OUP user menu

Correlation of oxidative stress with activity of matrix metalloproteinase in patients with coronary artery disease

Kunihiko Kameda , Toshiro Matsunaga , Naoki Abe , Hiroyuki Hanada , Hiroshi Ishizaka , Hirotsugu Ono , Masayuki Saitoh , Kozo Fukui , Ikuo Fukuda , Tomohiro Osanai , Ken Okumura
DOI: http://dx.doi.org/10.1016/j.ehj.2003.09.022 2180-2185 First published online: 2 December 2003

Abstract

Aim Oxidative stress is implicated in the progression of heart failure, and matrix metalloproteinase (MMP) activity is increased in patients with congestive heart failure. We examined the role of oxidative stress on MMP activity in humans.

Methods and results We measured the MMP activity and the level of 8-iso-prostagandin F2α (8-iso-PGF2α), a specific and quantitative maker of oxidant stress, in the pericardial fluid (PF) in 47 consecutive patients with coronary artery disease who underwent coronary artery bypass surgery. Zymography of PF showed bands at 92–85kDa (MMP-9) and 72–65kDa (MMP-2). The MMP activity was expressed as the ratio to MMP-2 standard. MMP-2, MMP-9 and total gelatinolysis activities were positively correlated with left ventricular end-diastolic volume index (LVEDVI), and MMP-2 and total gelatinolysis activities were also positively correlated with LV end-systolic volume index. Moreover, MMP-2, MMP-9 and total gelatinolysis activities were all positively correlated with pericardial level of 8-iso-PGF2α. Also, LVDEVI was positively correlated with pericardial level of 8-iso-PGF2α.

Conclusion Oxidative stress may play an important role in the regulation of MMP activity. Augmented MMP activity may be involved in the development of ventricular remodelling in patients with coronary artery disease.

  • Matrix metalloproteinase
  • Oxidative stress
  • Ventricular remodelling

1 Introduction

Ventricular remodelling, frequently observed in patients with large myocardial infarction, is an important contributory factor in the progression to congestive heart failure (CHF).1–3Oxidative stress has been implicated in the pathogenesis of CHF.4–6Mallat et al.6reported that pericardial level of 8-iso-prostaglandin (PG) F2α, a specific and quantitative biochemical marker for oxidative stress in vivo,7is increased with the functional severity of CHF. In addition, antioxidant agents such as allopurinol and carvedilol were shown to improve cardiac function in human failing heart.5,8

On the other hand, matrix metalloproteinases (MMPs) were shown to contribute to ventricular remodelling in experimental myocardial infarction model and pacing-induced heart failure model.9–11Increased MMP zymographic activity and species expression were also demonstrated in patients with end-stage dilated cardiomyopathy.12Taken together both oxidative stress and MMP activity seem to be involved deeply in the pathogenesis of CHF. Rajagopalan et al.13reported that both MMP-2 (72-kDa gelatinase A) and MMP-9 (92-kDa gelatinase B) activities in cultured smooth muscle cells are increased by reactive oxygen species produced by macrophage-derived foam cells. It is unclear, however, how oxygen stress and MMP activity are related to each other during the process of ventricular remodelling in humans. To clarify it, we measured the pericardial levels of MMP activity and 8-iso-PGF2α, indicative of their tissue levels, in patients with coronary artery disease who underwent coronary artery bypass graft surgery (CABG).

2 Methods

2.1 Patient's profiles

This study enrolled 47 consecutive patients with coronary artery disease (15 with previous myocardial infarction, 16 with unstable angina pectoris, and 16 with stable effort angina) who underwent CABG (male, n=36; female, n=11). The mean age of the patients was 67±1 years (mean±SEM). Nine patients had left main tract disease with double (n=7) or triple vessel disease (n=2), six double vessel disease and 32 triple vessel disease. Seventeen patients were diagnosed as diabetes mellitus, but none of the patients had any history of neoplastic, hepatic, infectious and autoimmune disease.

Before CABG, cardiac catheterization including coronary angiography was performed in all patients and biplane left ventriculography (LVG) in 40. Two cardiologists who wereunaware of the results of the MMP activity and 8-iso PGF2α level analysed the LVG. Ventricular silhouettes in the 30° rightanterior oblique and 60° left anterior oblique projections were digitized with a LVG analysis system (Cardio 500, KontronInstruments, Eching, Germany), and left ventricular end-diastolic (LVEDVI), end-systolic volume index (LVESVI) and ejection fraction (LVEF) were determined.

