The Thrill Can Kill: Murder by Methylation

【关键词】  Methylation

    Why babies of crack-cocaine mothers develop heart problems has always been a mystery. In this issue of Molecular Pharmacology, Zhang et al. (p. 1319) show that a specific methylation occurs at the protein kinase C (PKC) promoter of the babies born of mother rats exposed to cocaine. This reduces the expression of PKC, a naturally cardioprotective enzyme, which provides a plausible molecular mechanism for cardiac failure.

    The rampant use of cocaine as a recreational drug in the 1980s prompted the anticocaine commercial "The thrill can kill." The mood-enhancing effect of cocaine in the user is swift and fleeting. In contrast, the unfortunate children of crack cocaine mothers often inherit developmental and psychological disorders, such as low birth weight and learning disabilities (Bauer et al., 2005). They also show a higher incidence of heart failure and ischemic injury in adulthood. Although its exact cause remains unknown, a article in this issue of Molecular Pharmacology by Zhang et al. (2007) suggests a provocative molecular mechanism. These authors administered cocaine to pregnant laboratory rats and studied their offspring. Compared with pups of control rats not exposed to cocaine, those of cocaine-exposed mothers showed much lower cardiac levels of protein kinase C (PKC), an enzyme known for its cardioprotective power. The mRNA level of PKC was reduced even more dramatically. We were surprised to find that the nucleotide sequence of the PKC promoter was unchanged, pointing to a possible "epigenetic" mechanism, defined by heritable changes in gene regulation without a change in DNA sequence. The search eventually led to a methylation event in which a methyl group placed on a sequence near the PKC promoter prevented binding of the essential transcription factor activator protein-1 (AP-1), resulting in abrogation of PKC transcription. It is clear that the specific methylation in the fetal genome, triggered by the mother's use of cocaine, led to the loss of PKC expression, resulting in cardiac hypertrophy and susceptibility to ischemia and reperfusion injury later in life (Fig. 1). These results have important implications in the etiology and management of crack-cocaine children.

    Fig. 1. Mechanism of fetal PKC gene regulation by maternal cocaine use. The model is based on the findings of Zhang et al. (2007). Left, normal, drug-free mother. Right, mother addicted to cocaine, with the cocaine-related changes in blue color. In the absence of cocaine, the unmodified AP-1 enhancer site adjacent to the PKC gene promoter in the fetal heart binds AP-1 and the fetal PKC gene is expressed, offering cardio-protection. As a result of cocaine use by the mother, the same site is methylated, preventing AP-1 binding and PKC expression, leaving the child's heart vulnerable to future ischemic injuries.

    Cocaine, an alkaloid, prepared from the leaves of the plant Erythroxylon coca and first isolated by the German chemist Friedrich Gaedcke in 1855, has a long history of both medicinal and recreational use in the human society (http://en.wikipedia.org/wiki/Cocaine). World annual cocaine use currently stands at around 600 metric tons, with Americans consuming roughly half and Europeans, about a quarter. In its various forms (mainly "freebase" and "crack") cocaine is second only to cannabis as the most popular illegal recreational drug in the United States, amounting to an annual street business of approximately $40 billion. Like most psychostimulants and antidepressants, cocaine causes transient euphoria through the relatively well documented biochemical stimulation of the dopaminergic system. In contrast, little is known about the mechanism of the more lasting and inheritable effects of cocaine, which makes the studies of Zhang et al. (2007) significant.

    Cocaine kills thousands of its users every year. Within minutes of administration, cocaine is found in a variety of organs, including brain, lung, kidney, and heart (Benveniste et al., 2005). At the clinical and physiological levels, the most common cardiac disorders are ischemia, acute coronary syndrome, myocarditis, cardiomyopathy, and arrhythmias. Past studies have revealed some effects of cocaine at the genetic level in the heart tissue, but their exact clinical correlations await further study. These include reduction of sarcoplasmic reticular Ca2+-ATPase (SERCA2a) levels (Wang et al., 2002a), induction of various forms of cytochrome P450 proteins (Wang et al., 2002b), and transcriptional up-regulation of harmful complements, possibly by reactive oxygen species (Tanhehco et al., 2000). In recent studies, the Zhang group also showed that prenatal cocaine exposure increased apoptosis or programmed cell death in neonatal rat heart and increased the susceptibility to ischemia-reperfusion injury in month-old rats (Bae and Zhang, 2005; Li et al., 2005). What sets the current study (Zhang et al., 2007) apart is its ability to offer an in-depth molecular mechanism based on known premises.

