OUP user menu

Steroid treatment causes deterioration of myocardial function in the δ-sarcoglycan-deficient mouse model for dilated cardiomyopathy

R. Bauer, G.A. MacGowan, A. Blain, K. Bushby, V. Straub
DOI: http://dx.doi.org/10.1093/cvr/cvn131 652-661 First published online: 20 May 2008


Aims As oral corticosteroids have a beneficial effect on muscle strength in Duchenne muscular dystrophy, it has been suggested that they may also be a useful treatment in the pathologically related sarcoglycanopathies. The δ-sarcoglycan-deficient mouse (Sgcd-null) is a model for both limb girdle muscular dystrophy 2F (LGMD2F) and dilated cardiomyopathy.

Methods and results To study the effect of oral corticosteroids on cardiac function, we treated 8-week-old Sgcd-null mice with prednisolone (1.5 mg/kg body weight/day orally) for 8 weeks. In vivo cardiac function was assessed by pressure–volume loops using a conductance catheter. We found a well-compensated cardiomyopathy at baseline in Sgcd-null mice with decreased myocardial contractility, increased preload, and decreased afterload, maintaining a high cardiac output. Cardiac haemodynamics, surprisingly, did not improve in prednisolone-treated mice, but instead deteriorated with evidence of ventricular stiffening. On histology, after steroid treatment there was increased myocardial cell damage and increased myocardial fibrosis.

Conclusion Prednisolone led to a decompensation of cardiac haemodynamics in Sgcd-null mice and induced additional cardiac damage. On the basis of these findings, although mouse models may not completely replicate the human situation for LGMD2F, we conclude that careful cardiac monitoring is clearly indicated in patients on long-term corticosteroids.

  • Dilated cardiomyopathy
  • Muscular dystrophy
  • Corticosteroids
  • Delta-sarcogylcan

1. Introduction

Many cases of dilated cardiomyopathy have a genetic etiology and they are frequently associated with forms of muscular dystrophy or myopathy.1 Dissociation of the dystrophin–glycoprotein complex (DGC), an oligomeric complex spanning the plasma membrane of skeletal and cardiac muscle fibres, is one cause of genetic but also acquired cardiomyopathies.2 Within the DGC the sarcoglycans form a subcomplex of proteins that link the cytoskeleton with the extracellular environment. Mutations in either α-, β-, γ-, or δ-sarcoglycan are responsible for the autosomal recessive limb girdle muscular dystrophies (LGMD) 2D, 2E, 2C, or 2F, respectively.3 Particularly, patients with a mutation in the δ-sarcoglycan gene (LGMD2F) often develop a progressive and potentially fatal cardiomyopathy.4 Moreover, mutations in the human δ-sarcoglycan gene have been characterized in patients with familial and sporadic cases of cardiomyopathy without significant involvement of the skeletal muscle.5

It has been suggested that loss of the sarcoglycans results in increased susceptibility of the sarcolemma to contraction-induced injury leading to myocardial damage and an inflammatory response, which further exacerbates the disease process.6,7 An inflammatory component in DGC-associated muscular dystrophies has been best documented in Duchenne muscular dystrophy (DMD), which was the rationale for corticosteroid therapy in this most common form of muscular dystrophy.8 It is well established that inflammatory cells and cytokines, like TNF alpha, contribute to myofibre damage and that reduced TNF alpha activity inhibits breakdown of skeletal muscle in mdx mice, the animal model for DMD.9 It has also been shown that corticosteroids can directly reduce TNF alpha activity in several tissues, including the heart.10,11 Corticosteroid treatment is now the gold standard therapy for DMD and has shown to increase skeletal muscle strength in DMD patients as well as in α-sarcoglycan-deficient LGMD patients.12,13 A retrospective study also suggested that an anti-inflammatory therapy with oral deflazacort has beneficial effects on left ventricular function in DMD patients.14

So far, the effect of corticosteroids on cardiac function in patients with LGMD2C–F is unknown. In the present study, we wanted to investigate the effect of oral prednisolone on cardiac function in sarcoglycan deficiency. Both the δ-sarcoglycan-deficient cardiomyopathic hamster (BIO14.6 hamster)15 and the δ-sarcoglycan-null mouse (Sgcd-null)16 are well established animal models for sarcoglycan-deficient cardiomyopathy. Cardiac pathology in the Sgcd-null mouse can be detected from 12 weeks of age.16 At later stages the mice show many hallmarks of a severe cardiomyopathy including cell death, cardiomyocyte regeneration, inflammation, and fibrosis.16 In our study, we wanted to investigate whether oral corticosteroid treatment can prevent or delay the onset of cardiomyopathy in the Sgcd-null mice. Treatment of mice was started at 8 weeks of age and simultaneous measurements of left ventricular pressures and volumes by conductance catheter 8 weeks after initiation of steroid treatment were performed.17,18

Our data in 36 animals illustrate that 16 weeks old Sgcd-null mice have a well compensated cardiomyopathy, which decompensates after prednisolone treatment in association with sarcolemmal injury and myocardial fibrosis.

