Assessment of the Toxicity of Hydralazine in the Rat Using an Ultrasensitive Flow-based Cardiac Troponin I Immunoassay

Abstract

The purpose of this study was to correlate the histologic changes in the heart to serum cardiac troponin I (cTnI) concentrations assayed with the Erenna Immunoassay System in Wistar rats (Crl:Wi[Han]) using the hydralazine model of cardiotoxicity. A single dose of hydralazine caused an increase of cTnI concentrations at six hours post-dose, followed by a sharp decrease at twenty-four hours and a return to baseline at forty-eight hours. The second dose of hydralazine caused a smaller magnitude increase in cTnI concentrations at six hours as compared to the first dose. Also, cTnI concentrations returned to baseline at twenty-four hours after the second dose. The increased cTnI concentrations coincided with acute myocardial necrosis at histology. However, increased cTnI concentrations in the absence of microscopic lesions were identified in several rats. As cTnI concentrations decreased, microscopic changes in the heart matured to cardiomyophagy. In conclusion, the increases in cTnI concentrations six hours after the administration of hydralazine were indicative of a myocardial damage that did not consistently have a microscopic correlate. However, the window of increased cTnI concentrations was short, and only microscopic evaluation of the heart detected the damage at twenty-four to forty-eight hours after the episode of acute myocardial necrosis.

Introduction

The measurement of serum cardiac troponin I (cTnI) and T (cTnT) concentrations is the gold standard for the diagnosis of acute myocardial necrosis in man and has gained increasing recognition for the assessment of cardiotoxicity in safety studies (Wallace et al. 2003). The value of cTnI and cTnT concentrations in the assessment of acute myocardial damage resides in the unique functional characteristics of these proteins that, along with cardiac troponin C, form a complex bound to the myofibril of striated (skeletal and cardiac) muscles. In addition to this myofibril-bound pool, there is a small cytoplasmic pool that most probably serves as a precursor pool for the cTn complex (Adamcova et al. 2005). After myocardial damage, the cytoplasmic pool is rapidly released, whereas myofibril-bound cTnI is gradually degraded through proteolysis and released as fragments and portions of the original complex (Adamcova et al. 2005; Gaze and Collinson 2008). The release of these two pools of cTn may account for the biphasic kinetics of serum cTnI concentrations noted in some studies of acute myocardial damage, with peaks at one and six hours in the rat (Mikaelian et al. 2008), and at ten to eighteen and seventy to one hundred hours in man (Yu et al. 1995). However, some other studies in man (Bertinchant et al. 1996) and the rat (Schultze et al. 2008) have identified only one peak, and a recent in vitro study suggests that cTn is released only after irreversible myocardial damage (Hessel et al. 2008).

Currently, multiple cTnI immunoassays are available and the analytical characteristics have been evaluated in preclinical animal models (Apple et al. 2008), whereas only one cTnT assay is available, thus restricting the use of cTnT in preclinical settings. Cardiac troponin I concentrations in preclinical studies often are measured on the Beckman Coulter Access 2 platform, with a lower limit of detection (LoQ) of 30 pg/mL in the rat. These assays have been able to measure cTnI concentrations in rat serum after severe acute cardiac insults; however, they do not have sufficient sensitivity to measure cTnI concentrations in healthy, resting animals or animals after mild cardiac insults. The advent of high-sensitivity cTnI assays holds promise for providing insight around this otherwise intractable biology (Schultze et al. 2008; Todd et al. 2007). In this study, we employed the Erenna Immunoassay System (Erenna, high sensitivity single molecule immunoassay; LoQ of cTnI concentrations of 0.8 pg/mL) (Schultze et al. 2008; Todd et al. 2007) to reevaluate the established approaches and models of preclinical cardiotoxicity.

High doses of hydralazine cause subendocardial myocardial necrosis (Kemi et al. 1996). The postulated mechanism is that hydralazine causes hypotension and tachycardia, leading to hypoxia and eventually to myocardial necrosis (Balazs et al. 1981). In addition, calcium overload may have a role in the pathogenesis of the lesions because the calcium channel blocker verapamil provides partial protection against hydralazine-induced cardiotoxicity (Balazs et al. 1981). The purpose of this study was to compare, in their ablity to identify cardiotoxicity after one and two doses of hydralazine at 25 mg/kg, the relative sensitivity of microscopic evaluation of the heart to a highly sensitive, single-molecule immunoassay (Erenna) for cTnI concentrations.

The results of this study indicated that serum cTnI concentrations were more sensitive than microscopic evaluation of the heart in the identification of acute myocardial necrosis. However, cTnI concentrations returned to baseline forty-eight hours and twenty-four hours after one and two doses, respectively, at which time microscopic changes of cardiomyophagy were evident. This short window of increased cTnI concentrations after cardiac damage is not different with the Erenna assay than reported with less sensitive assays.

