Critical Pathways in Heart Function: Bis(2-chloroethoxy)methane-Induced Heart Gene Transcript Change in F344 Rats

Abstract

Gene transcript changes after exposure to the heart toxin, bis(2-chloroethoxy)methane (CEM), were analyzed to elucidate mechanisms in cardiotoxicity and recovery. CEM was administered to 5-week-old male F344/N rats at 0, 200, 400, or 600 mg/kg by dermal exposure, 5 days per week, for a total of 12 doses by study day 16. Heart toxicity occurred after 2 days of dosing in all 3 regions of the heart (atrium, ventricle, interventricular septum) and was characterized by myofiber vacuolation, necrosis, mononuclear-cell infiltration, and atrial thrombosis. Ultrastructural analysis revealed that the primary site of damage was the mitochondrion. By day 5, even though dosing was continued, the toxic lesions in the heart began to resolve, and by study day 16, the heart appeared histologically normal. RNA was extracted from whole hearts after 2 or 5 days of CEM dosing. After a screen for transcript change by microarray analysis, dose-response trends for selected transcripts were analyzed by qRT-PCR. The selected transcripts code for proteins involved in energy production, control of calcium levels, and maintenance of heart function. The down-regulation of ATP subunit transcripts (Atp5j, ATP5k), which reside in the mitochondrial membranes, indicated a decrease in energy supply at day 2 and day 5. This was accompanied by down-regulation of transcripts involved in high-energy consumption processes such as membrane transport and ion channel transcripts (e.g., abc1a, kcnj12). The up-regulation of transcripts encoding for temperature regulation and calcium binding proteins (ucp1 and calb3) only at the 2 low exposure levels, suggest that these adaptive processes cannot occur in association with severe cardiotoxicity as seen in hearts at the high exposure level. Transcript expression changes occurred within 2 days of CEM exposure, and were dose- and time-dependent. The heart transcript changes suggest that CEM cardiotoxicity activates protective processes associated energy conservation and maintenance of heart function.

Introduction

Environmental chemical exposure may contribute to heart disease. Chemicals that are metabolized to thiodiglycolic acid have been associated with cardiotoxicity in rodents and humans including bis(2-chloroethoxy)methane (Dunnick et al., 2004a, 2004b) (CEM), Ifosfamide (Visarius et al., 1998), monochloroacetic acid (National Toxicology Program, 1992), chloroacetaldehyde (Joqueviel et al., 1997), tri-chloroethane (Yllner, 1971), trichloroethylene (Anderson et al., 1987), 1,1-dichloroethylene (Anderson et al., 1987), cyclophosphamide (Joqueviel et al., 1997), vinylidene chloride (Jones and Hathway, 1978), and vinyl chloride (Green and Hathway, 1975). Fatty acids are a major source of energy in the heart (Ala-Rami et al., 2005), and because thiodiglycolic acid interferes with fatty acid metabolism (Visarius et al., 1998), CEM mitochondria damage may be due in part to depletion of nutrients (Edinger and Thompson, 2004; Lum et al., 2005).

In this study we characterized gene transcript expression patterns in the heart after CEM exposures in the rat. CEM is an organic compound used in the production of polysulfide polymers for sealant applications. CEM cardiotoxicity is characterized by cytoplasmic vacuolation of myocytes, necrosis, and inflammation (Dunnick et al., 2004a, 2004b). Thiodiglycolic acid is a metabolic product of several chemicals that target the heart in humans and/or rodents (Hofmann et al., 1991), and, thus, our model of CEM-induced cardiotoxicity also serves to characterize heart damage from other environmental chemicals. Because CEM causes heart mitochondria damage we characterize heart gene transcript expression to test the hypothesis that to combat heart failure, energy demand must equal energy supply (Hochacka et al., 1996; Katz, 1991, 1998). In addition, we characterize transcript changes involved with maintenance of heart function.