2. Methods

2.1 Animals

Sgcd-null mice were originally generated in the laboratory from K. Campbell (Howard Hughes Institute, Iowa University, IA, USA) and have been previously described.16 Age-matched C57Bl/10 wild-type mice (Jackson laboratories) were used as controls. Mice were housed under controlled temperature (17–28°C) and light conditions (12:12 h light:dark cycle). Animals had free access to food and water.

All experiments were performed at the animal care facility of the University of Newcastle upon Tyne, UK. Experiments described in this report have been performed under the terms of the Animals Scientific Procedures Act 1986, and were authorized by the Home Secretary, Home Office, UK. The work has been approved by the Ethical Review Committee of the University of Newcastle upon Tyne.

2.2 Administration of prednisolone

Soluble prednisolone sodium phosphate (Sovereign) at a dose of 1.5 mg/kg body weight/day was given to 8 weeks old male Sgcd-null mice (n = 17) through drinking water over 8 weeks. Allometric scaling in mdx mice for prednisone showed that 1.5 mg/kg/day was the best dose to obtain a good effect on cardiac function (atrial contractility) without adverse effects on growth. This work was done by Andrew Hoey's group (University of Southern Queensland, Toowoomba, Australia). The mice were weighed weekly and water was changed every 48 h. The concentration of prednisolone was adjusted to changes in body weight and the amount of consumed fluid. Also, a group of age-matched male C57Bl/10 wild-type mice (n = 12) was treated to evaluate the effects of prednisolone in healthy animals. There was no difference in water consumption between the groups. At the age of 16 weeks, after undergoing in vivo pressure–volume analysis, animals were sacrificed, heart weight was determined, and the heart snap frozen in ice cooled isopentane.

2.3 Quantitative reverse transcription polymerase chain reaction

Total RNA was extracted from hearts of treated and untreated Sgcd-null and C57Bl/10 mice using the RNeasy mini kit (Qiagen) according to the manufacturer's specifications. One mircogram of total RNA was randomly reverse transcribed to cDNA using SuperScript III Reverse Transcriptase (Invitrogen).

Quantitative PCR was carried out using a LightCycler System (Roche), and the detection was performed by measuring the binding of the fluorescence dye SYBR Green I to double-stranded DNA at 530 nm. The PCR reactions were set up in microcapillary tubes in a volume of 20 µL. The reaction components were 2 µL undiluted cDNA, 1× FastStart DNA Master SYBR Green I (Roche) and final concentrations of 5 mM for MgCl2, and 1 µM for primers. TNF alpha primers were forward 5′-TCGTAGCAAACCACCAAGTG-3′ and reverse 5′-AGATAGCAAATCGGCTGAG-3′ (207 bp) and GAPDH were forward 5′-AACTTTGGCATTGTGGAAGG-3′ and reverse 5′-ACACATTGGGGGTAGGAACA-3′ (223 bp). Reactions were incubated at 95°C for 10 min to activate the polymerase followed by 45 cycles at 95°C for 15 s, 30 s at 54°C, 72°C for 20 s, and 72°C for 10 s. Crossing points were determined by the LightCycler software (version 3.5) using the second derivative method. On completion of the PCR amplification, a melting curve analysis was performed. Relative quantification was performed by calculating relative expression ratios using GAPDH as a reference gene and the relative expression software tool (REST© Version 205).19

2.4 Conductance catheter studies

Measurements were made in closed-chest, spontaneously breathing mice.18 Prednisolone treated and untreated 16 weeks old Sgcd-null mice (n = 36) and age-matched healthy C57BL/10 mice (n = 31) were anaesthetized by intraperitoneal injection of a water solution with Hypnorm® (Fentanyl + Fluanisone) and midazolam (average dose per g BW: 3.8 µg fentanyl, 120 µg fluanisone, and 60 µg midazolam). Oxygen was orally supplied and body temperature was maintained at 37°C using a homoeothermic blanket (Harvard apparatus).

A 1.4-French conductance catheter (Millar) was introduced into the right carotid artery and advanced retrogradely across the aortic valve into the left ventricle. The catheter was advanced under continuous haemodynamic monitoring to ensure proper placement in the left ventricle. After stabilization, steady-state measurements were recorded.

