OUP user menu

Increased gene expression of collagen Types I and III is inhibited by β-receptor blockade in patients with dilated cardiomyopathy

Junsho Shigeyama, Yoshio Yasumura, Aiji Sakamoto, Yoshio Ishida, Tatsuya Fukutomi, Makoto Itoh, Kunio Miyatake, Masafumi Kitakaze
DOI: http://dx.doi.org/10.1093/eurheartj/ehi492 2698-2705 First published online: 4 October 2005


Aims To elucidate the cellular mechanisms of cardioprotection of β-blockers in patients with heart failure, we investigated the effects of β-blockers on collagen synthesis in patients with dilated cardiomyopathy (DCM).

Methods and results We examined the gene expression before and 4 months after the administration of a β-blocker in 17 DCM patients. The messenger ribonucleic acid expression of collagen Types I and III (Col I and III) and transforming growth factor-β1 (TGF-β1) of right ventricular tissues obtained by the endomyocardial biopsy were assessed by quantitative reverse transcriptase–polymerase chain reaction. Cardiac sympathetic nerve activity was assessed by the washout rate (WR) of 123I-metaiodobenzylguanidine from the heart. Left ventricular ejection fraction (21±7 vs. 35±9%) and WR (53±14 vs. 42±13%) improved significantly. Before the β-blocker treatment, the expressions of both Col I (r=0.560, P=0.041) and Col III (r=0.630, P=0.008) genes were correlated with WR. The expression levels of both Col I (1.08±0.72 vs. 0.65±0.26, P=0.024) and Col III (2.06±1.81 vs. 1.05±0.74, P=0.018) were reduced by a β-blocker. Changes in TGF-β1 correlated with those in WR (r=0.606, P=0.002).

Conclusion β-Blockers are considered to inhibit the expression of collagen-related genes in DCM, which seems to be mediated by TGF-β1.

  • Collagen synthesis
  • β-Blocker
  • Myocardial gene expression
  • Heart failure
  • Dilated cardiomyopathy


β-Blockers that decrease the mortality and morbidity in patients with heart failure are recognized to improve systolic ventricular function and reverse cardiac remodelling, in patients with dilated cardiomyopathy (DCM).1 However, the cellular mechanisms by which β-blockers exert these beneficial effects in heart failure are not fully elucidated. Both Lowes et al.2 and Yasumura et al.3 showed that the functional improvement in DCM by the β-blocker treatment was related to normalization of the expression of genes that regulate contractile function, increases in sarcoplasmic-reticulum calcium ATPase and α-myosin heavy chain, and decrease in β-myosin heavy chain of myocytes. The myocardial extracellular matrix, composed of a complex network of structural proteins, mainly collagen Type I (Col I) and Type III (Col III), provides architectural support for the cardiac myocytes and plays an important role in myocardial contractile function4 and cardiac remodelling.5 Therefore, β-blockers are expected to normalize the collagen metabolism in the process of cardiac reverse remodelling.

Hormonal factors, such as angiotensin II, transforming growth factor-β1 (TGF-β1), osteopontin, endothelin-1, and cathecolamines6 promote collagen synthesis. In DCM patients, the increased expressions of genes for both Col I and Col III were associated with a trend of increased expression of TGF-β1.7 In several experimental models, an increase in TGF-β1 messenger ribonucleic acid (mRNA) is reported to precede the increase in collagen mRNA, suggesting that TGF-β1 mediates the profibrotic effects of angiotensin II.79 A blockade of angiotensin II also prevents myocardial fibrosis along with the production of myocardial hepatocyte growth factor (HGF), suggesting that HGF may have a role in the prevention of myocardial injury as a result of angiotensin II blockade.10 Besides the effects of angiotensin II, remodelling of cardiac extracellular matrix has been reported to occur during β-adrenergic stimulation. The continuous β-adrenergic stimulation by either isoproterenol11 or norepinephrine,12,13 increased the expression of collagen mRNA in rat ventricles and the β-blockers are reported to reduce collagen deposition of infarcted rats14 or rats with chronic pressure overload.15

The aim of this study is to examine (1) the factors for the expression of cardiac collagen mRNA and (2) the effects of β-blockers on this expression in patients with DCM.