Just before the surgery, 19 patients were treated with furosemide, nine with spironolacton, 13 with angiotensin converting enzyme inhibitor, eight with angiotensin type 1 receptor blocker, 22 with beta-blocker, four with Ca channel antagonist and nine with HMG-CoA reductase inhibitor. All patients gave their written, informed consent. The study protocol was approved by the ethics committee on human research at our institution.

2.2 Sampling pericardial fluid during surgery

Immediately after a small incision of the pericardium, undiluted pericardial fluid was obtained. The samples were collected in sterile tubes and immediately placed on ice. The average amount of non-haemorrhagic pericardial fluid obtained from each patient was 3.2±0.3ml. After clarification of the cellular components by centrifugation at 3000×g for 10min at 4°C, these samples were rapidly stored at −80°C until use.

2.3 SDS-PAGE zymography

We performed the quantitative gelatin zymography.14Pericardial fluid samples (3μl) were separated by dilution into zymography sample buffer. The samples and MMP-2 standard (1.25×10−3unit/lane, Wako chemical, Japan) were loaded into the wells of a 7.5% gelatin gel and electrophoresed. The gel was removed and incubated for 1h at room temperature in 100ml of renaturing buffer (2.5% Triton X-100) on a rotary shaker. The buffer was decanted and replaced with 100ml of development buffer(50mM Tris, pH7.5, 200mM NaCl, 5mM CaCl2, 1μM ZnCl2, 0.02% Briji-35). The gel was incubated at 37°C for 18h. Each gel was stained with 100ml of 0.5% Coomassie blue G-250 in 30% methanol, and 10% acetic acid for 3h and then de-stained with three changes of 30% methanol, and 10% acetic acid. The gels were digitized using a scanning digitizing system andanalysed using NIH Image software (product of density and band area). Zymography of pericardial fluid showed bands at 92–85kDa (MMP-9) and 72–62kDa (MMP-2). In the preliminary study, we identified the 72kDa band as MMP2 and the 92kDa band as MMP-9 by Western blotting. The MMP-2 and MMP-9 activities were expressed as the ratio to MMP-2 standard to avoid the differences among gels.

Fig. 1

(A) Representative gelatin zymograms loaded by 1μl, 3μl and 9μl of pericardial fluid samples and MMP-2 standard (MMP-2 STD) at each lane. Zymography of pericardial fluid showed bands at 92–85kDa (MMP-9) and 72–62kDa (MMP-2). (B) Relationship between gelatinolysis activity with MMP-2 and pericardial fluid volume. (C) Relationship betweengelatinolysis activity with MMP-9 and pericardial fluid volume.

2.4 Measurement of 8-iso PGF2α

The 8-iso-PGF2α concentration in the pericardial fluid was measured by a specific enzyme immunoassay (EIA) kit (Cayman Chemical, Ann Arbor, MI) as previously described.15The antiserum used in this assay has a 100% cross reactivity with 8-iso-PGF2α, 0.2% with PGF2α, PGF3α, and PGE1 and 0.1% with 6-ketoPGF1α. The detection limit of the assay is 4pg/ml. We also measured the levels of tissue inhibitor of metalloproteinase (TIMP)-1 and TIMP-2 in pericardial fluid and B-type natriuretic peptide (BNP) in the plasma by the commercial available EIA kits (EIA kits of TIMP-1 and TIMP-2 from Fuji Chemical; and BNP from Shionogi).

2.5 Statistical analysis

All results are expressed as mean±SEM. Comparison of the variables between two patient's groups were performed using unpaired Student t test. Linear regression analysis was used to determine the relationship between MMP activity and each of LVEDVI, LVESVI and LVEF and between MMP activity and 8-iso-PGF2α level. A P value <0.05 was regarded as a significant.

3 Results

3.1 Measurement of MMP activity by gelatin zymography

Fig. 1shows sample volume-dependent increase in MMP activity with gelatin zymography. We loaded 1μl, 3μl and 9μl of pericardial fluid samples obtained from five patients at each lane. Gelatinolysis activities with MMP-2 and MMP-9 were increased in a volume-dependent manner in all five patients. Thus, gelatin zymography was useful for the quantitative and simultaneous measurement of MMP-2 and MMP-9 activities in human pericardial fluid.