Through a small laparotomy at the level of the xiphisternum, transient compression on the inferior vena cava was applied using a flexible instrument cap to reduce preload used in determining end-systolic elastance (Ees) and other load-independent indices of contractility. Measurements were taken under steady state and inferior vena cava occlusion, at baseline and during infusion of low and high dose dobutamine (5 and 10 µg/kg/min). Volume was calculated using the Relative Volume Units/Cuvette method in which external blood-filled standards of known volume are used to calculate a slope and intercept to convert the conductance catheter signal of relative volume units to volume.20 This blood conductance signal needs to be corrected for the parallel conductance attributed to the tissues surrounding the left ventricular cavity. Parallel conductance was estimated by the previously described hypersaline method in which a bolus of 10 µL of 10% hypersaline was injected through the left internal jugular vein.17,21

Pressure and volume data were recorded using the Powerlab® Chart5 software (ADinstruments) and analysed using the PVAN pressure–volume data analysis software package (Millar).

Data analysis: Indices of systolic function included peak systolic pressure (Pmax), stroke volume (SV), cardiac output (CO), ejection fraction (EF), and stroke work (SW). Myocardial contractility was assessed by maximal rate of pressure development (dP/dtmax), linear regression of dP/dtmax vs. end-diastolic volume (dP/dtmax-EDV slope),22 Ees,23 and preload-recruitable SW (linear regression of SW vs. end-diastolic volume, PRSW slope).24 In some circumstances, those indices that combine pressure and volume measurements may simply reflect systolic stiffness of the left ventricle as opposed to contractility.25 Diastolic function was assessed by minimum pressure, tau (Glantz method, regression of log of pressure vs. time),26 and the maximal rate of pressure decay (dP/dtmin). Preload was determined by maximal left ventricular volume and afterload by arterial elastance (Ea). Ventriculo-arterial coupling was determined by the ratio of Ea/Ees.

2.5 Histopathological analysis and evaluation of myocardial damage by Evans blue dye

Frozen, 6 µm transverse sections were cut from the mid portion of the heart at the level of the papillary muscles and stained with haematoxyllin and eosin (H&E). To quantify myocardial fibrosis, sections were scored blinded to drug treatment status using a modified already published scale as follows:27 0, no fibrosis; 1, fewer than three areas of fibrosis; 2, greater than three areas of fibrosis; 3, greater than three areas of fibrosis and at least one area >100 µm in diameter. Sections were each assigned a score, and the average score from each animal was determined.

Evans blue dye (EBD) (Sigma) was injected intraperitoneally in treated and untreated mice without anaesthesia 16–24 h prior to surgery. EBD was dissolved in PBS (0.15 M NaCl, 10 mM phosphate buffer, pH 7.4) and sterilized by passage through membrane filters. Twenty-five microliters of this solution per 10 g body weight with a concentration of 0.5 mg EBD/0.05 mL PBS were injected. EBD is considered to be non-toxic at the injected single dose and has even been used for human pregnancy studies.28 To investigate potential side effects of the injected dye volume on mouse hearts, we compared functional cardiac parameters and TNF alpha RNA expression levels between the injected and four non-injected mice. No significant differences were detected.

After in vivo cardiac measurements the mice were killed by cervical dislocation. Hearts were extracted and atrial tissue removed, washed in PBS, weighed, flash frozen in liquid nitrogen and stored at −80°C. After incubation in ice-cold acetone at −20°C for 10 min, the sections were washed 3 × 10 min with PBS, and mounted with Vectashield mounting medium (Vector).

All sections were examined and photographed under an Axioplan fluorescence microscope (Carl Zeiss). By fluorescence microscopy analysis, EBD staining shows a bright red emission. Observing fielded EBD uptake the same score as described for fibrosis analysis was used as follows: 0, no uptake; 1, fewer than three areas of uptake; 2, greater than three areas of uptake; 3, greater than three areas of uptake and at least one area >100 µm in diameter. For data analysis, sections were each assigned a score, and the average score from eight sections for each animal was determined.

2.6 Data and statistical analysis

All data are reported as means ± SD. Differences between the different animal groups were compared by an unpaired Student's t-test or using two-factor ANOVA with the Scheffé's test for individual subgroup comparisons. P < 0.05 was used as criteria for statistical significance.