    A major strength of the study derives from its focus on the offspring rather than the cocaine-using mother and the demonstration of a long-lasting epigenetic effect. The key molecular event is the specific methylation in a non-CpG-island region that normally functions as the enhancer for AP-1 binding. The abrogation of binding was demonstrated in a direct assay in vitro as well as by immunoprecipitation analysis of chromosomes in vivo. A number of control experiments were in place. For example, cocaine showed no effect on the binding of AP-1 on unmethylated PKC enhancer sequence, showing that AP-1 itself was unaffected. Reporter gene assays provided in vivo confirmation of the loss of AP-1-dependent transcription when the methylated enhancer was used. Enhancers of a number of other transcription factors were not methylated, ruling out their role. Finally, methylation of the promoter region of another PKC (i.e., PKC) was unaffected. Although not specifically shown in this article, there is a large volume of literature on the cardioprotective role of PKC that probably operates via phosphorylation of multiple downstream molecules (Liu et al., 1999; Inagaki et al., 2006; Morrow et al., 2006; Yue et al., 2006).

    As with any leading study, the article brings up as many questions as answers. An important question is whether the observed effect is due to cocaine in the fetal blood or to a secondary effector produced by the mother's system in response to cocaine, which then traveled to the fetus. Cocaine itself is known to freely traverse from the mother's bloodstream to the fetal circulation (Benveniste et al., 2005), and the same is possible for other small molecule effectors. Because the placental transport cannot be obstructed without harming the fetus, a better therapeutic target may be the putative methyltransferase activity itself, which needs to be characterized. A relatively unique activity with specialized regulatory roles will be a desirable drug target for short-term intervention during each episode of cocaine use. Another question is the stability of the methylation through adulthood and old age. Zhang et al. (2007) correctly speculate that as the cardiomyocytes do not divide after birth the acquired methylation is likely to persist. However, this should be tested at various ages because methylation can be a reversible process.

    The article also leaves the door open as to what activates the methylation in the first place. By far the two most common mechanisms of epigenetics are post-translational modifications of histone (e.g., acetylation, phosphorylation) and methylation of DNA. Both play important roles in a variety of inheritable phenomena in health and disease, including tissue-specific and parental imprinting, X-chromosome inactivation, development and cancer (Couture and Trievel, 2006; Wood and Oakey, 2006). It would be interesting to determine whether histone modification precedes and actually regulates the accessibility of the DNA to the methylation process discovered by Zhang et al. (2007). Pioneering studies have revealed that sets of adjacent genes are often regulated jointly, most likely through ultrastructural remodeling of the chromosome, mediated by local modification of histone (Spellman and Rubin, 2002). Each set may cover 20 to 200 kilobase pairs, encompassing 10 to 30 genes. Although the full physiological implication of this finding remains to be determined, an examination of the genes on both sides of the PKC gene on chromosome 2p21 reveals that they code for critical signaling proteins, such as transcription factors, phosphatases, and transporters (Fig. 2). These genes are highly expressed in cardiovascular as well as in hematopoietic, digestive, retinal, neural, and reproductive tissues. Indeed, it is probably not a coincidence that significant damage to all these tissues is observed in cocaine users (Summavielle et al., 2000; Darland and Dowling, 2001; Li et al., 2003; Wilson and Saukkonen, 2004; Ellis and McAlexander, 2005; He et al., 2006). In light of the findings of Zhang et al. (2007), it can now be tested whether the PKC-neighboring genes are also regulated in the cocaine-exposed children, positively or negatively. Finally, small noncoding RNAs such as microRNAs, transcribed from RNA PolII promoters, are now known to regulate a variety of cellular processes, including chromatin methylation and silencing (Almeida and Allshire, 2005; Barik, 2005). If a specific noncoding transcript plays a role in altering the chromatin structure for this particular methylation or for other epigenetic fetal changes, it will open new layers of regulation. Recent studies have indeed shown that altered miRNA levels may underlie disease states and that there seems to be a correlation between locations of miRNA genes with sites of genetic alteration (Calin and Croce, 2006; Huppi et al., 2007). Regardless of the mechanism, it seems that we are witnessing the dawn of a new era in which extensive studies of fetal epigenetic changes will be undertaken. Comprehensive genome-wide analyses may unravel the epigenetic signature of each intrauterine and/or environmental factor, which will affect both basic and clinical sciences.

    Fig. 2. The human genomic PKC locus. Top, location of PKC (PRKCE) and its neighbor genes on chromosome 2p21 based on GENATLAS (http://www.dsi.univ-paris5.fr/genatlas/). Bottom, short description of the same genes, including the tissue(s) in which they are expressed abundantly, their major domains, and known or putative functions.

【参考文献】
  Almeida R and Allshire RC (2005) RNA silencing and genome regulation. Trends Cell Biol 15: 251-258.

Bae S and Zhang L (2005) Prenatal cocaine exposure increases apoptosis of neonatal rat heart and heart susceptibility to ischemia-reperfusion injury in 1-month-old rat. Br J Pharmacol 144: 900-907.

Barik S (2005) Silence of the transcripts: RNA interference in medicine. J Mol Med 83: 764-773.