Methods and Materials

Chemical and Animal Exposures

Bis(2-chloroethoxy)methane (CAS No. 111-91-1; lot B004160277; Figure 1) (Karl Industries, Aurora, OH) (Figure 1) was 98.5% pure (Dunnick et al., 2004a, 2004b). Solutions of CEM were prepared in 95% ethanol for daily dermal administration to male F344 rats (Taconic Laboratories, Germantown, NY), 5 days per week, excluding weekends, for 2 weeks plus 2 consecutive dosages before sacrifice on study-day 16. Animals were placed on study at 5 weeks of age, and received a total of 12 CEM doses. Fur from the site of application was clipped weekly. Stock solutions prepared at concentrations of 0, 400, 800, and 1200 mg/ml were stored in amber glass bottles at room temperature. The administrations were applied to the skin of the male rats at 0.5-ml/kg body to deliver doses of 0, 200, 400, or 600 mg/kg body weight. All dose formulations were determined to be within ±10% of target concentrations. Approximately 45% of a dermal dose of CEM is adsorbed (NIEHS Contract NO1-ES-75407 2002). Male F344/N rats (Taconic Laboratories, Germantown, NY) were placed on study at 5 weeks of age and housed 1 per cage in polycarbonate cages in rooms maintained at temperatures between 69 and 75°F with 35–65% relative humidity and a 12-hour light/dark cycle. Control and treated groups received irradiated NTP-2000 diet (Zeigler Brothers, Gardners, PA) ad libitum. Hearts from six male rats/dose/per day (day 2 or day 5) were used for RNA extraction.

RNA Extraction Methods

RNA was extracted from hearts of 6 control rats, and from 6 rats from each treatment group (200, 400, or 600 mg/kg) at day 2 and day 5 just 1 hour after dosing. Animals designated for heart RNA extraction were anesthetized with CO₂/O₂ on study days 2 and 5, exanguinated, and their hearts infused with RNAlater. Total cardiac RNA was isolated from hearts using the QIAGEN Rneasy kit (QIAGEN, Valencia, CA). The RNA was quantified through optical density measurements and agarose gel electrophoresis, and kept frozen at −70°C.

Microarray Analysis

Gene transcript analysis was conducted using Agilent Rat Oligo arrays (Agilent Technologies, Palo Alto, CA). Total RNA was amplified using the Agilent Low RNA Input Fluorescent Linear Amplification Kit protocol <http://www.chem.agilent.com>. Starting with 1 ug of total RNA, Cy3 or Cy5 labeled cRNA was produced according to manufacturer’s protocol. For each 2-color comparison, 750 ng of each Cy3 and Cy5 labeled cRNAs were mixed and fragmented using the Agilent In Situ Hybridization Kit protocol. Hybridizations were performed for 16 hours in a rotating hybridization oven using the Agilent 60-mer oligo microarray processing protocol. Slides were washed as indicated in this protocol and then scanned with an Agilent Scanner. Data was obtained using the Agilent Feature Extraction software (v7.1), using defaults for all parameters.

RNA extracted from three 400 mg/kg male rat hearts (rats 50, 52, and 59) on day 2 was hybridized to RNA from 2 separate day 2 control male rats (control rat 2 and 22); and RNA extracted from 3 400 mg/kg male rat hearts (rats 49, 54, and 57) on day 2 was hybridized to RNA from 2 separate day 5 control male rats (control rats 9 and 19). Data from dye reversal hybridizations of treated vs. control RNA samples were combined using Rossetta Resolver 4.0 (RosettaBiosoftware, Seattle, WA).

Images and GEML files, including error and p-values, were exported from the Agilent Feature Extraction software and deposited into Rosetta Resolver (version 3.2, build 3.2.2.0.33) (Rosetta Biosoftware, Kirkland, WA). The resultant ratio profiles were combined into ratio experiments as described in Stoughton and Dai (2002). Intensity plots were generated for each ratio experiment and genes were considered “signature genes” if the p value was less than 0.001 (Stoughton and Dai, 2002).