3. Results

3.1 Oral prednisolone reduces body weight in Sgcd-null mice

Eight-week-old Sgcd-null mice were treated for 8 weeks with oral prednisolone. There was no significant difference in body and heart weight in Sgcd-null mice (n = 19) compared with their age-matched C57Bl/10 mice (n = 19) without steroid treatment (Table 1). Orally administered prednisolone over 8 weeks lead to a 17% reduction of body weight in Sgcd-null mice (n = 17) at the age of 16 weeks; however the ratio of heart and body weight remained unchanged. No significant changes in body weight were observed in C57Bl/10 controls (n = 12) after oral prednisolone.

View this table:
Table 1

Comparison of baseline haemodynamic variables for Sgcd-null and C57Bl/10 mice without and after oral prednisolone treatment (mean ± [SD])

C57 Bl/10 (n = 19)Sgcd-null (n = 19)C57 Bl/10 steroid (n = 12)Sgcd-null steroid (n = 17)
Body weight (BW), g30 [3]29 [2]28 [2]24 [2]**
Heart weight (HW), mg124 [14]118 [14]119 [11]106 [9]
HW/BW ratio, %0.42 [0.03]0.41 [0.03]0.43 [0.03]0.44 [0.02]
Parameters in steady state
Heart rate, bpm436 [81]452 [45]450 [84]422 [58]
Maximum volume, µL22.6 [5.3]29.5 [5.6]#20.9 [5.2]25.3 [7.6]
Minimum volume, µL7 [4.2]8.6 [4.1]6.4 [2.3]7.2 [4.7]
Stroke volume, µL15.7 [5]20.8 [3.3]#14.5 [4.4]18.1 [5.5]
Cardiac output, µl/min6789 [2312]9398 [1478]#6307 [1372]7571 [2197]*
Ejection fraction, %71 [17]72 [10]69 [10]74 [14]
Stroke work, mmHg × µL1257 [495]1511 [284]1179 [471]1279 [430]
Peak pressure, mmHg94.6 [10.4]86.3 [6]#99.9 [10]88.9 [7.2]
Minimum pressure, mmHg−2.4 [3.2]−1.7 [2.9]−0.7 [4.1]1.1 [3.1]*
End-diastolic pressure, mmHg1.0 [4.0]4.0 [4.0]1.6 [3.8]6.0 [4.0]
Ea, mmHg/µL6.0 [1.9]3.7 [0.8]#7.2 [2.3]5.0 [2.5]
dP/dtmax, mmHg/s10461 [2016]8327 [1471]#12108 [2862]8842 [1698]
dP/dtmin, mmHg/s−8717 [1913]−7747 [981]−8002 [1245]−7233 [1603]
Tau, ms5.0 [2.4]5.7 [1.2]6.6 [1.3]7.3 [1.4]*
Parameters obtained after temporary preload reduction
Ees, mmHg/µL10 [4.8]3.4 [1.2]##11.8 [3.7]7.8 [3.9]*
PRSW, mmHg91 [36]58 [16]##85 [13]57 [14]
dP/dtmax-EDV, mmHg/s/µL666 [271]269 [78]##1144 [565]+492 [261]*
  • Ea, arterial elastance; dP/dtmax, maximal rate of pressure development; dP/dtmin, maximal rate of pressure decline; tau, isovolumic time constant of relaxation; Ees, end-systolic elastance; PRSW, preload recruitable stroke work slope; dP/dtmax-EDV, slope of dP/dtmax to end-diastolic volume; Ea/Es, ventriculo-arterial coupling; #P < 0.05; ##P < 0.001 Sgcd-null untreated vs. C57BL/10 untreated; *P < 0.05; **P < 0.001 Sgcd-null before vs. after oral prednisolone; +P < 0.05 C57Bl/10 before vs. after oral prednisolone.

3.2 Oral prednisolone reduces myocardial TNF alpha RNA expression in Sgcd-null mice

Corticosteroids have been shown to reduce TNF alpha activity, a cytokine known to be elevated in muscular dystrophies29 and cardiomyopathy.30 It has also been postulated that TNF alpha activity can be used as a prognostic marker in heart failure.31 We investigated TNF alpha expression in treated and untreated control and Sgcd-null mice. In untreated Sgcd-null mice we found a 100-fold higher expression of myocardial TNF alpha RNA compared with C57Bl/10 mice (Figure 1). After oral prednisolone treatment myocardial TNF alpha RNA levels were significantly lower in Sgcd-null mice suggesting pharmacological effects of oral prednisolone in our experiments. No significant differences were found in C57Bl/10 mice with and without oral prednisolone.

Figure 1

Relative quantification of myocardial TNF alpha RNA performed by calculating relative expression ratios showed a 100-fold over-expression in Sgcd-null compared with C57l/10 mice and a significant down-regulation after orally administered prednisolone. No significant changes were found in C57Bl/10 mice after treatment (#P < 0.05 Sgcd-null untreated vs. C57BL/10 untreated; *P < 0.05 Sgcd-null before vs. after oral prednisolone).

