Small Heat Shock Proteins and Doxorubicin-Induced ... - Springer Link

1 downloads 0 Views 652KB Size Report
2006; 41(3):389–405. 40. Gambliel HA, Burke BE, Cusack BJ, et al. Doxorubicin and C-13 deoxydoxorubicin effects on ryanodine receptor gene expression.
Small Heat Shock Proteins and Doxorubicin-Induced Oxidative Stress in the Heart Karthikeyan Krishnamurthy, Ragu Kanagasabai, Lawrence J. Druhan, and Govindasamy Ilangovan

Abstract  Doxorubicin (Dox) and its derivatives are used as chemotherapeutics, either alone or in combination with other agents. Dilated cardiomyopathy and congestive heart failure due to cardiotoxicity continues to be the most serious side effect, imposing severe limitations in the use of these agents despite the arrival of new classes of Dox-derivatives and new formulations. In this chapter we summarize the recent understanding of the mechanism of Dox-induced cardiotoxicity and its relevance to the stress-inducible proteins, with special emphasis on the small heat shock proteins such as Hsp27, Hsp20, etc. The heat shock proteins are expressed as a response to the oxidative stress in the heart due to the redox reactions of these drugs and the generation of reactive oxygen species (ROS). On the other hand, ROS are also known to induce various MAP kinases and phosphorylate and activate the stress-responding transcription factors, including the heat shock factors (HSF). Activation of HSF-1 leads to the induction of a series of heat shock proteins, depending upon the type of exerted stress. Recent studies have confirmed that Dox-induced oxidative stress indeed leads to HSF-1 activation to induction of heat shock proteins, especially small Hsps in the heart. The Dox-induced small Hsps have been found to be involved in cell signaling and can be either cardioprotective or detrimental. Additionally, a few transgenic animal models have shown that selective overexpression of these proteins can be cardioprotective against Dox. These results establish the fact that proper regulation of the function of small Hsps could eliminate cardiotoxicity and serve as a potential therapeutic target to protect the heart from Dox-induced toxicity. Keywords  Adriamycin • Doxorubicin • Heart failure • Heat shock factor-1 • Oxidative stress • Small heat shock proteins

G. Ilangovan (*) Division of Cardiovascular Medicine, Davis Heart and Lung Research Institute, Ohio State University, 460 West 12th Ave, 43210 Columbus, OH, USA e-mail: [email protected] S. Basu and L. Wiklund (eds.), Studies on Experimental Models, Oxidative Stress in Applied Basic Research and Clinical Practice, DOI 10.1007/978-1-60761-956-7_5, © Springer Science+Business Media, LLC 2011

105

106

K. Krishnamurthy et al.

1 Introduction Doxorubicin (Dox, Adriamycin) is one of the original anthracyclines isolated in the early 1960s from the pigment-producing bacterium Streptomyces peucetius. To this day, adriamycin and its derivatives remain some of the most effective anticancer drugs [1, 2]. Adriamycin is a potent and broad spectrum antineoplastic agent used in the treatment of several cancers, such as Hodgkin’s lymphoma, aggressive nonHodgkin’s lymphomas, acute lymphoblastic leukemia, metastatic breast carcinoma, ovarian carcinoma, lung carcinoma, and sarcoma. It has continued to be considered a first-line antineoplastic drug for more than 30 years [3]. However, this drug has proven to be a double-edged sword because it also causes a serious side effect, namely, cardiotoxicity leading to cardiomyopathy and congestive heart failure (CHF). The risk of developing irreversible cardiomyopathy due to Dox treatment may be increased by several factors. These include the degree of anthracycline exposure (high-dose anthracycline infusion or higher cumulative doses are responsible for more frequent cardiac problems), age (the elderly and very young patients are at greater risk), a history of cardiac disease, previous cancer therapies (e.g., mediastinal radiation therapy), and the concurrent use of chemotherapy regimens that include paclitaxel or trastuzumab [4, 5]. The mortality rate of anthracyclinerelated CHF is greater than 50% within 2 years from diagnosis in patients with New York Heart Association class III/IV [6, 7]. In order to overcome this unwanted side effect, new developments such as new derivatives with less toxicity to the heart, new formulations in the form of encapsulation into delivery vehicles (such as microemulsions), new targeted delivery modalities to selectively deliver the drug to the desired site (avoiding the accumulation in the heart etc.) [8], have emerged. However, there has been no significant improvement towards avoiding cardiotoxicity and complete elimination of this side effect has not been achieved. Fortunately, the mechanisms of killing tumor cells (intended effect) and the cause of cardiomyopathy (unwanted effect) by Dox and its analogs are found to be different, indicating that the mode of action of these drugs depends upon the nature of tissues. This difference in action gives hope that by selectively targeting some of the pathways by novel approaches, the untoward development of cardiomyopathy and CHF among Dox treated patients can be avoided. Thus, a new generation of interest has arisen to combat the oxidative stress and to identify and target these pathways. In this chapter we summarize the research that has been carried out to determine the mechanisms of the induction of stress inducible proteins, specifically heat shock proteins (Hsps), in the heart upon treatment with Dox.