Testing for differential expression was performed with significance analysis of microarrays (SAM) (Tusher et al., 2001) for four comparisons (2 day vs. 2 day control, 5 day vs. 5 day control, 2 day vs. 5 day, and control vs. 2 day and 5 day combined). For each comparison, the set of genes corresponding to the minimal false discovery rate (FDR) was selected. The FDR for all comparisons was less than 0.01. Hierarchical cluster analysis was used to visualize expression of the genes found significant by SAM. The results of the SAM analysis were used to select genes for confirmation by PCR analysis including genes coding for high energy production/consumption, control of calcium levels, and maintenance of heart function.

Selection of Transcripts for qRT-PCR

The gene transcripts for RT-PCR analysis were selected from those shown to be significant by the SAM analysis of the microarray transcript data from rat hearts (400 mg CEM/kg vs. controls). Heart disease is characterized by mitochondria damage (Ballinger, 2005; Ballinger et al., 2002), as has been reported for the CEM-induced cardiotoxicity (Dunnick et al., 2004a). A balance between energy demand and energy supply must be maintained for the rat to combat the CEM cardiotoxicity (Katz 1991, 1998). Thus, we selected transcripts that are involved in synthesis of ATP, a function that occurs in mitochondria, and in high energy demand processes (Hochacka et al., 1996), to test the hypothesis that when confronted with mitochondria damage, high energy demanding processes are down regulated (e.g., ion pumping, protein synthesis) (Hochacka et al., 1996). Other transcripts, significant by the SAM analysis, were also selected for the dose-response study including transcripts for growth factors, proteins involved with signal transduction, and those involved in maintaining ion balance and heart function.

Quantitative Real-Time PCR

The expression levels of eighteen transcripts (Table 1) involved in energy production, heart function, and cell growth were evaluated by quantitative reverse transcription polymerase chain reaction (qRT-PCR). RNA from cardiac muscle from 6 individual rats at each of the dose groups (0, 200, 400, and 600 mg/kg) was collected at day 2 or day 5. The qRT-PCR reactions were run in triplicate. The quality of each RNA sample was checked on the Agilent Bioanalyzer by analyzing 1 μl aliquots. RNA was reversed transcribed into first strand cDNA using the High-Capacity cDNA Archive kit (Applied Biosystems, Foster City, CA). For each RNA sample, 2.5 μg RNA in volume of 50 μl was combined with an equal volume of the 2X RT Master-Mix (Applied Biosystems, Foster City, CA), containing random primers, dNTP mixture, and Multi-scribe RT enzyme for a total reaction volume of 100 μl per well in a 96-well reaction plate. The plate was incubated for 10 minutes at 25°C and then at 37°C for 2 hours on 9700 ABI thermocycler. The cDNA was stored at −20°C until further use.

The cDNA was amplified using primer and probe sets from Assays on Demand, (Applied Biosystems, Foster City, CA) on an ABI 7900 Sequence Detection System (Applied Biosystems, Foster City, CA). Universal master mix (Applied Biosystems, Foster City, CA) with the specified Taqman Primer Probe set was added to each well on a 384 well reaction plate using a Precision 2000 liquid handler. Fifty ng of each cDNA was loaded into the master mix in each reaction well to a final volume of 20 μl. The samples were amplified by incubation for 2 minutes at 50°C, then 10 minutes at 95°C, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. SDS Software version 2.1 and Microsoft excel software was used for analysis of the resulting data from the relative quantitation assay. Manual threshold values were used and expression of each gene was normalized to Acidic Ribosomal Phosphoprotein (arbp) (Hafer-Macko et al., 2005) and expressed relative to a calibrator (•C_Tcb = CT control sample_Gene − CT Control sample_arbp) with the use of the formula 2^−ΔΔCt to calculate fold change as described by K. Livak: •C_Ttrt = (CT Treatment group sample_Gene − CT Treatment group sample_arbp); −ΔΔC_T = − (•C_Ttrt −•C_Tcb): fold change = 2^−ΔΔCT(Livak and Schmittgen, 2001).

