inflammation. The large difference between placebo and sham myocardium suggests that the elevated ischemia tolerance of newborn rats depends partly on a humoral or paracrine/autocrine signaling, which in its broadest sense can be described as a response mechanism (as in this case an intraperitoneal injection) and may be related to the phenomenon of ischemic preconditioning. This interesting phenomenon was confirmed by a very recent study on mice hearts: in isolated adult mice hearts, a delayed preconditioning-like effect in terms of reduced infarct size and better ventricular function was obtained by an intraperitoneal injection of Ringer solution (Labruto et al., 2005). Our study demonstrates this effect not only in neonatal rats but also provides information about the nature of this mechanism.
In the present study, naloxone, a nonselective opioid receptor antagonist, reduced the cardioprotective effect present in placebo group as indicated by pronounced sarcoplasmic and mitochondrial swelling. Although information exists about protective effects of naloxone on different organs subjected to ischemia (Elkadi et al., 1987; Machuganska et al., 1989; Chen et al., 2001), a large amount of recent studies confirmed that opioids, either endogenously secreted during ischemic preconditioning (Schultz et al., 1997) or pharmacologically administered (Bell et al., 2000; Romano et al., 2004) before ischemia display protective effects on adult cardiomyocytes in vivo and in vitro. These effects are characterized by limitation of infarct size and improved postischemic functional recovery during ischemia (Bolling et al., 1997; Okubo et al., 2004). Although it is yet not possible to say whether this effect is related to opioid receptor antagonism or to a distinct effect of naloxone, the similarities to other physiological states of increased ischemia tolerance are noticeable.
If administered in micromolar doses, 5-hydroxydecanoic acid selectively blocks a certain type of mitochondrial potassium channel (Jarburek et al., 1998). This K+ (ATP) channel is physiologically regulated, among other features, by the cellular ATP concentration. Opening of the channel allows potassium ions to enter the mitochondrial matrix space, thus helping mitochondria to maintain functional and structural integrity (Carreira et al., 2005). Opening of these channels by pharmacological influences (e.g., diazoxide) has proved to be a powerful tool to protect the heart from ischemic injury (Garlid et al., 1997). Studies on the signaling pathways of ischemic preconditioning have shown that 5-hydrxoydecanoate, at least in part, blocks the beneficial effects of ischemic preconditioning, indicating that opening of K+ (ATP) channels is an important feature of cardioprotection induced by ischemic preconditioning (Kevelaitis et al., 1999). Furthermore, the opioid-induced cardioprotection known from adult myocardium and cardiomyocyte culture can be blocked by 5-hydroxydecanoate (Fryer et al., 1999; Huh et al., 2001). Diazoxide had a protective effect on cultured neonatal rat cardiomyocytes (Kicinska and Szewczyk, 2003). In accordance with these data and our hypothesis, 5-hydroxydecanoate induced a larger ischemic injury in newborn rat hearts than that observed in placebo group. Interestingly, this effect was much stronger with respect to mitochondrial volume regulation than to cardiomyocyte swelling.
Both naloxone and 5-hydroxydecanoate were not able to abolish the preconditioning-like effect observed in the placebo group completely. In particular, cardiomyocyte swelling was more pronounced in sham than in the other groups. This may indicate an immediate response and effect of opioids in response to the intraperitoneal injection so that the resorption of the experimental substances could only in part block the beneficial effects.
In summary, the present study provides evidence for a preconditioning-like response mechanism in the newborn rat that increases the ischemia tolerance of the heart and helps to maintain better control of cardiomyocyte and mitochondrial volume regulation. This endogenous cardioprotection is independent of myocardial contraction since only arrested hearts were used in the present study. Although more work is needed on this topic, our data indicate that the endogenous opioid system and mitochondrial ATP-regulated potassium channels are highly important in the signaling of the increased tolerance of the newborn rat heart to ischemia.
The authors thank Ms. S. Freese, Ms. H. Hühn, and Ms. S. Wienstroth for their reliable and expert technical assistance and Ms. C. Maelicke for proofreading our English. Supported by a research grant from the Deutsche Stiftung für Herzforschung (F/01/03; to H.D., J.R., and M.M.) and by a grant from the Boehringer Ingelheim Fonds (to C.M.).