Study protocol

Thirty four patients referred to the National Cardiovascular Center, Osaka, Japan, because of congestive heart failure and accepted to undergo cardiac catheterization for the diagnosis of DCM, were enrolled. Four patients complicated with obvious hypertension and three patients with suspected secondary DCM were excluded. Three patients who could not tolerate the target dose of β-blockers and seven patients who did not undergo the second cardiac catheterization were also excluded. Finally, 17 patients were included in this study. The average time after the onset of the symptom was 3.1 years (3 months to 6.5 years). After their functional classes being stabilized by conventional therapy (Table 1), at least 1 month elapsed after the acute exacerbation, they underwent cardiac catheterization and were diagnosed as DCM. DCM was diagnosed by the absence of significant coronary artery disease as determined by coronary angiography, the absence of specific heart muscle disease or active myocarditis as determined by endomyocardial biopsy, and the reduction of left ventricular ejection fraction (<40%). Patients with obvious systemic hypertension, diabetes mellitus, a history of excess alcohol consumption, prior myocardial infarction, valvular heart disease, thyroid dysfunction, or other systemic diseases were excluded from this study. Written informed consent was obtained from all patients participated in this study according to a protocol, which was in agreement with the Ethical Committee at our institution.

View this table:
Table 1

Patients' demographics and baseline characteristics

Number (n)17
Age (years)48±11
Sex (male/female)16/1
NYHA classes I/II/III4/12/1
 Furosemide, n(%)14(82)
 Spironolactone, n(%)9(53)
 Digitalis, n(%)9(53)
 ACE inhibitor, n(%)16(94)
 ARB, n(%)1(6)
 Vasodilator, n(%)1(6)
HR (beats/min)75±14
sBP (mmHg)112±12
EDVI (mL/m2)153±32
LVEF (%)21±7
BNP (pg/mL)165±146

IHD, ischaemic heart disease; HR, heart rate; sBP, systolic blood pressure.

Before the administration of a β-blocker, the plasma concentration of brain natriuretic peptide (BNP) was measured, 123I-metaiodobenzylguanidine (MIBG) imaging was performed to measure the activation state of cardiac sympathetic nerve activity,1618 and the gene expressions of right ventricular myocardial samples obtained by the endomyocardial biopsy were assessed by the real-time reverse transcription–polymerase chain reaction (RT–PCR). Both left ventricular end-diastolic volume and ejection fraction were determined by left ventriculography. After cardiac catheterization, the patients were administered either carvedilol (12 patients) or bisoprolol (five patients). Attending physicians administered the β-blocker to the patient, according to the standardized regimen of β-blocker therapy. The initial doses of carvedilol and bisoprolol were 2.5 mg bid and 0.625 mg bid, respectively. These doses were doubled weekly until target dose (20 mg bid of carvedilol and 5.0 mg bid of bisoprolol). Four months after the onset of the treatment, all patients were re-admitted to repeat all of the medical examination performed at baseline. Follow-up MIBG imaging could not be performed in two patients who refused it.

Sixteen out of 17 patients received angiotensin-converting enzyme (ACE)-inhibitor and one patient received angiotensin II receptor blocker (ARB). Enalapril of 2.5–10 mg, quinapril of 10 mg, and temocapril of 2 mg daily were administered in 14, one, and one patients, respectively. Losartan as ARB was used at a dosage of 25 mg daily in one patient.

MIBG imaging

After an overnight fast, patients underwent cardiac MIBG imaging. Anterior planar and single photon emission computerized tomography images were acquired 15 min (early) and 180 min (delayed) after injection of 111 MBq of general MIBG. A scintillation camera (GCA 901A/HG, Toshiba, Tokyo, Japan) with a parallel-hole general purpose collimator (low energy, high resolution type) was used. The camera was interfaced to the digital data acquisition system (GMS 5500 UI, Toshiba). The energy window of 123I was centred at 159 keV with a 20% window. Anterior planar images were acquired using 512×512 matrix format for 180 s. Single photon emission computerized tomography images were reconstructed from a total of 30 projection images over 180° in 6° increments with 30 s per view using 64×64 matrix format. Global cardiac MIBG uptake was determined by the heart-to-mediastinum (H/M) count ratio on the planar image in both early and delayed imaging. Cardiac washout rate (WR) of MIBG was also determined from the equation {([H]−[M])early−([H]−[M])delayed}/{([H]−[M])early}×100%, 16–18 as an index of cardiac sympathetic nerve activity.

