Dox is a potent broad-spectrum anticancer drug and is crucial to the treatment of variety of solid and hematologic tumor such as acute leukemia, Hodgkin’s and non-Hodgkin’s lymphoma, and breast cancer (Takemura and Fujiwara, 2007). However, the well-known chronic cardiomyopathy of Dox greatly limits the usefulness. It may causes dilated
cardiomyopathy in a dose-dependent manner and lead to heart failure in patients even after a long duration of discontinuous administration (Singal et al., 1997). Therefore, we attempted to utilize Dox to exam the hypothesis that alterations of ECM play a role in Dox-induced
chronic cardiomyopathy and may be influenced by MMPs/TIMPs system. To our best knowledge, this issue has been studied rarely to this day.
In the present study we used a well-established murine model of Dox-induced
cardiomyopathy. A low dosage administration by 4 mg/kg Dox hydrochloride per week was performed (Figure 1). Ultimately, five weeks treatment produced a 20 mg/kg accumulated dose and was equivalent to 1400 mg in a 70-kg man. In patients, the usual cumulative maximum dosage is about 500 mg/m2, or 1000 mg for a 70-kg man. Beyond this dose, the frequency of cardiomyopathy rapidly escalates (Yi et al., 2006). Although the accumulated dose is highly related to the severity of cardiomyopathy, the greater dosage can lead to higher mortality due to several adverse effects such as diarrhea, weight loss, colitis, and bone
marrow suppression (Van Vleet et al., 1979; Speth et al., 1988; Christiansen and Autschbach, 2006). In this study, the mortality came to about 30% during 12 weeks in Dox-treated mice.
Most of them died in non-treated period, which reflected the long-term toxic effect of Dox.
Our results showed that Dox treatment in mice significantly reduced body weight (Figure 2), ventricle weight and the ventricle weight/tail length ratio (Table 6). Changes in body weight may result from the general toxicity of Dox (Klimtova et al., 2002), reduction of
food intake and inhibition of protein synthesis (Tong et al., 1991). The tail length of a rodent is a better indicator in malnutrition than body weight as a control (Kizaki et al., 2006).
Therefore, we used ventricle weight/tail length ratio to estimate the Dox-induced
cardiotoxicity. The decreases of ventricle weight and the ventricle weight/tail length ratio began from week 3 and further descended later, which may be the result of cardiomyocytes damage, Dox-induced apoptosis (Narula et al., 1996; Sharov et al., 1996) and inhibition of myocardial protein synthesis, confirming the prolonged cardiotoxic effects (Minotti et al., 2004).
Alteration of electrocardiogram on Dox-treated mice had been demonstrated previously (van Acker et al., 1996; Fisher et al., 2005). In our measurement, an increased S-T interval was observed in mice with Dox treatment and had a significant difference compared with control from week 5 initially (Figure 3). The prolongation of the repolarization phase may be related to the prolonged action potential in Purkinje fibers after incubation with Dox (le Marec et al., 1986). In another study, oxygen-derived free radicals generated by Dox can increase the duration of the action potential in isolated myocytes (Jabr and Cole, 1993).
However, the increase of S-T interval in our study did not appear as early as the previous report (van Acker et al., 1996), which began after 2 weeks administration at a cumulative dose of 12 mg/kg. Additionally, the smaller degree of the increasing in this study at the relative time points, these may indicate a more moderate disorder in our mice. On the other hand, these divergences may also result from under anesthesia in different strain of mice.
In addition to ECG, we introduced a second cardiotoxic monitor by measuring the natriuretic peptides mRNA level in LV tissues (Figure 4). Synthesis and secretion of NPs by heart is responses to myocardial stretch due to volume expansion or pressure overload, and they are regarded as good biomarkers for the diagnosis of heart failure (Daniels and Maisel, 2007; Lainscak et al., 2007). The elevation of ANP and BNP in mRNA level in this study
provided evidence as hemodynamic deterioration which usually occur in heart failure (Daniels and Maisel, 2007). Nevertheless, the remarkable increase on ANP with a 1.5 fold slightly induced BNP may again indicate the moderate impairments on ventricular function rather than overt heart failure in our mice, and this difference may not be distinguished from NPs level in plasma (Langenickel et al., 2000).
By histology, the cytoplasmic vacuoles existed apparently in the cardiomyocytes of mice with Dox treatment after 12 weeks (Figure 5), it was consistent with that in patients
(Takemura and Fujiwara, 2007). The morphology of the heart, however, showed no significant difference between the group of treatment and control. Typical dilation,
representing the end stage remodeling of Dox-induced cardiomyopathy, was absent during the 12 weeks development in our study.
The mRNA expression level of type I and type III collagen in LV tissue was surveyed (Figure 6). The results showed no changes on these two predominant ECM components between Dox-treated and control groups. The consistent expression of collagen may result from the constitutional situations; however, we could not exclude the possibility that it was the antagonistic effect of profibrotic stimulations and Dox-induced oxidative stress which may decrease collagen synthesis and the abundance of mRNAs for procollagens (Tanaka et al., 1993; Siwik et al., 2001).