2 Differential Mechanisms of Action of Doxorubicin: The Oxidative Stress in the Heart Anthracyclines are known to kill cancer cells by DNA intercalation and topoisomerase II inhibition, while the loss of cardiomyocytes in the heart has been attributed to oxidative stress, caused by oxidants such as oxygen-derived free radicals

Dox and Small Hsps in the Heart

107

DNA Intercalation/ Topo II Inhibition

Cell Death

DOX CANCER CELLs

NADPH NADP+

ROS

Oxidative stress-p53 activation/mTOR inhibition

Apoptosis/ Cell Death

Heart Failure

CARDIOMYOCYTES

Fig. 1  Differential mechanisms of Dox-induced cell death in malignant tissue and the heart. In cancer cells, the primary mechanism of cell death involves DNA intercalation, FasR mediated apoptosis, and topo II inhibition. Cardiomyocyte death in the heart is mainly by mechanisms involving oxidative stress-related apoptosis (endogenous p53/Bax pathway), mTOR inhibition, TLR 2 activation, etc. The differences in the mechanisms are mainly due to differences in the nature of the cells (differentiating versus terminally differentiated cardiomyocytes)

(O2•−, OH•) and generation of H2O2 (Fig. 1) [9]. If cell death in the heart is avoided by selectively targeting the cardiomyocyte death pathways, the intended outcome of Dox treatment could be greatly improved. Indeed, supplementation of antioxidants or specific overexpression of endogenous antioxidants has been found to be very effective in alleviating Dox-induced heart failure in animal models [10]. However, some clinical trials with antioxidants such as vitamin E showed no significant improvement [11, 12]. Thus, to realize the full potential of antioxidantbased adjuvant therapy, the complete mechanism of cardiomyocyte death needs to be understood. Even the precise mechanisms of action of anthracyclines in tumor cells remain a matter of controversy [13]. Suggested mechanisms include (1) intercalation into DNA, leading to inhibition of synthesis of macromolecules; (2) generation of reactive oxygen species (ROS), leading to DNA damage or lipid peroxidation; (3) DNA binding and alkylation; (4) DNA cross-linking; (5) interference with DNA unwinding or DNA strand separation and helicase activity; (6) direct membrane effects; (7) initiation of DNA damage via inhibition of topoisomerase II; and (8) induction of apoptosis in response to topoisomerase II inhibition [9, 13, 14]. Likewise, several mechanisms have been proposed to explain adriamycin-induced myocardial damage; the most important among them is generation of oxygen-derived ROS and subsequent “oxidative stress.” Increased oxidative stress and antioxidant deficit have been suggested to play a major role in adriamycin-induced cardiomyopathy and CHF. Previously, EPR spin-trapping has been successfully used to show the free radicals generated during Dox treatment [15]. There are two major pathways by which Dox causes free radical formation [16–19], an enzymatic mechanism catalyzed by several mono-electronic oxidoreductases and a nonenzymatic mechanism involving anthracycline–iron complexes. In the enzymatic mechanism, one electron reduction of the quinone moiety in the C ring of anthracyclines results in