Extraction of total RNA and synthesis of cDNA

We obtained the myocardial samples by endomyocardial biopsy from right ventricle. All of the myocardial samples were immediately frozen in liquid nitrogen and then stored at −80°C until use. The samples were homogenized in 1.0 mL ISOGENTM reagent (Nippon Gene, Tokyo, Japan), thoroughly mixed with 0.2 mL chloroform, and centrifuged at 15 000 g for 15 min at 4°C. The aqueous supernatant was transferred into a micro test tube, mixed with 0.6 mL isopropanol, and centrifuged at 15 000 g for 15 min at 4°C. The precipitated total RNA was rinsed with 70% ethanol, air-dried, and then resuspended in RNase-free water. About 2 µg total RNA was treated with DNase FreeTM reagent (Ambion, Austin, TX, USA) for 60 min, and then reverse-transcribed with Superscript IITM (Invitrogen, Carlsbad, CA, USA) at 37°C for 60 min using random primers (TaKaRa, Tokyo, Japan). The integrity of each complementary deoxyribonucleic acid (cDNA) mixture was checked by polymerase chain reaction (PCR) for the cDNAs of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and dystrophin, using ExTaq (TaKaRa, Tokyo, Japan) and the following primer sets: 5′-ACCACAGTCCATGCCATCAC-3′/5′-TCCACCACCCTGTTGCTGTA-3′ and 5′-GTACAAGAGGCAGGCTGATG-3′/5′-CTGAGCTGGATCTGAGTTGG-3′, respectively.

Oligonucleotides of primers and probes

Published cDNA sequences for human ACE, angiotensinogen, angiotensin II Type 1 receptor, mineralcorticoid Type 1 receptor, TGF-β1, HGF, and Col I and III were used for construction of the primer sets and TaqMan probes as described subsequently. By Primer ExpressTM software (Applied Biosystems, Foster, CA, USA), several sets of primers were designed for each gene. The primer set with most efficient amplification of a target cDNA by PCR was selected for final use, which was estimated by conventional electrophoresis and ethidium-bromide staining. Subsequently, the TaqMan probe inherent to each primer set was prepared, which is an oligonucleotide labelled with a reporter dye (FAM) at the 5′-end and a quencher dye (TAMRA) at the 3′-end. The nucleotide sequences of the primers and corresponding TaqMan probes used in this study are shown in Table 2.

View this table:
Table 2

Primers and probes

Col I

ATNG, angiotensinogen; MR1, mineralcorticoid Type I receptor.

Real-time RT–PCR

The amount of cDNA of interest was analysed using ABI 7700 Prism Sequence detection system (Applied Biosystems), which was based on RT–PCR. The reaction solution was assembled in a volume of 25 µL, which comprised Platinum Quantitative PCR Super Mix-UDG (Invitrogen), forward and reverse primers (final concentration 300 nM each), TaqMan probe (final concentration 200 nM), and cDNA mixture (about 10 ng). The conditions for RT–PCR were pre-heating at 50°C for 2 min and at 95°C for 10 min, followed by 40 cycles of shuttle heating at 95°C for 15 s and at 60°C for 1 min. We used normal heart samples commercially available as normal controls. From human heart total RNA (Invitrogen), cDNA mixture was synthesized and used as a normal standard (normal-1) to generate the working standard for quantification of the cDNA of interest, which plots the relationship between the dilution of the standard cDNA mixture and the corresponding Ct value (the number of cycles necessary to obtain a threshold fluorescent signal). The initial quantity of the cDNA of interest, in a certain cDNA mixture, was calculated from the working standard and then normalized to that of GAPDH determined with Pre-developed TaqMan Assay Reagent Endogenous ControlTM (Applied Biosystems). The normalized value for each target cDNA is considered to reflect the mRNA expression levels of the corresponding gene in a test sample relative to normal-1. The quantitative assay for each cDNA was performed at least twice in triplicate and statistically analysed. In order to test the validity of normal-1 cDNA as normal control, cDNA mixtures were also synthesized from the other commercially available total RNAs from normal hearts (Clontech and Origen technologies). Changes in the mRNA expression levels by β-blocker treatment were expressed by the ratio of post- to pre-values.