Statistical Analysis of Gene Expression Data

Expression of the 18 transcripts (Table 2) was analyzed to determine differences from control expression, and differences across time and dose. Because fold-changes were not normally distributed with equal variances, nonparametric statistical methods were used (Conover, 1971). Trends with dose were tested with the Spearman rank-order correlation coefficient. Fold-changes were compared between each pair of dose levels using an exact Mann–Whitney test (Table 2). All p-values are two-sided.

Histopathology

Histologic and ultrastructural analysis of hearts from male rats after 2, 3, 5, and 16 days of CEM treatment have previously been reported (Dunnick et al., 2004a, 2004b).

Results

CEM induced heart toxicity in all regions of the heart (Figure 2) (Dunnick et al., 2004a, 2004b). The CEM-induced heart damage occurred as early as day 2, and began to resolve by day 5, and by day 16 had completely resolved (Figure 2). Ultrastructural analysis showed that the primary site for heart damage was the mitochondrion. This included mitochondrial swelling with disruption of cristae, loss of membrane structure, and distention of sarcoplasmic reticulum (Dunnick et al., 2004a).

Microarray analysis of RNA extracted from the heart of control or 400 mg/kg male rats on day 2 and day 5 showed that unique gene transcript changes were dependent on day of examination. Transcripts found to be significantly different from controls at p < 0.01 resulted in 240 up-regulated transcripts, day 2; 101 down-regulated transcripts, day 2; 113 up-regulated transcripts, day 5; and 102 down-regulated transcripts, day 5. Cluster analysis segregated the day 2 and day 5 significant genes transcripts. There were 27 gene transcripts in common for day 2 and day 5, with most of these being down regulated.

There was generally a statistically significant dose response for the 18 transcripts evaluated by qRT-PCR (Table 2, Figure 3). There was an increase in transcript expression for adra2, calb3, egr1, kcnj12, nectin3, tcf4, ucp1, vegf relative to control expression on day 2 and/or day 5, particularly in the 200 and 400 mg/kg groups. Cardiotoxicity was associated with a decrease in transcript expression for adra2, abcb1a, ATP5j, ATP5k, b2m, bsn, edg5, psma2, slc4a3, tcf4, and vegf relative to control expression on day 2 and/or day 5, particularly in the 600 mg/kg group. Pip5k2a and rgs2 transcript levels were close to control levels at all doses on day 2 and day 5. Transcript expression for the endogenous control (arbp) was similar across all samples as demonstrated by a standard error of 0.038.

Many of the significant transcripts produce proteins that function at the cell membrane level (edg5, kcnj12, rgs2, slc4a3, nectin3, adra2, transcripts). Thus, the response to CEM exposure involved changes that take place at the cell membrane (e.g., response to extracellular signals, ion transport, cell adhesion).

Discussion

Cardiac disease leads to impairment of energy supply through mitochondria damage (Ballinger, 2005; Ballinger et al., 2002; Bulteau et al., 2005; Kang and Hamasaki, 2005; Rosenberg, 2004). Thus, in the CEM rat model of heart toxicity, where mitochondria damage occurs, we explored changes in heart transcripts associated with energy supply and demand, and how changes in heart transcripts may help to maintain heart function.