3.2 Relationship between MMP activity and left ventricular volume

Mean MMP-2, MMP-9 and total gelatinolysis (MMP-2+MMP-9) activities were 0.471±0.025, 0.070±0.012 and 0.541±0.032, respectively. Fig. 2shows relationships between MMP activity and LVEDVI, LVESVI and LVEF. MMP-2 activity was positively correlated with LVEDVI (r=0.61, P<0.0001) and LVESVI (r=0.53, P<0.001). MMP-9 activity was positively correlated with LVEDVI (r=0.40, P<0.05). However, any MMP activity was not correlated with LVEF.

Fig. 2

Relationship between MMP activities and left ventricular end-diastolic volume index (LVEDVI, ml/m2), end-systolic volume index (LVESVI, ml/m2) and ejection fraction (LVEF,%). MMP-2 activity was positively correlated with LVEDVI and LVESVI (A). MMP-9 activity was also positively correlated with LVEDVI (B).

To analyse whether MMP activity was correlated with any clinical characteristics and haemodynamic parameters, we divided the patients into two groups, i.e., patient group with total gelatinolysis activity ⩾0.54 (high group, n=22), and total gelatinolysis activity <0.54 (low group, n=25) and compared the characteristics between the two groups. Table 1shows the patient characteristics of each group. Both LVEDVI and LVESVI were significantly greater in high group than those in low group. Any of the age, haemodynamic parameters, blood chemistry, pericardial levels of TIMP-1 and TIMP-2, and plasma level of BNP was not different between the two groups except for HDL cholesterol level. Fig. 3A shows example of zymography of MMP-2 and MMP-9 in patients with high and low activities.

View this table:
Table 1

Matrix metalloproteinase activity and 8-iso-PGF2α level on pericardial fluid

High (n=22)(total gelatinolysis activity ≥0.54)Low (n=25)(total gelatinolysis activity <0.54)P value
Age68±2a67±20.6482
Total gelatinolysis activity0.700±0.0440.399±0.017<0.0001
MMP-20.608±0.0260.349±0.022<0.0001
MMP-90.092±0.0240.050±0.0800.0864
8-iso-PGF2a (pg/ml)57±826±4<0.001
Haemodynamics
Mean BP (mmHg)100±3102±50.7299
Heart rate (beats/min)74±567±50.5048
Cardiac index (l/min/m2)2.4±0.12.4±0.20.7154
PCW (mmHg)11±18±10.0807
LV volume
LVEDVI (ml/m2)108±682±4<0.001
LVESVI (ml/m2)61±541±4<0.01
LVEF (%)45±342±30.091
Blood chemistry
Total cholesterol (mg/dl)180±8184±70.7536
HDL-cholesterol (mg/dl)37±246±30.0233
LDL-cholesterol (mg/dl)115±7111±70.6862
Triglyceride (mg/dl)143±21126±150.5119
Blood sugar (mg/dl)125±8126±90.9180
BNP (pg/dl)249±61133±470.1372
TIMP level in PE
TIMP-1 (μg/ml)1.8±0.11.7±0.10.4264
TIMP-2 (ng/ml)78±1192±90.3475
  • PCW; pulmonary capillary wedge pressure, BNP; B type natriuretic peptide; TIMP; tissue inhibitor of metalloproteinase.

  • a Values are mean±SEM.

Fig. 3

(A) Representative gelatin zymographies in patients with high gelatinolysis activity ≥0.54 (left) and low gelatinolysis activity <0.54 (right). (B) Pericardial fluid level of 8-iso-PGF2α (pg/ml) in high (open bar) and in low gelatinolysis activity groups (closed bar). (C) Relationship between 8-iso-PGF2α level and MMP-2 activity. (D) Relationship between 8-iso-PGF2α level and MMP-9 activity. *P<0.01 vs high group.

3.3 Relationship between MMP activity and 8-iso-PGF2α in pericardial fluid

Pericardial level of 8-iso-PGF2α (pg/ml) was significantly greater in high gelatinolysis activity group (57±8) than that in low group (26±4) (P<0.001) (Fig. 3B). Both MMP-2 and MMP-9 activities were all positively correlated with pericardial level of 8-iso-PGF2α (MMP-2, r=0.55, P<0.0001; MMP-9, r=0.52, P<0.001) (Fig. 3C, D, respectively). Moreover, LVDEVI, and not LVESVI or LVEF, was also positively correlated with pericardial level of 8-iso-PGF2α (r=0.34, P<0.05).