108

K. Krishnamurthy et al.

the formation of a semiquinone radical that quickly regenerates its parent quinone by reducing molecular oxygen to superoxide anion (O2•–) and subsequently to hydrogen peroxide (H2O2) via dismutation, thus increasing ROS concentrations above the physiologic levels generated during normal aerobic metabolism [20, 21]. This futile cycle is catalyzed by a number of flavin-centered oxidoreductases located primarily in mitochondria (NADH-ubiquinone oxidoreductase) [22, 23], as well as in microsomes (NADPH-cytochrome P450 or NADPH-cytochrome b5 reductases) [24, 25], nucleus (NADPH-cytochrome b5 reductases) [19] and cytosol (xanthine dehydrogenase, nitric oxide synthase) [26–28]. Anthracycline-free radicals may arise via a nonenzymatic mechanism involving reactions of anthracyclines and iron. For example, iron (III) readily interacts with Dox. This is followed by a redox reaction, wherein the iron atom accepts an electron, generating an iron (II)-doxorubicin-free radical complex to shuttle electrons to O2 forming redox cycle [29]. Anthracycline aglycones may generate ROS as well, as another pathway of nonenzymatic ROS generation. Due to their higher lipophilicity, aglycones can intercalate into mitochondrial membranes better than the parent anthracyclines and thus divert more electrons towards oxygen, thereby generating ROS in close proximity to sensitive targets [30, 31]. Additionally, Dox treatment may also increase nitric oxide (NO) production by upregulating the expression of the inducible isoform of NO synthase (iNOS) [32]. The concurrent over-production of O2•– and NO yields reactive nitrogen species (RNS) and, in particular peroxynitrite (ONOO–) [33, 34], a powerful oxidant that may directly damage every type of cellular macromolecules, resulting in lipid peroxidation, protein covalent modifications, and DNA strand breaks. Anthracycline-related oxidative and nitrosative stress have been reported to interfere with many aspects of cardiac function by inducing mitochondrial dysfunction [5, 35–37], energy imbalance [38, 39], disruption of the cardiac specific gene expression program [40, 41], and apoptosis [42–44]. In most cells, ROS and RNS formation is kept to a minimum by the presence of detoxifying enzymes, like superoxide dismutase, catalase, and glutathione peroxidase. Unfortunately, cardiomyocytes are more susceptible than other cells to oxidative and nitrosative damage for several reasons (a) high metabolic activity [45]; (b) high concentration of cardiolipin, a mitochondrial membrane-bound phospholipid critical to cell respiration for which anthracyclines have a high affinity [46, 47]; and (c) weak antioxidative defenses due to their low content of catalase, superoxide dismutase and glutathione peroxidase [48, 49]. A growing number of reports indicate that free radical damage to the mitochondria and defenses against free radicals may contribute to cumulative cardiomyopathy [4]. This concept of oxidative stress is directly supported by the detection of Dox-generated ROS using electron spin resonance spectroscopy [27, 50, 51]. In addition, indirect support is offered by studies showing that Dox increases levels of tissue malondialdehyde, which is a product of lipid peroxidation [52, 53]. There is also evidence that Dox metabolites induce myocardial damage. Doxorubicinol, the primary C-13 alcohol metabolite of Dox, has been found to accumulate in ­cardiac tissue in a time- and dose-dependent manner [54]. Anthracycline alcohol metabolites can affect myocardial energy metabolism, ionic gradients, and Ca2+

Dox and Small Hsps in the Heart

109

movements. All of these effects impair cardiac contraction and relaxation. Research has proven that, at concentrations that are relatively low when compared to those of Dox, Doxorubicinol compromises both systolic and diastolic function of isolated heart preparations and block ATPase activity of the sarcoplasmic reticulum, mitochondria, and sarcolemma [55].

3 Enhanced Heat Shock Proteins Expression in Dox-Treated Hearts Cells respond to stressful conditions by activating genetic programs whose evolutionarily conserved mechanisms have common ancestral origins, from the simplest bacteria to complex organisms including humans [56]. One such genetic program that has gained increased attention is the expression of Hsps under genotoxic environments such as Dox treatment. Hsp is expressed in all living cells, in response to exerted stress. Both in vivo and in vitro studies have shown that various stressors transiently increase the production of Hsp as a protection against harmful insults. Hsp induction was first identified in the salivary glands of Drosophila upon application of heat shock [57]. Since then, studies of individual genes have shown that the cellular heat shock response is conserved across kingdoms and is characterized by the strong induction of numerous heat shock proteins. Hsps expression was found to correlate with the acquisition of thermotolerance and cytoprotection. Thus initially it was thought that Hsps expression was a unique mechanism to adapt to hypothermia [58]. Later, the interest in Hsps expanded tremendously as more and more processes in normal resting cells involving Hsp were uncovered. The multifunctional roles of Hsps in cells show that Hsps are major regulatory proteins in the cell, and vital functions of the cell, such as the maintenance of the cell cycle, are associated with Hsps [59, 60]. In mammalian cells, the stress response involves the induction of four major classes of Hsp families, namely, the small Hsp exemplified by Hsp27, Hsp60, Hsp70, and Hsp90 [61–64]. Increased expression of Hsps in Dox-treated hearts was reported as early as 1990. The Schoeppe group reported increased expression of a 30 kDa “stress protein” in cultured cardiomyocytes, upon treated with cisplatin [65] and inhibition of this protein upon being treated with Dox. Although initially the purpose of the expression of these proteins was unknown, only later was it proven that Hsp induction was an endogenous defensive mechanism of cells in response to oxidative stress. Initial experiments demonstrated that co-injection of carnitine with Doxenhanced Hsp25 expression threefold, and this provided significantly enhanced cardioprotection [65, 66]. These studies were further supported by different groups demonstrating the effect of thermal preconditioning [67] and physical exercise [68, 69] on the Dox-induced cardiotoxicity. Despite the fact that these early studies established increased expression of Hsps as cardioprotective, the mechanistic details of such enhancement, the signaling that followed such an increased expression, were lacking prior to the identification of the transcription factors that translated the stress into expression of Hsps.