From the right ventricular muscle biopsied in six patients, whole homogenate protein as well as total RNA were extracted. The myocardial samples were homogenized in 1.0 mL ISOGENTM reagent (Nippon Gene), thoroughly mixed with 0.2 mL chloroform, and centrifuged at 15 000 g for 15 min at 4°C. The interphase and organic phase were mixed with 0.3 mL ethanol and centrifuged at 2000 g for 5 min at 4°C. The supernatant was transferred into a micro test tube, mixed with 1.5 mL isopropanol, and centrifuged at 12 000 g for 10 min at 4°C. The precipitate was mixed with 2 mL 0.3 M guanidine hydrochloride in 95% ethanol, stored for 20 min at room temperature, and centrifuged at 7500 g for 5 min at 4°C. After the process was repeated three times, the precipitate with 2 mL ethanol was centrifuged at 7500 g for 5 min at 4°C. The precipitated total protein was air-dried and dissolved in sodium dodecyl sulfate (SDS) solution. The integrities of extracted proteins were examined by SDS–polyacrylamide gel electrophoresis (PAGE) with silver staining. Extracted proteins were loaded onto 15% SDS–polyacrylamide gel and separated. Proteins were electrotransferred onto nitrocellulose membrane (Invitrogen) and blocked with 1% Block Ace (Serotec, Oxford, UK) in phosphate-buffered saline Tween-20 (PBS-T) at 4°C. After washed with PBS-T, the membrane was exposed for 1 h at room temperature to monoclonal antibody against TGF-β1 (MAB1032, CHEMICON, Temecula, CA, USA) in PBS-T at a dilution of 1:150. The membrane was washed with PBS-T and incubated along with a horseradish peroxidase (HRP)-conjugated donkey anti-mouse secondary antibody (715-035-151, Jackson Immuno Research, Baltimore, PA, USA) for 1 h in PBS-T at a dilution of 1:50 000. After the membrane was washed with PBS-T, the immunological detection of TGF-β1 protein was performed with ECL Plus (Amersham, Chicago, IL, USA) and the band density was captured and degitalized by EDAS290 (Eastman Kodak, Rochester, NY, USA). The expression levels of TGF-β1 protein were normalized to that of actin. The amount of TGF-β1 protein was determined relative to the mean amount of the samples.

Statistical analysis

The sample size was obtained as follows: the primary variable was the mRNA level of Col I or III and the sample size was based on a two-sided paired t-test with a significance level of 0.05, a power level of 0.80, and with an anticipated effect size d=difference of means/standard deviation=0.75. The required sample size was 16. We decided 17 as sample size for avoiding the lack of required sample size during the follow-up.

Changes in gene expressions by the β-blockers were compared by the two-sided paired t-test. The relationship between gene expression and clinical indexes was determined by linear regression analysis. A value of P<0.05 was considered significant. Data are expressed as mean±SD.


Clinical indices of chronic heart failure

End-diastolic volume index (EDVI) (153±32 vs. 113±23 mL/m2, P=0.0007), left ventricular ejection fraction (LVEF) (21±7 vs. 35±9%, P<0.0001), plasma BNP concentration (165±146 vs. 45±79 pg/mL, P=0.0009), and WR (53±14 vs. 42±13%, P=0.002) significantly improved 4 months after the administration of β-blockers.