Material and Methods

Animals

Male Wistar rats (Crl:Wi[Han]), aged eight weeks, were obtained from Charles River Laboratories (Raleigh, NC, USA) and were acclimated two weeks prior to dosing. Rats were co-housed in polycarbonate, solid-bottom cages in a controlled environment (temperature maintained at 22°C ± 2°C and humidity at 50% ± 20%) with ad libitum access to Purina Certified Rodent Diet #5002-9 (pellets) and reverse osmosis–filtered water. All experiments were conducted in accordance with the guidance of the Roche Animal Care and Use Committee. The Hoffmann-La Roche, Inc. campus is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International.

Treatment and Sample Collection

Hydralazine (Sigma-Aldrich, St. Louis, MO, USA) was prepared at 12.5 mg/mL in 0.9% NaCl in water for injection. Hydralazine was administered intravenously at 25 mg/kg at a volume of 2 mL/kg, which constituted a dose of 25 mg/kg for hydralazine. One group of rats received a single dose of hydralazine and one group received two doses (n = 12 rats/group).

Blood collection was performed no more than twice on any group of rats. Sampling time points were six, twenty-four, and forty-eight hours after a single dose of hydralazine, and six and twenty-four hours after the second dose for rats receiving two doses of hydralazine. Rats were anesthetized with isoflurane/O₂ anesthesia for blood collection from the retro-orbital sinus into serum separator tubes. The blood was allowed to clot at room temperature for at least fifteen minutes and then was centrifuged at 3,000 rpm for fifteen minutes. Serum samples were frozen at −70 °C and shipped to Singulex to be analyzed.

Necropsies were performed six, twenty-four and forty-eight hours after dosing for rats dosed once and six and twenty-four hours after the second dose for rats dosed twice. At necropsy, the anesthetized rats were exsanguinated by catheterization of the abdominal aorta and a pneumothorax was induced.

Clinical Chemistry

Samples for cTnI concentrations were analyzed using the Erenna Immunoassay System (Singulex, Alameda, CA, USA), as previously described (Schultze et al. 2008; Todd et al. 2007). The baseline reference range (95% confidence interval) established in our facility was 0.5 to 9.1 pg/mL (eight-week-old male rats, N = 56).

Microscopic Changes

The heart was fixed in 10% neutral buffered formalin overnight. It was then parallel cross-sectioned into four cross-sections 2 to 3 mm apart, processed to slides, and stained with hematoxylin-eosin and phospho-tungstic acid hematoxylin.

Statistical Analysis

Prior to analysis, the data were screened for normality and the presence of outliers. No outliers were detected, and the data conformed to the assumptions of normality; therefore, no transformations were applied. Group differences in cTnI concentrations were examined using the general linear model procedure of SAS 9.1 (SAS Institute, Cary, NC, USA) followed by the appropriate treatment contrasts to detect specific group differences.

Results

Cardiac troponin I concentrations in vehicle-treated rats were 4.3 ± 2.7 pg/mL (mean ± standard deviation). In the vehicle-treated group, three of twenty-four samples were above the baseline 95% confidence interval, yielding concentrations of 9.2, 10.0, and 11.0 pg/mL (Figure 1 and Table 1). There was a trend toward a decrease of cTnI concentrations over time in the vehicle-treated rats after a single dose of saline, and statistical significance was found between the six-hour and the twenty-four- and forty-eight-hour time points (p < 0.04 and p < 0.001, respectively), and the six-hour and thirty-hour time points (p < 0.03).

After a single dose of hydralazine, cTnI concentrations in all (eighteen of eighteen) rats were dramatically increased above the baseline range, with a large variability in responses (121.4 to 4094.9 pg/mL) at six hours. By twenty-four hours, cTnI concentrations had decreased (16.6 ±11.2 pg/mL) but still remained above the baseline range in ten of twelve rats. By forty-eight hours, cTnI concentrations had returned to baseline in all but one rat, which had cTnI concentrations marginally above the baseline range (11.3 pg/mL). Consistent with the distribution reported by others (Balazs et al. 1981; Kemi et al. 1996), microscopic changes were centered on the subendocardial areas of the apical portions of the left ventricle. At six hours, there were a few small foci of acute myocardial necrosis, characterized by hypereosinophilia of myocardial fibers with pyknosis or karyolysis. This change was present in three of six rats, which correlated with the rats that had the highest cTnI concentrations at this time point. There were no microscopic changes in the three other rats. At twenty-four and forty-eight hours, the cytoplasm of the necrotic cardiomyocytes had been replaced by macrophages and globular acidophilic debris, which is consistent with the diagnosis of cardiomyophagy.