110

K. Krishnamurthy et al.

4 Transcription of Hsps in Dox-Treated Hearts: The Role of Heat Shock Factors Environmental insults provoke a variety of adaptive physiological responses to help organisms cope with specific stressors. The dramatic induction of Hsps is an important unifying component of most of these responses for survival under stressful conditions. Regulation of Hsp expression is intricate, with multiple layers of redundancy and feedback control (Fig. 2). Systemic stresses, genotoxic (i.e., Dox treatment) and proteotoxic (accumulation of denatured proteins) stresses, activate a small family of transcription factors called heat shock factors (HSFs) that are the primary regulators of stress-inducible expression in eukaryotic cells. HSFs, originally found to respond to heat shock, but they are now known to respond to any environmental stress, are activated in toxic environments such as the genotoxic agents by a complicated mechanism. The structure and function of HSFs have been conserved for more than a billion years [70]. Several HSFs are present in mammalian cells (i.e., murine and human HSF-1, -2, and -4 and a unique avian HSF-3), but HSF-1 is clearly the dominant factor controlling cellular responses to stress.

a Monomers of HSF

DOX

MAP Kinases activation

ROS

NADPH

PPP

Phosphorylated HSF

NADP+

HSP27 HSP20

mRNA

CYTOSOL

PPP

HSE NUCLEUS

b

ERK GSK3β

P 230

N

HR-A/B

DBD 16

120 137

S

JNK/PKC

P P

RD 212

P

298 303 307 363

HR-C 310

378

395 407

AD

C 503

Fig. 2  Oxidative stress-induced activation of HSF-1. (a) Schematics of the HSF-1 activation upon stress. External stresses phosphorylate inactive monomeric HSF-1 which translocates into nucleus as trimers. The HSE bound HSF-1 trimer becomes transcriptionally active upon hyperphosphorylation and leading to transcription of mRNA for Hsps. (b) Different domains and possible phosphorylation sites in HSF-1

Dox and Small Hsps in the Heart

111

HSF-1 was first identified in cellular extracts from Saccharomyces cerevisiae and Drosophila melanogaster. It oligomerizes on activation and binds to heat shock element (HSE) sequences to control stress-inducible expression of Hsp genes [71]. In mice, HSF-1 is ubiquitously expressed but is particularly abundant in ovary, placenta, heart, and fetal brain [72]. HSF-1 is dispensable for growth and survival under controlled laboratory conditions but essential for survival following stresses such as high temperature and endotoxin challenge [73]. Transgenic animal models have demonstrated the importance of HSFs in reproduction and survival [74]. HSF1-deficient mouse embryos suffer from defects in placental development and are recovered from cross-breeding in lower numbers than expected by Mendelian segregation. Other than being ~20% smaller than wild-type mice, however, HSF-1deficient mice display no overt organ system abnormalities and, in the absence of acute stress, live to late adulthood [73]. Although the HSF-2 transcription factor is refractory to typical stress stimuli, its ability to bind DNA is specifically enhanced in certain development situations. Studies using specific polyclonal antibodies clearly showed that only HSF-1 can respond to heat and other physiological stress [75–77]. HSF-1 was found to be activated by diverse forms of stress and the activation of HSF-1 occurs via a multistep process [78]. Under nonstress conditions, mammalian HSF-1 appears as a monomeric form that exhibits little DNA binding activity. Activation of HSF-1 is accomplished by a stress-induced conversion of monomers to trimers, translocation of HSF-1 from cytosol to the nucleus, posttranslational modifications including phosphorylation of serine residues, and transcriptional competence with enhanced heat shock gene expression [79] (Fig.  2). HSF binds to a conserved DNA sequence motif, the HSE, which is characterized by an array of inverted repeats of the motif nGAAn. Copies of the HSE are found in the promoters of genes encoding several known Hsps [80]. HSF–HSE DNA binding is not sufficient to elicit maximal transcription of the Hsp genes, and it is necessary for HSF-1 to be modified by phosphorylation and sumoylation to increase its transcriptional activity [71, 81]. It has been suggested that HSF-1 is repressed by GSK-3b (Ser303), ERK (Ser307), and JNK (Ser363) under normal conditions, whereas it is activated by hyperphosphorylation (Ser230) upon exposure to various stresses (Fig. 2) [81, 82]. As illustrated in Fig. 2, HSF-1 contains two distinct carboxyl-terminal activation domains, AD1 and AD2, which are under the control of a centrally located, heat-responsive regulatory domain (RD) [83]. Since AD1 does not appear to be heat-regulated by itself, the regulatory domain of HSF-1 has been proposed to play a key role in sensing heat stress in humans [84]. Constitutive phosphorylation of two specific serine/proline motifs, S303 and S307, is important for the function of the RD and may be critical for negative regulation of HSF-1 transcriptional activity at normal temperatures, since substitution of serine to alanine causes constitutive transcriptional competence [84–86]. Constitutive phosphorylation of S363 has also been found to negatively regulate HSF-1 under normal growth conditions [87]. The positive role of phosphorylation in regulation of HSF-1 transcriptional activity is only emerging. So far, one phosphorylated site (S230), which has a positive effect on HSF-1 transactivating capacity, has been characterized [88]. A number of studies have shown that HSF-1