Consideration for ECM degradation, the expression and protein activity level of MMP-1, MMP-2, MMP-9 and TIMPs were measured. MMP-1 is one of the collagenase members which are the main enzymes as well as MT-MMP-1 capable of degrading intact fibrillar collagen (Ohuchi et al., 1997; Ala-aho and Kahari, 2005), suggesting MMP-1 is critical in tissue remodeling. However, it is interesting that this major collagenase in many species including human had been unable to identify in rodent until year 2001 (Balbin et al., 2001).
Before the time, MMP-13 was the main interstitial collagenase in rodent, which had been
considered a counterpart to human MMP-1. Nowadays, the newly finding interstitial
collagenases (Mcol-A and Mcol-B), the most possible orthologue of human MMP-1 finding so far, have been suggested playing roles in reproductive tissues in mice (Nuttall et al., 2004;
Chen et al., 2007). However, few studies in cardiac related issues had been reported. The MMP-1 antibody used in this study recognizes an epitope from amino acid 332-350 of human MMP-1. Sequence of the recognizable site on human MMP-1 is highly similar to mice
MMP-1 rather than MMP-13, and the specificity of the antibody for discrimination between mice MMP-1 and MMP-13 in myocardium had been mentioned in previous study (Seeland et al., 2007). In our finding by immunoblotting, the reduced 57 kDa latent MMP-1 accompanied by the increased 47 kDa active MMP-1 during the Dox treatment (Figure 7). This interesting opposite trends of changes began from the middle stage of the course and further altered later indicating the advanced activation of pro-MMP-1. However, the total MMP-1 protein was unchanged except at week 3. Recently, several studies reported the forming of ROS in the late onset cardiomyopathy even if the fact of intramyocardial short half-life of Dox and its
metabolites (Lebrecht et al., 2003; Lebrecht et al., 2007). Other studies pointed that continuously induced ROS may be involved in MMPs activation by modification of the cysteine on MMPs propeptide domain (Fu et al., 2001; Okamoto et al., 2001; McCarthy et al., 2008). According to above, we supposed that the alteration of MMP-1 protein in our mice with Dox-induced cardiomyopathy was mainly due to posttranslational regulation and the activation may be partially associated with ROS induced by Dox. In addition, our finding may provide an explanation for the previous reports that the relative abundance of 57 kDa MMP-1 was significantly reduced (Thomas et al., 1998) but with increased collagenolytic activity (Gunja-Smith et al., 1996; Tyagi et al., 1996), at least partially contributed from active MMP-1, in patient with DCM. In this study, we found the activation of MMP-1 throughout the myocardium in Dox-induced cardiomyopathy, which may play an important role in ECM turnover and cardiac remodeling as well as human MMP-1 overexpression in mice (Kim et al.,
2000). Nevertheless, the transcript of MMP-1 based on the sequence from Balbin (Balbin et al., 2001) was limited to detection in heart of mice (Nuttall et al., 2004) and neither was our examination by PCR. Thus, the uncertainty of native MMP-1 in mice is necessary to be characterized further.
MMP-2 and MMP-9 so called gelatinases possess the capacity to degrade denatured collagen and a number of components from interstitial and basement membrane (Hijova, 2005). Moreover, intracardiomyocyte disorganization of the contractile apparatus in DCM by gelatinases was also reported (Rouet-Benzineb et al., 1999) suggesting the great importance of these MMPs in heart diseases. In this study, real-time PCR and zymography were used to determine the gelatinases expression level (Figure 9) and enzyme activity (Figure 8), respectively. At the end of the first week, the increased MMP-9 activity was observed with the unchanged MMP-2 mRNA and activity level. The results well connect with the previous studies, in which an early induced MMP-2 activity (Bai et al., 2004) and up-regulation of MMP-2 and MMP-9 mRNA expression in mice with acute Dox treatment were reported (Kizaki et al., 2006). According to the latter, the MMP-2 expression level returned to baseline earlier than MMP-9 at day 4 (Kizaki et al., 2006), it may interpret the later induced MMP-9 activity but not MMP-2 at day 6 in our case. This acute induction of gelatinases may be the respondence to inflammation induced by myocardium damage. In chronic phase, the activity of MMP-2 and MMP-9 was decreased at week 12 following the reduced mRNA expression at week 9. These unexpected results compared to most of the heart failure reports, usually up-regulation on gelatinases in heart failure, may be due to different stage of diseases and variant etiologies. For instance, MMP-2 was increased in nonischemic DCM but was
unchanged in ischemic DCM (Spinale et al., 2000), and decreased MMP-9 mRNA expression with no differences in MMP-2 in DCM was also reported (Batlle et al., 2007). In addition to the reduction of both gelatinases, however, the smaller degree of alteration on MMP-2
displays stable and constitutive expression in myocardium (Spinale, 2007).
The activity of TIMPs was measured by reverse zymography (Figure 10). Three inhibitive bands were identified as TIMP-2,-3 and -4 according to molecular weight, which approximately increased at week 5 and further induced later. This increase of TIMPs at the protein activity was not coincident with the mRNA level (Figure 11) suggesting a role for posttranscriptional regulation.