Myocardial gene expression

The expressions of genes that belong to renin–angiotensin system before the β-blocker treatment in DCM patients were close to that in normal controls except for angiotensin II Type 1 receptor (Table 3). The expression level of angiotensin II Type 1 receptor was down-regulated (0.30±0.17) compared with that of normal control.

View this table:
Table 3

Changes in the expression of factors involved in renin–angiotensin system


Expression levels of each gene are expressed as the relative value to those in the normal control. AGTN, angiotensinogen; MR1, mineralcorticoid Type I receptor; DCM, dilated cardiomyopathy; Before, before 4 months treatment with β-blocker; After, after 4 months treatment with β-blocker.

The mRNA expression levels of both Col I (1.08±0.72 vs. 0.65±0.26, P=0.024) and Col III (2.06±1.81 vs. 1.05±0.74, P=0.018) were significantly reduced by the β-blocker (Figure 1). The expression levels of either Col I or Col III did not correlate with EDVI, LVEF, or plasma BNP concentrations. Nevertheless, both Col I (r=0.560, P=0.041) and Col III (r=0.630, P=0.008) expressions correlated with WR before the β-blocker treatment, (Figure 2A and B). The expression levels of Col I (r=0.662, P=0.032) correlated with that of TGF-β1 before the β-blocker treatment. After the β-blocker treatment, either Col I or Col III expressions did not correlate with WR. However, the extent of changes in Col I (r=0.542, P=0.036) mRNA expression was positively correlated with that in WR. The extent of changes in Col III (r=0.455, P=0.089) mRNA expression has tendency of correlation with that in WR.

Figure 1 Changes in collagen mRNA expression levels 4 months after the treatment with β-blockers. The abundance of each mRNA expression in DCM patients is expressed by the relative value to that in the normal control. Before, before the β-blocker therapy, 4Mo after, 4 months after the treatment with β-blocker, nau, normalized arbitrary unit.

Figure 2 Relation between collagen mRNA expression and WR before the β-blocker therapy. The expression level of collagen in DCM patients is expressed by the relative value to that in the normal control. Pre, before the β-blocker therapy, nau, normalized arbitrary unit.

The extent of changes in both Col I (r=0.813, P<0.0001) and Col III (r=0.619, P<0.0001) mRNA expressions were positively correlated with that in TGF-β1 (Figure 3A and B). Moreover, the extent of changes in TGF-β1 (r=0.606, P=0.002) mRNA expression correlated with that in WR (Figure 4), although it did not correlate with that of the other components of renin–angiotensin system.

Figure 3 Relation between changes in collagen mRNA expression and TGF-β1 mRNA expression by β-blocker therapy. Post/pre, the ratio of the value after to before the treatment with β-blocker.

Figure 4 Relation between changes in the expression level of TGF-β1 and the value of WR. Post/pre, the ratio of the value after to before the treatment with β-blocker.

The expression levels of HGF correlated with that of mineral corticoid receptor 1 both before and after 4 months treatment with the β-blocker (Figure 5A and B). The extent of changes in HGF mRNA expression was also correlated with that in mineralcorticoid Type 1 receptor mRNA expression (Figure 5C).

Figure 5 Relation between the expression levels of HGF and MR1. The abundance of each mRNA expression in DCM patients is expressed by the relative value to that in the normal control. Post/pre, the ratio of the value after to before the treatment with β-blocker, MR1, mineralcorticoid Type 1 receptor, nau, normalized arbitrary unit.

Protein expression

The integrities of extracted proteins were confirmed by SDS–PAGE with silver staining (Figure 6A). The expression level of TGF-β1 protein was analysed by immunoblotting. We observed bands corresponding to TGF-β1 protein at 12.5 kDa in each lane (Figure 6B). The β-blocker reduced the protein expression levels of TGF-β1 (1.15±0.64 vs. 0.85±0.62, P=0.017) as well as the mRNA expression levels (Figure 6C).

Figure 6 (A) Representative SDS–PAGE with silver staining in extracted whole homogenate protein. (B) Representative immunoblot for TGF-β1 in failing hearts before (b) and after (a) the β-blocker therapy. (C) Summarized data from immunoblot analysis of TGF-β1. Before, before the β-blocker therapy, 4Mo after, 4 months after the treatment with β-blocker, nau, normalized arbitrary unit.