3.3 Compensated cardiomyopathy in Sgcd-null mice

After the age of 3 months, Sgcd-null mice develop a histological severe cardiomyopathy.16 So far, in vivo cardiac function in this animal model for LGMD2F has not been described. Before the start of steroid treatment we therefore studied in vivo cardiac function in untreated control and Sgcd-null mice by conductance catheter investigations. Measurements of systolic function and contractility such as maximal left ventricular pressure, dP/dtmax, Ees, and preload recruitable SW were significantly reduced in the Sgcd-null mice compared with control animals (Table 1; Figures 2 and 3A and C). Due to a compensatory increase in preload (maximum left ventricular volume) (Figure 2B and D) and decreased afterload (Ea), Sgcd-null mice maintained a high SV and CO. No significant differences in measures of diastolic function were found between control and Sgcd-null mice (Table 1).

Figure 2

Representative pressure–volume loops at baseline in C57Bl10 wild-type and Sgcd-null mice without and after oral prednisolone treatment. Hearts in Sgcd-null mice show significantly higher diastolic volumes (B), and lower systolic pressures compared with C57Bl/10 mice (A). After prednisolone treatment ventricles are smaller and diastolic pressures significantly increase in Sgcd-null mice (D). Pressure–volume loops in C57Bl/10 mice remained unchanged after oral prednisolone (C).

Figure 3

Load-independent myocardial contractility in Sgcd-null compared with C57Bl/10 mice. Linear regression analysis was applied to limited data over physiological pressure volume ranges.25 The slopes of ESPVR (Ees) (A) and PRSW (C) are significantly higher in C57Bl/10 compared with untreated Sgcd-null mice. Ees increases after oral prednisolone treatment in Sgcd-null mice (B), whereas PRSW remains unchanged (D). This may reflect increased myocardial stiffness after oral prednisolone treatment (##P < 0.001 Sgcd-null untreated vs. C57BL/10 untreated; *P < 0.05 Sgcd-null before vs. after oral prednisolone; n.s., not significant).

3.4 Abnormal β-adrenergic responses in C57/Bl10 wild-type mice after steroid therapy

Effects of steroid therapy on in vivo cardiac function in wild-type mice are not well described. In order to see whether wild-type mice develop steroid-associated side effects that could affect cardiac function, we analysed pressure volume loops in 31 age-matched, healthy C57Bl/10 mice with and without steroids. At baseline, we found no significant differences in the steroid-treated C57/Bl10 group with an exception of an isolated increase in the dP/dtmax-EDV relationship (Table 1). There were, however, marked differences after intravenous dobutamine infusion relative to non-treated controls, with reductions in indices of global systolic function (SV, CO, SW), elevated afterload (Ea), and abnormal diastolic function (dP/dtmin and tau) (Table 2; Figure 4).

Figure 4

Impaired beta-adrenergic response in C57Bl/10 and Sgcd-null mice after oral prednisolone with reduction in indices for global systolic function [stroke volume (A) and stroke work (B)] and abnormal diastolic function [dP/dtmin (C) and tau (D)] (*P < 0.05; **P < 0.001 Sgcd-null before vs. after oral prednisolone; +P < 0.05 C57Bl/10 before vs. after oral prednisolone).

View this table:
Table 2

Comparison of haemodynamic variables after intravenously administered dobutamine (10 µg/kg/min) for Sgcd-null and C57Bl/10 mice without and after oral prednisolone treatment (mean+[SD])