After two doses of hydralazine twenty-four hours apart, cTnI concentrations were increased in eleven of twelve samples six hours after the second dose. These levels were approximately fivefold lower than those found after a single dose. At twenty-four hours after the second dose, cTnI concentrations returned to baseline in six of six rats. Upon microscopic evaluation of the heart, the changes six hours after the second dose consisted predominantly of cardiomyophagy in six of six rats, with lesser acute myocardial necrosis in four of six rats. At six hours after the second dose, cardiomyophagy was interpreted as the result of the first dose, whereas acute myocardial necrosis was interpreted to be the result of the second dose. Twenty-four hours after the second dose, the only microscopic change was cardiomyophagy in four of six rats.

Discussion

This study confirmed that cTnI concentrations are a more sensitive and accessible marker than microscopic evaluation of the heart to identify acute myocardial necrosis. As reported with less sensitive methods (Bertinchant et al. 2000; Mikaelian et al. 2008), cTnI concentrations peaked early and returned to baseline as microscopic changes of cardiomyophagy developed. However, because of the sensitivity of the measurements, we were able to accurately detail the kinetics of cTnI concentrations after cardiac insult.

At the six-hour time point, all hydralazine-treated rats had increased cTnI concentrations, but only half of them had microscopic evidence of myocardial damage. Thus, cTnI concentrations are more sensitive than microscopic evaluation of the heart in the early stages of myocardial damage. The lower sensitivity of microscopic evaluation of the heart as compared to cTnI concentrations may be the result of the limited representativeness of the histologic planes of section, although more were examined in this study than in conventional toxicologic studies. The representativeness of histologic sections is in question especially when the lesions are few and small, which they were in this study.

Increased concentrations of cTnI in the absence of a histologic correlate may result from the release of the cytoplasmic pool of cTnI that occurs during reversible myocardial damage (Adamcova et al. 2005; Gaze and Collinson 2008; Patane, Marte, and Di Bella 2007). Cardiac troponin I degradation products released after irreversible myocardial damage are differentiated from undegraded cTnI released after reversible myocardial damage by Western blot analysis. Evaluation of cTnI fraction concentrations was not performed in this study. The significance of reversible and irreversible myocardial damage needs to be further understood in the context of overall cardiotoxicity.

After hydralazine insult, cTnI concentrations declined rapidly, reverting to baseline concentrations at forty-eight hours after a single dose and at twenty-four hours after two doses of hydralazine, whereas microscopic changes were present in most rats. Possible causes for the rapid decrease of serum cTnI concentrations after their peak include cTnI depletion from the damaged myocardium and its proteolysis into degradation products undetectable by the antibodies of the assay (Hessel et al. 2008; Katrukha et al. 1998). Thus, microscopic evaluation of the heart remains the marker of choice after the relatively short temporal window during which cTnI increased concentrations occurs.

The second dose of hydralazine caused less myocardial damage than the first one, as assessed microscopically and by cTnI concentrations. This difference between the first and second doses was also observed in an isoproterenol rat model, and was attributed to internalization of the β-adrenergic receptor following the first dose (Feng et al. 2005). The hypothesis of internalization of the β-adrenergic receptor is contradicted by this study because hydralazine is not a β-adrenergic agonist. Rather, the milder response of the heart to the second dose of hydralazine as compared to the first dose may be the result of preconditioning, a mechanism by which an acute episode of hypoxia partially protects the heart from a subsequent episode of hypoxia. Alternatively, the areas of the myocardium most sensitive to hypoxia may have undergone necrosis as the result of the first dose of hydralazine, and therefore may no longer be available for necrosis following the second dose.

In this study, cTnI concentrations in vehicle-treated rats were 4.3 ± 2.7 pg/mL, with a range of 1 to 11 pg/mL. These concentrations are below those reported in an earlier study using the same method, with a mean of 9.5 pg/mL and a range of 9 to 20 ng/mL (Schultze et al. 2008). The use of Crl:WI(Han) rats in this study and F344/NHsd rats in the earlier study may account for this difference. In addition to possible differences in concentrations from different strains of rats, the effect of housing, age, and sampling techniques has yet to be sufficiently characterized. Also, cTnI concentrations decreased over time after a single administration of saline. Possible explanations for this observation may be a decrease in the level of stress or the effects of the intravenous injection of saline itself.

In conclusion, serum cTnI concentrations were more sensitive than microscopic evaluation of the heart to identify acute myocardial damage. However, once past the acute stage, myocardial damage was identified only by microscopic evaluation of the heart.