We found that, in the failing hearts of mild to moderate degrees, the gene expression level of collagen correlated not only with that of TGF-β1, but also with the sympathetic nerve activity. The collagen mRNA expression was decreased 4 months after the onset of the treatment with the β-blocker. Changes in collagen mRNA expression were positively correlated with those in TGF-β1, closely related to changes in cardiac sympathetic nerve activity by the β-blocker treatment.

Expression of the genes related to renin–angiotensin system in DCM patients

Although ACE mRNA in this study did not show significant increase, ACE mRNA is reported to increase in patients with aortic stenosis and aortic regurgitation using right ventricular biopsy samples,19 or in patients with advanced heart failure using excised heart.20 This difference may be attributed to the difference of the severity of heart failure, because New York Heart Association (NYHA) functional classes in this study were mild to moderate. The down-regulation of cardiac angiotensin II Type 1 receptor mRNA was observed in this study, consistently with the previous report,21 which may reflect an increase in tissue angiotensin II content. As ACE and angiotensinogen mRNA expression did not increase significantly, the down-regulation of cardiac angiotensin II Type 1 receptor mRNA may suggest that ACE activity is enhanced even in patients with mild to moderate severity of heart failure.

Regulation of collagen synthesis

Several studies targeted at DCM have demonstrated changes in the collagen content both at the protein7,2224 or mRNA level,7 characterized primarily by an accumulation of Col I. Indeed, the mRNA levels of both Col I and Col III mRNAs increased in some patients in this study. Different extent of increases in Col I and Col III contributes an increased Col I/Col III ratio in the myocardium of DCM patients.7 However, the relative increase in the mRNA level of Col III to I in this study seems to be inconsistent with these previous findings. We did not clarify this disparity. One explanation is that the ratio may be different between the stages or extent of progression of pathophysiological states of DCM.

Norepinephrine induces the accumulation of collagen in the myocardium independent of haemodynamic changes produced by norepinephrine,25 and was attributed to a direct effect of norepinephrine on cardiac adrenoreceptor.12 We could show that cardiac sympathetic nerve activity represented by WR of MIBG correlated with the Col I and Col III mRNA expression. This result, in combination with the results of the previous studies, provides one aspect of the theoretical background of the β-blocker treatment for chronic heart failure.

The increase in collagen synthesis by norepinephrine is reported to be accompanied by the increase in TGF-β1 mRNA abundance in rats heart.13 In addition to catecholamines, activation of renin–angiotensin system is also believed to play an important role in the collagen metabolism. A primary mediator of angiotensin II effects on collagen metabolism is also thought to be TGF-β1, because it has been shown to stimulate collagen synthesis in vitro,26 and activates a wide array of processes that collectively increase extracellular matrix production.27 Proteins and mRNA of TGF-β1 are increased in patients with aortic stenosis and regurgitation.9 In DCM patients, TGF-β1 mRNA expression was not increased significantly.7 TGF-β1 mRNA in this study was decreased compared with normal control. When myocardial fibrosis progresses in the early stage of DCM, the expression of TGF-β1 is up-regulated. However, at the end-stage of DCM, the expression of TGF-β1 may be down-regulated even when the TGF-β1 protein is accumulated in the myocardium as was observed in the present study. Functional classes of patients were stabilized mild to moderate by conventional therapy, but most of the patients had relative severe DCM and long clinical course before the catheterization. Therefore, basal levels of the expression of TGF-β1 may be already down-regulated. Importantly, the treatment with the β-blocker further reduced the expression level of TGF-β1 in the present study, suggesting that the small amount of progression of fibrosis may contribute to the progression of disease states.