C57 Bl/10 (n = 19)Sgcd-null (n = 19)C57 Bl/10 steroid (n = 12)Sgcd-null steroid (n = 17)
Heart rate, bpm609 [11]620 [67]565 [13]593 [48]
Maximum volume, µL25 [6.9]25.2 [6]21.1 [3.8]23.1 [7.3]
Minimum volume, µL3.7 [3]3.0 [2.7]5.7 [3.3]6.2 [5.7]
Stroke volume, µL21.3 [6.2]22.2 [5.5]15.4 [3.4]+16.9 [4.2]*
Cardiac output, µL/min13010 [4015]13564 [3075]8712 [2050]+10135 [3230]*
Ejection fraction, %87 [19]88 [10]74 [14]76 [16]
Stroke work, mmHg × µL1689 [621]1586 [490]1112 [274]+1065 [331]*
Peak pressure, mmHg93.1 [10.4]86.5 [8.4]92.7 [8.8]83.9 [7.7]
Minimum pressure, mmHg−2.4 [3.2]−3.4 [2.5]−0.5 [2.1]1.1 [3.3]**
End-diastolic pressure, mmHg−0.7 [3.0]0.2 [2.7]1.4 [2.3]3.0 [3.0]**
Ea, mmHg/µL3.6 [1.1]3.0 [0.8]5.4 [1.6]+4.2 [1.3]*
dP/dtmax, mmHg/s16823 [2493]13936 [2735]#16215 [2309]13349 [1746]
dP/dtmin, mmHg/s−8050 [2282]−7921 [1179]−6343 [1379]+−5973 [1117]*
Tau, ms3.7 [1.7]3.9 [0.7]5.9 [1.4]+6.0 [1.4]**
Parameters obtained after temporary preload reduction
Ees, mmHg/µL14.6 [4]9.9 [4.4]#15.2 [5.9]12.9 [5.3]
PRSW, mmHg110 [35]102 [23]102 [29]87 [16]
dP/dtmax-EDV, mmHg/s/µL858 [380]700 [352]934 [479]679 [283]
  • #P < 0.05 Sgcd-null untreated vs. C57BL/10 untreated; *P < 0.05; **P < 0.001 Sgcd-null before vs. after oral prednisolone; +P < 0.05 C57Bl/10 before vs. after oral prednisolone.

3.5 Impaired haemodynamics after oral prednisolone in Sgcd-null cardiomyopathy

In our study we wanted to investigate whether oral corticosteroid treatment can improve cardiac function in Sgcd-null mice. We treated 19 Sgcd-null mice for 8 weeks. Steroid treatment was started at 8 weeks of age, when the heart was not showing any histological abnormalities. Instead of an improved heart function after 8 weeks of steroid treatment we observed a significant reduction in CO in Sgcd-null mice but the values remained above the values in wild-type mice. This can at least in part be explained by a loss in the compensatory increase in preload and reduction in afterload that we found in the Sgcd-null mice without steroids (Table 1; Figure 2). It is also possible that steroid treatment corrects the prior ‘high output situation’ and reverses haemodynamics in Sgcd-null mice in direction to those of the wild-type mice.

Diastolic function was significantly impaired after steroid therapy with an increase in minimal pressure and tau. Amongst parameters usually thought to indicate contractile state, Ees, and dP/dtmax-EDV were significantly increased after treatment with steroids, though preload recruitable SW was not (Figure 3B and D). An explanation for this finding is that parameters such as Ees may simply reflect systolic stiffness of the left ventricle25 rather than increased contractility, which would be consistent with the impaired diastolic function (also stiffness) and reduced CO. Figure 4 illustrates that β-adrenergic responses are similarly abnormal in both C57/Bl10 controls and Sgcd-null mice indicating a common effect of steroids on β-adrenergic mediated changes in global function and relaxation. The inadequate stress response might be a result of suppressed endogenous corticosteroid synthesis and an indicator for effective oral steroid intake. Whereas steroids did not generate any changes in global heart function in wild-type mice, Sgcd-null steroid treated mice developed marked abnormalities in both systolic and diastolic parameters.

3.6 Cardiac damage and fibrosis in Sgcd-null mice after prednisolone treatment

In order to assess cardiac pathology in response to steroid treatment both control and Sgcd-null mice were intraperitoneally injected with EBD 16–24 h prior to surgery. EBD is a non-toxic, widely used tracer that has been demonstrated to label damaged cardiomyocytes in vivo.32 We first analysed the heart muscle of untreated Sgcd-null mice for EBD uptake. Fluorescent microscopic analysis revealed only a few EBD-positive areas in the myocardium of untreated 16 weeks old Sgcd-null mice and corresponding necrotic and fibrotic areas on H&E (Figure 5). This suggested that 16-week-old Sgcd-null mice show little tissue damage at this stage of cardiomyopathy. In contrast, 16-week-old prednisolone-treated Sgcd-null mice displayed multiple areas of EBD uptake, which corresponded to regions of acute myocardial necrosis and fibrosis on H&E stained tissue sections. EBD-positive areas and fibrosis were detected within the left and right ventricular wall in a patchy pattern without any anatomical preference. The score we used to quantify EBD uptake and fibrosis in cardiomyocytes was found to be significantly higher in prednisolone-treated Sgcd-null mice compared with untreated Sgcd-null mice (Figure 6A and B). Our data showing that Ees in treated and untreated Sgcd-null mice is increased with severe fibrosis (Figure 6C), strongly suggest that the haemodynamic derangements are related to the underlying fibrosis. We did not detect any EBD uptake or fibrosis in untreated and treated C57/Bl10 control mice.