112

K. Krishnamurthy et al.

and Hsps confer protection against cardiovascular disease. Induction of HSF-1 and Hsp expression by various stimuli, such as heat shock, reduces the size of infarcts after ischemia/reperfusion [89]. Transgenic mice overexpressing a constitutively active form of HSF-1 or inducible Hsp70 in the heart show more resistance to ischemia/reperfusion injury compared with wild-type mice [90]. In contrast, the cardiac function of inducible Hsp70 knockout mice is markedly impaired by ischemia/reperfusion injury [91]. In addition to a protective effect against ischemia/ reperfusion injury and Dox-induced cardiomyopathy, it has been reported that Hsps have a beneficial role in myocardial infarction and atrial fibrillation [92, 93]. ROS activates HSF-1, which in turn increases the expression of Hsps including Hsp90, Hsp70, Hsp60 and Hsp25. In addition to their well-characterized role as molecular chaperones, previous studies have found involvement of Hsps in various cellular signaling [94]. Hsp90 has been extensively studied in cellular systems demonstrating that it activates nitric oxide synthase (NOS) [95–97]. A recent study using cardiac cell has reported that Hsp90 activates endothelial NOS to increase NO generation. The excess NO produced by such an activation can block the mitochondrial electron transport chain (ETC) by binding at the cytochrome c oxidase of complex IV [98]. Hsp27 and its murine ortholog Hsp25 have been shown to play a very important role in the protection of cardiomyocytes under oxidative stress. Ischemia–reperfusion injury and hypertrophy were found to be reduced by small Hsps in the heart and cardiomyocytes [99–106]. Studies have also shown that ROS and lipid peroxidation could be attenuated by the induction and phosphorylation of small Hsps [107, 108]. Dox has been found to show a different effect on the activation of HSF-1 in cells and animal models. HSF-1 activation by heat shock is different from that induced by anticancer drugs such as Dox. Unlike the extensive phosphorylation and transcription of a full spectrum of Hsps produced by heat shock, anticancer drugs induced oxidative stress which induces a selective expression of Hsps. For example the anticancer drug vincristine and vinblastine activate HSF-1 by selective phosphorylation via the c-JNK pathway [109]. HSF-1 has also been implicated in MDR-1 expression in Dox-resistant cells. However, with the exception of our initial report [110], the Dox-induced activation of HSF-1 in the heart or in cardiomyocytes has not been studied in detail. Such an activation of HSF-1 is expected to have a wide range of implications and needs to be studied in detail to unravel the complete mechanism of Dox-induced cardiotoxicity.

5 Mechanism of Heat Shock Protein Cardioprotection Against Dox Toxicity Small Hsps have been found to play many critical roles in cardioprotection. Various in vitro and in vivo models have shown convincingly that these proteins are vital in protecting the structural integrity of myocardium, in myogenesis, and in preventing various cardiovascular diseases. Gene deletion/mutations in both the HSF-1 as well the small Hsps such as aB-crystallin or Hsp25 (the murine ortholog of human Hsp27),

Dox and Small Hsps in the Heart

113

resulted in CHF with age [111]. However, very little insight has been gained on the underlying mechanism of how small Hsps provide cardioprotection against oxidative stress including that caused by Dox. Using in vitro cardiomyocyte cultures and in  vivo small animal models, various groups have studied this problem. These ­studies have examined the issue from two angles (1) studies on the induction mechanism of the small Hsps, such as Hsp27, due to Dox-induced oxidative stress; and (2) studies on the effect of preinduction of Hsps by activation of HSFs due to heat shock and other activation procedures or by overexpression in mice. In neonatal cardiomyocytes-derived H9c2 cells, Dox was found to enhance Hsp27 expression [92]. In low concentrations, Dox-induced toxicity was higher and with increase in Dox concentration its survival reached a leveling off (Fig. 3a). Hsp27 induction increased in parallel with increasing Dox concentrations, suggesting that the Dox-induced Hsp27 offers protection from the Dox in these higher concentrations. Moreover, in preheat shocked cells (heat shock for 4  h at 42°C and incubated for the next 24 h at 37°C in order to activate HSF-1 and induce the Hsps), where the basal Hsp27 was higher (Fig. 3b), the cells showed increased resistance against Dox-induced toxicity [92]. EMSA analysis of HSF-1 binding to consensus HSE showed increased DNA binding of the transcription factor (Fig. 3c). Overall, it appears that Hsp27 induction in response to Dox treatment is likely a response aimed at protecting the cells (Fig.  3d). However this study did not establish a definitive mechanism for either the increased Hsp27 expression, or how Hsp27 offered protection from Dox. In another study, we saw that heat shock can protect the cells from mitochondrial-based apoptotic pathways, attenuating OH radical generation from mitochondrial aconitase [112]. We found that the deactivation of aconitase and subsequent generation OH radicals from the free Fe is reduced in Hsp27 overexpressing cells. Thus, it seems that the heat shock-induced activation of HSF-1 and induction of Hsps play important roles in retaining the mitochondrial integrity, even in the presence of Dox. While this could be one of the possible mechanisms, it is clear that there may be additional mechanisms involved. We recently reported that Dox-induced Hsp25 and its phosphorylation may protect myocytes by modulating p53 activity [110]. In cancer cells, Hsp25 overexpression is known to increase p53 degradation through increased ubiquitination and protect the cells by triggering cell cycle arrest/repair mechanism. Also Hsp25 is known to protect Akt and enhance its phosphorylation to activate p21, to induce cell cycle arrest/release and cell survival. As such, the question arises whether a similar Hsp25/p53 interaction is possible in the heart. However, the mechanism cannot be simply extended to cardiomyocytes because cardiomyocytes are terminally differentiated cells, and thus, unlike in cancer cells, there is no active cell cycling in cardiomyocytes. Thus Hsp25 cannot protect cardiomyocytes by altering cell cycle regulation, but it can protect cardiomyocytes by affecting p53-regulated apoptosis, by modulating p53 stability. However, studies with the use of cardiac H9c2 cells [113] were required for the initial understanding of Hsp25/p53 signaling upon treatment with Dox. The immortalized cardiac H9c2 cells that undergo active proliferation, and cardiac fibroblasts (with wild-type HSF-1 and HSF-1 knockout), which cannot be directly