ATP synthase activity is essential for energy production (Tomasetig et al., 2002), and subunits ATP synthase were down-regulated after CEM exposure in a dose-related manner. The down-regulation of subunit e (ATP5k), a protein required for maintaining the ATP synthase dimeric form (Arsselin et al., 2004), and the regulation of mitochondrial H⁺-ATP synthase activity via a Ca²⁺-dependent regulatory region (Arakaki et al., 2001), suggests that energy supply was decreased in the CEM-treated heart. A general survival principal is that when energy production is decreased, the cell/species responds by decreasing high energy consuming processes such as ion pumping and/or protein synthesis and degradation (Hochacka et al., 1996). Our findings suggest that the down-regulation of high energy consuming processes did occur based on the decreased transcript expression for ion pumps, transport processes, and protein degradation (abcb1a, kcnj12, slc4a3, psma2). For example, ATP binding cassette subunit 1a (abcb1a) protein functions to transport phosphatidylcholine from the inside of the cell to the outside (Kalin et al., 2004), and the down-regulation of this transport function may be one way to conserve energy.

The down-regulation of proteosome alpha subunit 2 (psma2), and, therefore, protein degradation, may be another way to conserve energy.

Other changes in heart transcripts, suggest changes in ion flow. The Na⁺-independent Cl⁻HCO⁻ exchanger gene, 3 slc4a3 (also known as AE3), was down regulated on day 2 and 5, suggesting that CEM cardiotoxicity is characterized in part by failing to control alkalosis. Members of the AE3 anion exchanger family are involved in the recovery from alkaline loads in myocardial tissue (de Cingolani et al., 2001; Orchard and Kentish, 1990). Chloride/bicarbonate exchangers prevent intracellular alkalosis (Papageorgiou et al., 2001), and the slc4a3 exchanger is prominent in the recovery from heart toxicity (Alvarez et al., 2004).

Transcripts for G protein-regulated inwardly rectifying K⁺ channels (kcnj12), which play a role in regulation of heart rate (Nikolov and Ivanova-Nikolova, 2004), were up regulated on day 2, and may have resulted in hyperpolarization and a slowing of heart rate (Leaney et al., 2004). Inward rectifier potassium channels conduct currents at voltages around the reversal potential, and can stabilize the resting membrane potential and repolarize the myocyte (Kaibara et al., 2002).

An increase in the transcript for a calcium binding (calb3) protein was noted on day 2 and 5 in the 2 lower exposure levels, and this protein may help to control intracellular calcium levels, a critical function for maintaining cardiac function and cycling (Wang and Goldhaver, 2004; Wehrens et al., 2005) (Hagihara et al., 2005) Chacon et al., 2001). At the high CEM exposure level the heart was unable to launch this response.

Other studies in the literature suggest that nectin 3, ucp 1, beta2microglobulin, and alpha 2 adrenergic receptors (adra2) help to maintain heart function, and the upregulation of transcripts for these proteins, suggest that this may be part of the defensive mechanisms in the heart to combat CEM heart toxicity. Nectin 3, is a calcium independent member of the cadherin superfamily (Satoh-Horikawa et al., 2000), which participates in the recruitment of other cadherin proteins to promote cell-cell adhesion (Irie et al., 2004) and in the maintenance of end-to-end connections in the myoctes (Ferreira-Cornwell et al., 2002). Ucp1 uncouples mitochondria respiration from ATP production resulting in a net increase in expenditure of caloric energy as heat (Klingenberg, 2001; Rousset et al., 2004), and a maintenance of temperature (Hoerter et al., 2004). Beta2-microglobulin knockout mice have improved cardiac function and less hypothermia when exposed to the cardiac toxin, anti-asioloBM1, than control mice (Tao and Sherwood, 2003). This suggests that the down regulation of beta2-microglobulin (b2m) on day 2 and 5 may help to maintain heart function.

Expression of the transcript for alpha 2 adrenergic receptor, a G protein coupled receptor that mediates vasoconstriction, inotropy, and remodeling (Varma and Deng, 2000), was maintained close to control levels. Activation of these receptors result in positive inotropic effects and maintenance of cardiac output (Brede et al., 2002; Klouche et al., 2002; Niederhoffer et al., 2004). Thus, the down-regulation of the adra2 gene tranascript only at the high exposure level, suggests that in the 2 lower dose groups heart function was maintained in part because alpha 2 adrenergic receptor function was maintained.