Figure 5

Representative hearts of Sgcd-null and C57Bl/10 mice without and after prednisolone treatment (5-fold magnification). Compared with untreated Sgcd-null mice (A and B), after oral prednisolone (C and D) pronounced histopathology and increased uptake of Evans blue dye (EBD) were found. There was no histopathology and EBD uptake in untreated (E and F) and treated (G and H) C57Bl/10 mice.

Figure 6

Mean scores (±SD) for quantification of (A) Evans blue dye (EBD) uptake (1.9 ± 1 vs. 0.5 ± 1) and (B) fibrosis (H&E) (2.9 ± 0.2 vs. 1.7 ± 1.3) in hearts of Sgcd-null mice after oral prednisolone treatment (n = 17) were found to be significantly higher compared with untreated Sgcd-null mice (n = 15) (*P < 0.05, unpaired Student's t-test). (C) End-systolic elastance (Ees) is increased with severe fibrosis in treated and untreated Sgcd-null mice [mild = fibrosis score 0-1 (n = 9) and severe = fibrosis score 2-3 (n = 23)] (§P < 0.05, unpaired Student's t-test).

4. Discussion

It is well documented that an inflammatory reaction is part of the pathogenic process in muscular dystrophy and muscular dystrophy-associated cardiomyopathy.8 It has also been shown that steroid treatment improves skeletal muscle function in muscular dystrophy patients.12 In addition, it has been assumed that treatment with steroids may have beneficial effects on cardiac function in these diseases.14

The rationale for our study was to investigate if early prednisolone treatment can improve or even prevent cardiomyopathy in δ-sarcoglycan-deficient mice, together with the cardiomyopathic hamster BIO14.6 a well established animal model for LGMD2F and cardiomyopathy. Both the hamster and the Sgcd-null mouse do not express any δ-sarcoglycan protein. Concomitantly, all other sarcoglycans are lost from the sarcolemma, which gives the mouse a rather severe phenotype. In patients with LGMD2F, many of the mutations are missense mutations resulting in a residual expression of the sarcoglycans and, consequently a broader spectrum of phenotypes. Not all patients with LGMD2F do therefore develop a cardiomyopathy. We collected in vivo baseline data for untreated Sgcd-null mice through conductance catheter studies and were able to show significant impairment in systolic and diastolic parameters. Mice were then treated for 8 weeks with oral prednisolone but did not show any improvement of their cardiomyopathy. In contrast, prednisolone treatment induced a decompensation of global heart function which was associated with an increase in myocardial pathology. In addition, we found an impaired β-adrenergic response in both C57Bl/10 wild-type mice and Sgcd-null mice after prednisolone treatment.

The cardiomyopathy in Sgcd-null mice has been well characterized histologically,16 whereas in vivo cardiac function has so far not been examined. For the first time, a conductance catheter method was used to assess cardiac function in Sgcd-null mice at the age of 16 weeks. According to our findings at baseline the cardiomyopathy is well compensated by favourable loading conditions at this stage of development. Low myocardial contractility is disguised by increased preload (end-diastolic volume) and reduced afterload (Ea) thereby maintaining high SV and CO. Oral treatment with prednisolone over 8 weeks induced a decrease in global cardiac function with lower CO and in particular diastolic dysfunction (Figure 7). On the other hand, the decrease in CO could also be interpreted as a correction of a ‘high output situation’, as CO remained above the values of wild-type mice and ventriculo-arterial coupling improved. Nevertheless, the histological deterioration after steroid therapy in Sgcd-null mice compared with untreated Sgcd-null mice and wild-type mice strongly suggests negative effects even if some of the correlative haemodynamics are ambiguous.

Figure 7

Schematic pressure volume loops showing the effect of oral prednisolone on cardiac haemodynamics (grey dotted = C57Bl/10, black long dashed = Sgcd-null, black = Sgcd-null steroid). After prednisolone treatment loops are shifted towards higher diastolic and systolic pressures and lower diastolic filling volumes resulting in reduced stroke volume.

Interestingly, despite a higher Ees cardiac function declined after prednisolone treatment, suggesting ventricular stiffness in Sgcd-null cardiomyopathy. Ees, the slope of the end-systolic pressure–volume relationship is calculated during transient preload reduction and is often used as a parameter for load-independent myocardial contractility. On the other hand, it has been reported that structural changes with hypertrophy or fibrosis can also increase Ees.25 In our study, this could be confirmed by histological analysis of cardiac tissue sections. The functionally increased ventricular stiffness was reflected by increased myocardial damage and fibrosis after steroid treatment in Sgcd-null mice with our data clearly showing that Ees in Sgcd-null mice is increased with severe fibrosis. This shows the importance to assess Ees in context with other parameters, since an ‘isolated’ increase in Ees can be misinterpreted assuming better myocardial contractility. Finally, no histopathology was found in C57Bl/10 wild-type mice after prednisolone therapy in accordance to an unchanged Ees.