Bauer CR, Langer JC, Shankaran S, Bada HS, Lester B, Wright LL, Krause-Steinrauf H, Smeriglio VL, Finnegan LP, Maza PL, et al. (2005) Acute neonatal effects of cocaine exposure during pregnancy. Arch Pediatr Adolesc Med 159: 824-834.[Abstract/Free Full Text]

Benveniste H, Fowler JS, Rooney W, Ding YS, Baumann AL, Moller DH, Du C, Backus W, Logan J, Carter P, et al. (2005) Maternal and fetal 11C-cocaine uptake and kinetics measured in vivo by combined PET and MRI in pregnant nonhuman primates. J Nucl Med 46: 312-320.[Abstract/Free Full Text]

Calin GA and Croce CM (2006) MicroRNA signatures in human cancers. Nat Rev Cancer 6: 857-866.

Couture JF and Trievel RC (2006) Histone-modifying enzymes: encrypting an enigmatic epigenetic code. Curr Opin Struct Biol 16: 753-760.

Darland T and Dowling JE (2001) Behavioral screening for cocaine sensitivity in mutagenized zebrafish. Proc Natl Acad Sci USA 98: 11691-11696.[Abstract/Free Full Text]

Ellis CN and McAlexander WW (2005) Enterocolitis associated with cocaine use. Dis Colon Rectum 48: 2313-2316.

He F, Lidow IA, and Lidow MS (2006) Consequences of paternal cocaine exposure in mice. Neurotoxicol Teratol 28: 198-209.

Huppi K, Volfovsky N, Mackiewicz M, Runfola T, Jones TL, Martin SE, Stephens R, and Caplen NJ (2007) MicroRNAs and genomic instability. Semin Cancer Biol 17: 65-73.

Inagaki K, Churchill E, and Mochly-Rosen D (2006) Epsilon protein kinase C as a potential therapeutic target for the ischemic heart. Cardiovasc Res 70: 222-230.[Abstract/Free Full Text]

Morrow CE, Culbertson JL, Accornero VH, Xue L, Anthony JC, and Bandstra ES (2006) Learning disabilities and intellectual functioning in school-aged children with prenatal cocaine exposure. Dev Neuropsychol 30: 905-931.

Li G, Xiao YH, and Zhang L (2005) Cocaine induces apoptosis in fetal rat myocardial cells through the p38 MAPK and mitochondrial/cytochrome c pathways. J Pharmacol Exp Ther 312: 112-119.[Abstract/Free Full Text]

Li H, Dunbar JC, and Dhabuwala CB (2003) Expression of cAMP-responsive element modulator (CREM) in rat testes following chronic cocaine administration. J Environ Pathol Toxicol Oncol 22: 111-116.

Liu GS, Cohen MV, Mochly-Rosen D, and Downey JM (1999) Protein kinase C-epsilon is responsible for the protection of preconditioning in rabbit cardiomyocytes. J Mol Cell Cardiol 31: 1937-1948.

Summavielle T, Silva-Araujo A, Silva MC, and Tavares MA (2000) Effects of neonatal exposure to cocaine in the development of the neurotransmitters retinal systems: an immunocytochemical and neurochemical study. Ann N Y Acad Sci 914: 418-430.

Spellman PT and Rubin GM (2002) Evidence for large domains of similarly expressed genes in the Drosophila genome. J Biol 1: 5.

Tanhehco EJ, Yasojima K, McGeer PL, and Lucchesi BR (2000) Acute cocaine exposure up-regulates complement expression in rabbit heart. J Pharmacol Exp Ther 292: 201-208.[Abstract/Free Full Text]

Wang JF, Yan X, Min J, Sullivan MF, Hampton TG, and Morgan JP (2002a) Cocaine downregulates cardiac SERCA2a and depresses myocardial function in the murine model. Can J Physiol Pharmacol 80: 1015-1021.

Wang JF, Yang Y, Sullivan MF, Min J, Cai J, Zeldin DC, Xiao YF, and Morgan JP (2002b) Induction of cardiac cytochrome p450 in cocaine-treated mice. Exp Biol Med (Maywood) 227: 182-188.[Abstract/Free Full Text]

Wilson KC and Saukkonen JJ (2004) Acute respiratory failure from abused substances. J Intensive Care Med 19: 183-193.[Abstract/Free Full Text]

Wood AJ and Oakey RJ (2006) Genomic imprinting in mammals: emerging themes and established theories. PLoS Genet 2: e147.

Yue Y, Qu Y, and Boutjdir M (2006) Protective role of protein kinase C epsilon activation in ischemia-reperfusion arrhythmia. Biochem Biophys Res Commun 349: 432-438.

Zhang H, Darwanto A, Linkhart T, Soers LC, and Zhang L (2007) Maternal cocaine administration causes an epigenetic modification of protein kinase C gene expression in fetal rat heart. Mol Pharmacol 71: 1319-1328.

作者单位：Department of Biochemistry and Molecular Biology, University of South Alabama, College of Medicine, Mobile, Alabama