114

K. Krishnamurthy et al.

b 100

c

NO UV

6 hr after HT

Heat-shocked

75 HSF-1+HSE

Control

50 25 0

2

4 6 8 DOX (mM)

10

12

NADP+

d

e

2+ /F

Fe

O2

DOX

3+

Reduced DCFDA fluorescence & Free radicals generation

NADPH

Heat Stress

ROS p38MAPK(a)

Hsp27 Oligomers

p38MAPK(b )

Bcl2 Bax Caspase 3

Increased p38MAPK (b) activity

MAPKAP 2

Phosphorylation

Cell Viability (rel. %)

a

Reduced Bax/BCl-2 S-15 & S-85 phosphorylation & deoligomerisation

APOPTOSIS F-actin remodeling Retention of F-actin networks

Phospho Hsp27 Monomers & dimers

NECROSIS Fig. 3  Toxicity, HSF-1 binding to HSE and Hsp27 expression in Dox-treated cardiac H9c2 cells. (a) MTT assay. Control and heat-shocked cells, at 70–80% confluency were treated with various concentrations of Dox for 6 h followed by incubation in drug-free medium for 24 h and subjected to MTT assay. Data presented as relative percentage of cell viability, with respect to control cells with no drug treatment. In all the Dox concentrations studied (0.25–10.0 mM), significant protection was observed in heat-shocked cells (p  0.05). (c) Nuclear extracts were prepared from control and Dox-treated cells and incubated with HSE, tagged with biotin. Binding was analyzed by blotting of biotin. Compared to control, HSF-1 binding to HSE is higher in Dox-treated mice. Consistently, addition of HSF-1 antibody (ab) shifted the band. (d) Schematic illustration of the proposed mechanism of protection from Dox toxicity. Doxinduced cell death occurs through activation of stress-induced protein kinases and subsequent signal transduction for apoptosis. On the other hand, hyperthermia also induces similar stressinduced kinases and small heat-shock proteins, which can be phosphorylated by MAP kinase. The phosphorylated small Hsps protect the cells

116

K. Krishnamurthy et al. DOX

Hsp27

ROS

p38MAPK

K

AP

M

p53 p-Hsp27

p53 Bax/Bcl2 p21 Cell death

Bax/Bcl2

Cell Cycle Arrest Cell Survival

Fig.  4  Regulation of p53 transcriptional activity by Hsp27 in Dox-treated H9c2 cells. Dox generates ROS which in turn activates p53. Activation of p53 results in increased transcription of the Bax and increases the p53 depended apoptosis and cell death [114]. However, increased expression of Hsp27 (by modes such as heat shock, oxidative stress or any other mode of HSF-1 activation) and phosphorylation by MAP kinase, results in p53 binding and increased transcription of p21. Such an increased transcription of p21 results in cell cycle arrest, DNA repair and cell survival [113]

Hsp25 was found to stabilize the p53 and transactivate it to increase Bax transcription. Thus, in the heart, Dox treatment increases accumulation of Hsp25 due to activation of HSF-1 and the accumulated Hsp25 transactivates p53 to increase the transcription of proapoptotic protein Bax. Although several aspects of the mechanism of Dox-induced heart failure, using various cellular and animal models, and human heart studies [9] have been determined, none of these studies has revealed the complete role of HSF-1 activation and dynamic induction of Hsp25 in Dox-treated hearts, as a response to the oxidative stress. Our study [110] was the first to demonstrate a link between HSF-1 activation and Dox-induced heart failure, and that Hsp25 could be triggering a signaling cascade causing the loss of cardiomyocytes. More importantly, HSF-1–/– showed increased survival against Dox, indicating that targeting HSF-1 or suppression of Hsp25 accumulation may be a potential therapeutic approach to prevent heart failure in Dox-treated patients. Very recently induction of proapototic Bax increase by increased activation of p53 was Fig. 5  (continued) increased more than 10 times in Dox-treated samples. Cleaved PARP-1 (89 kDa band) was observed to be higher in Dox-treated heart lysates due to increased Caspase-3 activity, consistent with higher Bax. (f) Schematic of the HSF-1 activation and cardiomyocyte loss in the heart upon Dox treatment. Dox-derived ROS activates HSF-1 to increase the transcription of Hsps, including Hsp25. Due to chronic oxidative stress, Hsp25 is phosphorylated by p38-MAP kinase (MAPKAP-2) and associated with p53. Then p53 activates the transcription of the proapoptotic protein Bax, which triggers apoptosis via the mitochondria-dependent signaling [110]