We were also interested to examine whether Sgcd-null mice were likely to develop a more pronounced cardiac abnormality after β-adrenergic stimulation by intravenous dobutamine. Whereas β-adrenergic response in untreated Sgcd-null mice was normal, we found an abnormal β-adrenergic response of global heart function in Sgcd-null mice after treatment with prednisolone. Additionally, prednisolone-treated C57Bl/10 wild-type mice showed similar alterations after β-adrenergic stimulation with systolic and diastolic dysfunction. These findings can either be interpreted as a direct toxic effect of steroids on healthy hearts or as a suppressed endogenous corticosteroid activity. Indirect results that might be the consequence of fluid retention were not apparent, as determined by body weights.

The pharmacodynamic effects of steroids in DGC-associated muscular dystrophies and cardiomyopathies are still poorly understood. It has been hypothesized that prednisolone modulates calcium handling in muscle cells and that alterations in mitochondrial function and other cellular signalling pathways (e.g. nitric oxide) are involved.33 Quantitative RT–PCR revealed increased TNF alpha expression levels in Sgcd-null mice, which were significantly down-regulated after prednisolone treatment. This suggests that a steroid induced modulation of inflammatory pathways may contribute to the observed deterioration of Sgcd-null mice cardiomyopathy.

It has been shown by experimental studies that physiological levels of TNF are necessary for cardiovascular homeostasis.34,35 Accordingly, sustained lowering of TNF alpha levels may have contributed to the myocardial damaging after prednisolone therapy. Our results suggest that TNF alpha is not a therapeutic target in muscular dystrophy-associated cardiomyopathy, which is compatible with the results of a large heart failure clinical trial showing that preventing TNF from binding to its receptors does not have any beneficial and in some cases even detrimental effects.36 Furthermore, detrimental direct myocardial effects of steroids must be considered, as it is known that there is a close relation between plasma cortisol levels and mortality in patients with congestive heart failure.37

There is no doubt that steroids are beneficial for boys with DMD38,39 and they possibly also improve left ventricular function in DMD patients.14 Interestingly, mdx mice do not seem to benefit much from steroid treatment if one looks at histological changes in skeletal muscle.40 This illustrates that treatment effects in humans can not be easily extrapolated to mice and vice versa. We do not postulate that corticosteroids are detrimental for patients with sarcoglycan-deficient muscular dystrophies or cardiomyopathies but rather want to point out that treatment effects in mouse models might be different from the effects one might observe in patients. The steroid dose we used in our study was slightly higher than normally used for DMD boys and steroids were not administered in a single oral dose, which might influence their pharmacodynamic effect. So far, no controlled trials in LGMD2D–F patients using steroids have been published and it is feasible that the effects of steroid therapy in those patients are different from the ones in Duchenne boys.

However, according to our findings a comprehensive set of parameters seems to be crucial for the interpretation of heart function and clinical treatment outcomes. In our study echocardiography would show smaller ventricles after prednisolone treatment suggesting beneficial therapeutic effects. Smaller end-diastolic volumes may on the other hand reflect ventricular stiffness and it is therefore not sufficient to only assess ventricular size. Even generally used myocardial contractility parameters such as Ees may not be specific enough to asses ‘intrinsic’ myocardial contractility. In addition, the commonly used EF, a parameter for global heart function, may not record therapy-induced changes sensitively. Despite an increase in histopathological changes in the steroid-treated Sgcd-null mice, the EF remained completely unchanged.

In conclusion, in our study oral prednisolone deteriorated cardiac haemodynamics in Sgcd-null mice, an animal model for LGMD2F and cardiomyopathy, and induced cardiac damage and fibrosis.


The study was funded by Heart Research UK, Special Trustees of the Newcastle University Hospitals and the German Ministry of Education and Research (BMBF, Bonn, Germany). R.B. and V.S. are members of the German Muscular Dystrophy Network (MD-NET 01GM0601) funded by the BMBF; www.md-net.org. MD-NET is a partner of TREAT- NMD (EC, 6th FP, proposal # 036825; www.treat-nmd.eu).


We thank Professor Andrew Hoey from the Centre for Biomedical Research at the University of Southern Queensland in Toowoomba, Australia, for his advice on steroid dosage in muscular dystrophy mouse models.

Conflict of interest: none declared.


View Abstract