4 Discussion

4.1 MMP activity in the pericardial fluid

Gelatin zymography is an established method for the quantitative analysis of both MMP-2 and MMP-9 activities.14We estimated MMP activity in the pericardial fluid obtained from the patients with coronary artery disease because many studies demonstrated that the pericardial fluid was useful to investigate the severity of ischaemic or failing heart in humans.6,16,17Recently, Etoh et al.18reported that MMP activity in myocardial interstitial fluid was increased after acute myocardial infarction in a canine model. Their report suggests that pericardial fluid sample is suitable for the measurement of tissue level of MMP activity in humans. In the present study, gelatinolysis activity with MMP in the pericardial fluid was increased in a volume-dependent manner. Our findings indicate that gelatin zymography can be used for quantitative and simultaneous analysis for the activities of MMP-2 and MMP-9 in human pericardial fluid.

4.2 Role of oxidative stress in ventricular remodelling

Ventricular remodelling is associated with increased morbidity and mortality in patients with myocardial infarction and heart failure.1–3Oxidative stress is related to the progression of heart failure and atherosclerotic plaque instability.6,13Oxidative stress is increased in the ischaemic myocardium and failing heart.6,7,19Mallat et al.6reported that pericardial level of 8-iso-PGF2α was increased with the functional severity of CHF. In the present study, 8-iso-PGF2α was positively correlated with LVEDVI. Previous and our present results indicate that oxidative stress plays an important role in the development of ventricular remodelling and progression to heart failure.

4.3 Role of oxidative stress in MMP activity

Matrix metalloproteinases, endogenous zinc-dependent enzymes, are important for progression of atherosclerosis, plaques vulnerability and coronary angiogenesis.20–24Furthermore, many experimental studies indicated that MMP contributed to ventricular remodelling in myocardial infarction and pacing-induced heart failure.9–11Clinically, MMP zymographic activities were increased in LV myocardium in patients with end-staged dilated cardiomyopathy.12Miyamoto et al. also reported that pericardial levels of MMP-2, which were measured by enzyme immunoassay, increased in association with an increase in LV volume and with a decrease in LV function in patients with coronary artery disease.17In the present study, MMP-2 and total gelatinolysis activities were positively correlated with LVEDVI and LVESVI, and MMP-9 was also positively correlated with LVEDVI. It is known that MMP-2 and -9 are inhibited by TIMP-2 and TIMP-1, respectively.25However, the pericardial expression levels of TIMP-1 and -2 were not different between higher and lower gelatinolysis activity groups. The present results suggest that the increases in MMP activities are related to the progression of ventricular enlargement (remodelling) in patients with coronary artery disease.

According to previous and present results, both oxidative stress and MMP activity seem to be involved deeply in the pathogenesis of ventricular remodelling. Rajagopalan et al.13reported that reactive oxygen species was important for the activities of MMP-2 and MMP-9 in cultured smooth muscle cells. In the present study, both MMP-2 and MMP-9 activities were all well positively correlated with the level of 8-iso-PGF2α inthe pericardial fluid. Thus, oxidative stress, which is increased by myocardial ischaemia or mechanical strain following ventricular dysfunction, may upregulate the activities of MMPs, resulting in the development of ventricular remodelling. Recently Nakamura et al reported that carvedilol decreased the elevated oxidative stress in human failing myocardium.8It is also known that carvedilol improves the morbidity and mortality in patients with chronic heart failure.26These mechanisms may be caused by the suppression in MMP activity resulted from the reduction in oxidative stress with carvedilol.

In conclusion, oxidative stress seems to play an important role in the regulation of MMP activity, and augmented MMP activity may be involved in the development of ventricular remodelling in patients with coronary artery disease. Anti-oxidant therapy may be a strategy for the prevention from ventricular remodelling and congestive heart failure via reduction of MMP activity in the heart.

Acknowledgments

The authors would like to thank Dr Hiroshi Osawa, Dr Shingen Owada, Dr Hirofumi Tomita, Dr Koichi Oikawa, The Second Department of Internal Medicine, Hirosaki University School of Medicine, for advising us in the technique of Zymography.

References

View Abstract