Expression of transcripts coding for proteins that promote cell proliferation was maintained or upregulated in the 2 lower exposure groups (egr-1, pip5k2a, edg5, tcf4). Early-growth factor response transcript (egr-1) that codes for a protein that activates down stream growth factors (Khachigian et al., 1996), was up-regulated 2 days after CEM exposure. Egr-1 is commonly activated after vascular and mitochondria injury, and is cardioprotective by controlling calcium levels (Wang et al., 2005). Tcf4 transcript levels were maintained in the 200 and 400 mg/kg groups, but down-regulated in the 600 mg/kg group where cardiotoxicity was most severe. The tcf4 protein signaling is important in maintaining heart function (Graham et al., 2001), regulating cardiac valve formation (Hurlstone et al., 2003), and in turning on genes essential to myocyte cell survival (Blais et al., 2004) including c-myc (He et al., 1998), cyclin D1 (Shtutman et al., 1999; Tetsu and McCormick, 1999), metalloproteinases (Brabletz et al., 1999), and vegf (Zhang et al., 2001). Phosphatidylinositol-4-phosphatekinase (Pip5k2a) catalyzes the synthesis of phosphatidylinositol 4,5-bisphosphate (PIP2). Phosphoinositide 3-kinase (PI3K) phosphorylates PIP2 to form phosphatidyl-3,4,5-triphosphate (PIP3) which activates downstream targets to promote cell survival and proliferation (Luo et al., 2003; Ueno et al., 2003). Pip5k2a transcript expression levels were maintained despite CEM cardiotoxicity, and this maintenance of function may be essential for regulating potassium channels and currents (Ding et al., 2004; Loussouarn et al., 2003). Rgs2 has been reported to play a role in vascular smooth muscle relaxation (Tang et al., 2003), and the expression levels of this transcript were also maintained throughout the CEM cardiotoxicity process.

Spingosine 1-phoshate activates the edg5 receptor, located in the cell membrane, and the activation of edg5 receptors appears to be crucial for myocyte progenitor cell migration (Kupperman et al., 2000). Edg5 signaling is expressed in myoblasts and undifferentiated cells (Meacci et al., 2003). Edg 5 receptor signaling promotes cell proliferation and inhibits apoptosis, in part through activation of phosphatidylinositol 3 kinase–Akt/protein kinase B pathways and activation of Rho (Banno et al., 2001; Yau et al., 2003). Rho signals promote cell proliferation and survival (Radeff-Huang et al., 2004) and play a role in ventricular remodeling after infarction (Hattori et al., 2004).

Vegf protein promotes angiogenesis, a process essential for new cell growth (Ferrara, 2004). Thus, the maintenance of vegf gene expression levels at day 2 in the two lower exposure levels suggests normal cell growth processes were maintained. The Vegf protein, through binding to its receptors, turns on phosphatidylinostiol 3-kinase pathways, and subsequently other cell growth/proliferative processes (Ferrara, 2004).

A number of transcripts coding for proteins active in neonatal cardiac development were upregulated or maintained close to control expression levels (Ucp1 (Erlanson-Albertsoon, 2002); kcnj12 (Zaritsky et al., 2001); adra2 (Porter et al., 2003); tcf4 (Hurlstone et al., 2003)). The maintenance of transcript levels for proteins/signals involved in cell growth (i.e., egr-1, Pip5K2a, tsf4, rgs2, edg5, vegf) supports the hypothesis that there are heart signal pathways that may promote regeneration (Beltrami et al., 2001, 2003; Urbanek et al., 2005).

CEM cardiotoxicity resulted in down regulation of ATP synthase transcripts. This suggested reduced capacity for energy production, and was accompanied by down regulation of transcripts coding for proteins involved in high energy consuming processes, and up regulation of transcripts involved in maintenance of heart function. The heart gene transcript changes observed suggest that cardioprotective processes include energy conservation, maintenance of heart function, and promotion of growth.