The effects of β-blockers on the expression of collagen mRNA expression

The β-blockers are reported to reduce collagen production in rat heart with left ventricular dysfunction.14,15 This study, to our best of our knowledge, is the first to have investigated the effect of β-blockers on the collagen mRNA expression in DCM patients. This effect may contribute to the effect of β-blockers of left ventricular reverse remodelling,28 in which the function of cardiac fibroblast exerts an important role.6

Results of this study may provide some useful information to elucidate the mechanism of the decreases in collagen mRNA abundance by the β-blocker treatment. The extent of changes in Col I mRNA abundance, tended to correlate with the changes in WR and correlated with TGF-β1 mRNA abundance by the β-blocker treatment, which again implies that both cardiac sympathetic nerve activity and TGF-β1 was largely related to the collagen synthesis. Because the changes in TGF-β1 was associated with that in WR, the decrease in collagen synthesis by β-blocker may be mediated by TGF-β1, which is consistent with previous report.13

HGF is reported to attenuate collagen synthesis or promote collagen degradation.10,29 However, HGF did not seem to play the major role for the reduction of collagen by the β-blocker treatment, because HGF mRNA expression was irrelevant to collagen mRNA expression. HGF mRNA expression, in contrast, closely correlated with mineralcorticoid receptor Type 1 mRNA expression, which suggests a close relationship between tissue aldosterone concentration and HGF production. The combination treatment with β-blocker and spironolactone is expected to be more effective than the monotherapy in respect of the inhibition of collagen synthesis.

The dose of β-blocker and clinical improvement

We observed the marked improvements of cardiac function despite a low dose of β-blocker. Indeed, the target dose of β-blocker in the present study was less than that used in other clinical trials.3033 However, a low dose of β-blocker is also reported to improve left ventricular function and increase survival rates in patients with congestive heart failure.34,35 On the other hand, we cannot deny the possibility that some patients could improve cardiac function spontaneously or uncommonly during the study period. Another possibility for the effectiveness of a low-dose β-blocker is that all patients in this study were DCMs without coronary artery disease, and there is a tendency for more improvement in left ventricular function in the non-ischaemic patient than in the ischaemic patient by carvedilol.34 Since we did not set the control group for ethical reasons, we cannot perform the quantitative analysis of the improvements of cardiac function, although the cardiac improvements due to a β-blocker in the present study seem to be remarkable compared with the control data that have been already published.3035

Study limitations

Out of 34 referred patients, 17 were not included for various reasons. The exclusion of patients based on our criteria, after the start of the study, could introduce selection bias into the results.

Because of the lack of control groups without the use of β-blockers, changes in mRNA expression cannot be solely attributable to the effects of β-blockers. Moreover, because of the small sample size, the results of this study must be interpreted with caution. However, the decrease in collagen mRNA expression after β-blocker treatment in this study is consistent with the results examined in rat hearts with chronic pressure overload.15 By extrapolating from these results, we may attribute the changes in other gene expressions in this study to the effects of β-blockers.

GAPDH is stably expressed in human myocardium from patients with various cardiac conditions, thereby validating their use as a reference gene for internal normalization in the study using small biopsy samples. Although previous reports have documented variations in GAPDH gene expression in circumstances such as hypoxia,36 other studies state that GAPDH is unaltered in heart failure and GAPDH is usually used as reference gene in real-time RT–PCR for not only non-failing heart but also failing heart.37,38 Further study is required to validate using GAPDH as a reference gene because it is not determined definitely whether or not there is a change in the absolute abundance of GAPDH after β-blocker treatment.

We added β-blockers on conventional therapy including ACE-inhibitors, which had been administered at least 1 month before the beginning of β-blocker treatment. ACE-inhibitors or the combination of β-blockers with ACE-inhibitors might be partially responsible for some of these changes in gene expression, because the observation time in the present study may not be enough to confirm that the remodelling effect is only due to the β-blocker and not to the ACE-inhibitors.


The collagen mRNA expression, which is related to the cardiac sympathetic nerve activity, is inhibited by the treatment with β-blockers. This inhibition seems to be mediated by cardiac TGF-β1. The clinical benefit of β-blockers can be partly explained by the effect of β-blockers on this collagen metabolism.


This work was supported in part by the Promotion of Fundamental Studies in Health Science of the Organization for Pharmaceutical Safety and Research (OPSR) of Japan.

Conflict of interest: none declared.


View Abstract