117

**p = 0.001

0.0

Bax Bcl2

0.4

MDM2

0.0

p21

DOX

f

**

0.8

Con

s-86 p-Hsp25

1.2

e

Con

d

1.0

DOX

0.0

2.0

Con

0.4

P-s 86

DOX

Hsp90

0.8

**

3.0

DOX

p-s 15

1.2

Con

Hsp70

c s-15 p - Hsp25

1.6

Hsp25 (au)

Total

**

b

DOX

Con

DOX

a

Con

Dox and Small Hsps in the Heart

Fig. 5  Hsp25 overexpression in Dox-treated failing hearts of mice. (a) Representative western blots of total, s-15 and s-86 phosphorylated Hsp25, Hsp70 and Hsp90. (b–d) Quantitative densitometric plots of total and s-15 and s-86 phosphorylated Hsp25, respectively, normalized with respect to GAPDH (n = 5). The error bars are the SEM. (e) Western blots of various p53 target proteins. Among the p53 target proteins probed, namely Bax, p21, MDM2, Bcl2, only the Bax

118

K. Krishnamurthy et al.

independently reported [117]. This paradoxical observation, compared to the findings in in vitro secondary cardiomyocytes-derived cell lines, may be due to the difference in the activation mechanism of p53. Perhaps in cardiac cell lines, p53 preferably enhances the transcription of cell-cycle-arrest proteins/survival, as in the case of cancer cells, whereas in the terminally differentiated cardiomyocytes the same p53 is enhancing the transcription of proapoptotic proteins. This hypothetical duality of the function of p53 requires further experiments to confirm. On the other hand, small Hsps overexpressing mice such Hsp27 or Hsp20 overexpressing mice also showed improved cardiac function compared to their wild-type counterparts [118, 119]. In Human Hsp27 overexpressed mice, Liu et al reported an improved cardiac function compared to wild-type litter mates [120]. They found far-reduced ROS generation and protein carbonylation in these mice. In another study, overexpression of Hsp20 was found to improve cardiac function and reduced oxidative stress upon treating with Dox [118]. Although these studies reported overall protection of the heart from Dox-induced toxicity, the actual mechanism of such protection still remains elusive. The signaling pathways that are associated with such a protective mechanism have begun to emerge. Several of these studies determined the magnitude ROS generation in Hsp27 overexpressing mice or cell culture. We recently determined that the redox reaction of Dox-generated ROS could be minimized in cardiac cells [92]. The EPR spectra of DMPO spin adducts obtained from cells treated with various concentrations of Dox showed spectral features that are similar to the DMPO-OH adduct, indicating that the drug treatment causes generation of OH radicals either as a direct product or that there is iron and H2O2 for sets Fenton reaction. It is important to note that the magnitude of this spectrum decreased for heat-shocked cells as shown in Fig. 3d [92]. The quantitative plots have indicated lower values for the trapped radicals in heat-shocked cells, even in the higher Dox concentration range, indicating that the free radical burst generated by Dox is reduced by heat shock. Further studies on the nature of the free radicals generated (•OH or O2•−), their correlation to the magnitude of Hsp27, and the magnitude of the Dox-induced injury are needed. Another important observation that we have made is that p53 and s-86-phosphoHsp25 coimmunoprecipitate, indicating that phosphorylation of Hsp25 induces an interaction with p53, and in Dox-treated hearts this association is higher (Fig. 5). This observed Hsp25/p53 interaction was previously unknown, especially in the heart; however it was recently shown that HSF-1 knock out suppressed cutaneous tumerogenesis [70] and it was proposed that HSF-1 is a multifaceted factor that regulates p53 transcriptional activity. Hsps are also known to interact with many other receptor molecules. For example, recent studies have demonstrated that TLRs are activated by endogenous signals, including Hsps and oxidative stress, which may contribute to CHF [121]. TLRs are members of the interleukin-1 receptor family and are involved in the responsiveness to pathogen-associated molecular patterns. TLRs were also found to be taking part in the Dox-induced heart failure. Cardiac dysfunction, induced by a single injection of Dox (20  mg/kg IP) into wild-type (WT) mice and TLR-2-knockout (KO) mice, showed better cardiac function TLR-2 KO mice than WT mice [121]. Nuclear factor-kappaB activation and production of

Dox and Small Hsps in the Heart

119

proinflammatory cytokines after Dox were also shown to be suppressed in TLR KO mice compared with WT counterparts. Survival rate was significantly higher in TLR-2 null mice than in WT mice 10 days after Dox injection. These findings suggest that TLR-2 may play a role in the regulation of inflammatory and apoptotic mediators in the heart after Dox administration. These results agree, in general with the hypothesis that Dox-induced apoptosis is primarily responsible for the loss of cardiomyocytes in the heart, that eventually lead to heart failure. However, it was recently reported that activation of the mammalian target of rapamycin, mTOR, protects the heart from Dox-induced heart failure. These authors have proposed that apoptosis could only be a secondary effect of the activated mTOR signaling in the overall pathogenesis [122]. However, there is no reported link between Hsp25 to mTOR function and hence further studies are required to delineate the relationship between the Hsp25 and mTOR in Dox-treated hearts.

6 Doxorubicin-Activated MAP Kinase and Small Hsps Phosphorylation Dox-induced ROS not only activate HSF-1 as discussed above, but it also activates protein kinases. ROS, generated by Dox-redox reactions, are aggravated by many factors in cardiac cells. For example, free redox cycling iron species have been found to play important roles in the development of Dox-induced cardiomyopathy [123]. Evidence of iron participation was obtained from the addition of membranepermeable iron chelators and their ability to attenuate the Dox-induced apoptosis. Mitogen-activated protein kinase (MAP kinase) signaling pathways are the primary intermediator of induction of apoptosis by oxidative stress. There are three major MAP kinase families, including extracellular signal-regulated kinases (ERKs), p38, and c-Jun NH2-terminal kinases (JNKs). In the cardiovascular system, ERK1/2 are potently and rapidly activated by growth factors and hypertrophic agents, thereby mediating cell survival as well as offering cytoprotection [124, 125]. Conversely, JNKs and p38-MAP kinases are activated by cellular stresses, including oxidative stress, and are thought to correlate with cardiomyocyte apoptosis and cardiac pathologies [126–128]. It has also been reported that inhibition of p38-MAP kinase by selective cardiac-specific overexpression of cysteine-rich metallothioneins reduced the Dox-induced apoptosis in the heart [129]. Interestingly, Hsps, which are induced as a stress response, have been found to be phosphorylated by MAP kinase activated protein-2 (MAPKAP-2) (downstream of p38-MAP kinase), and phosphorylated Hsps prevent apoptosis [130–133]. Elaborate studies have been carried out on the phosphorylation of small Hsps by different MAP kinases. Since MAP kinases are also activated by Dox, we have carried out studies on how these kinases especially MAPKAP-2 can phosphorylate Hsp27 in the Dox-treated cardiac H9c2 cells [92]. Similarly, proteomic analyses of Hsp25 in the Dox-treated hearts have shown that extensive oligomerization of Hsp25 occurs when the protein is overaccumulated due to activation of

120

K. Krishnamurthy et al.

HSF-1 [110]. Phosphorylated Hsp25 is increased in Dox-treated hearts (Fig.  5); however, the ratio of phosphorylated Hsp25 to nonphosphorylated Hsp25 is not altered. At the same time, various MAP kinases are also activated due to oxidative stress, and p38-MAP kinase has been shown to phosphorylate the downstream target MAPKAP-2 which in turn phosphorylates Hsp25 [134]. Indeed phosphorylation of Hsp25/27 determines aggregation/disaggregation and functionalization of Hsp25/27 phosphorylation of Hsp25 in Dox-treated mouse hearts. Although the aggregation/disaggregation behavior of Hsp25 has been found to be regulated by phosphorylation of Hsp25, the observation of six spots in the 2D western blots separated by very narrow pI (observed almost as a continuum) could not be interpreted based on phosphorylation status alone [92]. Similar to our observation of multiple spots in 2D western blots, the Dohke research group observed an increase in the number of Hsp27 spots in 2D western blots of congestive failure hearts lysates, caused by increased pacing (tachycardia) [135]. Three prominent Hsp27 spots were observed in control hearts. These three spots were assigned to nonphosphorylated, monophosphorylated and diphosphorylated Hsp27 [136, 137]. In some studies Pro-Q staining (a marker of phosphorylated proteins) have stained all three of the spots, and these studies have assigned these spots as mono-, di- and tri-phosphosphorylated Hsp27 [135]. However, our study using MS/MS and phospho-specific antibodies has confirmed that all three spots were predominantly nonphosphorylated; rather, only a fraction of each spot represents a phosphorylated species in these aggregates [92]. This indicates that separation of Hsp27 in the first dimension (i.e., pH gradient) of the 2D gel is primarily determined by Hsp25 concentration and local pH dependent aggregation (especially at pH