High prevalence of plasma lipid abnormalities in human and canine Duchenne and Becker muscular dystrophies depicts a new type of primary genetic dyslipidemia

Abstract

BACKGROUND:

Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD) are allelic X-linked recessive muscle diseases caused by mutations in the DMD gene, with DMD being the more severe form. We have recently shown that increased plasma low-density lipoprotein–associated cholesterol causes severe muscle wasting in the mdx mouse, a mild DMD model, which suggested that plasma lipids may play a critical role in DMD. We have also observed that loss of dystrophin in mice causes unexpected elevations in plasma lipoprotein levels.

OBJECTIVE:

The objectives of the study were to determine whether patients with DMD and BMD also present with clinically relevant plasma lipoprotein abnormalities and to mitigate the presence of confounders (medications and lifestyle) by analyzing the plasma from patients with DMD/BMD and unmedicated dogs with DMD, the most relevant model of DMD.

METHODS:

Levels of low-density lipoprotein–associated cholesterol, high-density lipoprotein cholesterol, and triglycerides were analyzed in patients with DMD and BMD and female carriers. Samples from unmedicated, ambulatory dogs with DMD, unaffected carriers, and normal controls were also analyzed.

RESULTS:

We report that 97% and 64% of all pediatric patients with DMD (33 of 36) and BMD (6 of 11) are dyslipidemic, along with an unusually high incidence in adult patients with BMD. All dogs with DMD showed plasma lipid abnormalities that progressively worsened with age. Most strikingly, unaffected carrier dogs also showed plasma lipid abnormalities similar to affected dogs with DMD. Dyslipidemia is likely not secondary to liver damage as unaffected carriers showed no plasma aminotransferase elevation.

CONCLUSIONS:

The high incidence of plasma lipid abnormalities in dystrophin-deficient plasma may depict a new type of genetic dyslipidemia. Abnormal lipid levels in dystrophinopathic samples in the absence of muscle damage suggest a primary state of dyslipidemia. Whether dyslipidemia plays a causal role in patients with DMD warrants further investigation, which could lead to new diagnostic and therapeutic options.

Introduction

Duchenne muscular dystrophy (DMD) is the most common form of muscular dystrophy (MD), affecting approximately 9–15 per 100,000 live male newborns worldwide.¹ DMD is caused by the loss of the X-linked, 427-kDa protein dystrophin,² a key member of the dystrophin-associated glycoprotein complex (DGC), which spans the plasma membrane and links the extracellular matrix to the intracellular cytoskeleton.³ Absence of dystrophin results in increased sarcolemmal fragility and a greater susceptibility of muscle fibers to mechanical injury.³ Affected boys present with abnormal gait, progressive muscle weakness, calf hypertrophy, and elevated serum creatine kinase (CK).^1,4,5 Continuous muscle degeneration/necrosis, inflammation, and fibro-fatty deposition generally result in loss of ambulation by the age of 12 years^6,7 as well as severe cardiorespiratory dysfunction that is often lethal if untreated. On the other hand, Becker muscular dystrophy (BMD) is a less severe form of dystrophinopathy. Patients with BMD exhibit slower progression than those with DMD as muscles contain reduced quantities of a truncated yet partially functional dystrophin protein.⁴

Our group recently challenged the mdx mouse, a dystrophin-null but clinically mild DMD model by knocking out their apolipoprotein E (ApoE) gene, which increased their atherogenic plasma lipid profile.⁸ This resulted in severe exacerbation of the clinically mild dystrophic phenotype of mdx mice with prominent gait abnormalities and human-like muscle wasting only observed in mdx/ApoE double-deficient mice, particularly when fed a high-fat diet.⁸ Other types of MD, such as dysferlin-dependent LGMD2B, showed similar exacerbation after ApoE inactivation,⁹ suggesting that atherogenic lipids might exacerbate MD-associated muscle wasting. Interestingly, in addition to devastating skeletal, cardiac, and respiratory muscle wasting, a recent study has reported the presence of various lipid-related abnormalities in patients with DMD using advanced proton nuclear magnetic resonance (NMR) spectroscopic techniques.¹⁰ Indeed, purified serum extracts from pediatric patients with DMD were shown to contain elevated levels of phospholipids and cholesteryl esters, amongst others, when compared with healthy individuals.^10,11 While it might be assumed that these abnormalities are secondary to lifestyle changes (such as loss of ambulation), medication and the muscle wasting process observed in patients with DMD, to our knowledge, the cause of these changes have not been investigated.

To better understand the etiology of muscle wasting in dystrophinopathies and translate the potential link between elevated plasma lipids and exacerbated muscle wasting previously highlighted by our animal studies to patients, plasma samples from patients with DMD and BMD and unaffected female carriers with heterozygous DMD mutation were analyzed using routine clinical approaches with the objective of showing signs of abnormal lipoprotein metabolism in this unique patient population. This study was initiated based on the hypothesis that little to no lipoprotein-related abnormalities were to be found in dystrophinopathies. Owing to the presence of multiple confounders in these patients, plasma samples from unmedicated dystrophin-null dogs, a relevant preclinical model of DMD due to manifestation similarities with human DMD,¹² were also analyzed. Our data suggest that dystrophinopathies such as DMD and BMD may represent a new type of primary genetic dyslipidemia. These findings are potentially significant for the diagnosis and pharmacological management of this patient population and depict DMD as a systemic metabolic disease. Whether dyslipidemia plays a causal role in DMD muscle wasting warrants further investigation.

Methods

Participants

Plasma samples were retrospectively obtained from human patients from the MRC Biobank for Rare and Neuromuscular Diseases curated at Newcastle University. The National Research Ethics Service Committee Newcastle and North Tyneside 1 approved prior collection of samples from all patients and their subsequent use in research (REC Number: 19/NE/0028). Informed consent was obtained from all subjects. Ambulatory status, corticosteroid use, and measurements for age, weight, and height (where applicable) were retrospectively obtained from United Kingdom National Health Service (NHS) records. Samples were processed at the University of British Columbia affiliated, St Paul’s Hospital, (Vancouver, Canada) under Ethics number H19–01573.

Canine cohort

Canine plasma samples were obtained from dogs on a mixed genetic background of golden retriever, Labrador retriever, beagle, and Welsh corgi, generated in house by artificial insemination. The genotype (DMD carry null dystrophin mutations, and carriers are female heterozygous mutants) was determined as published.^13–15 All experimental dogs were housed in a specific pathogen-free animal care facility and kept under a 12-hour light/12-hour dark cycle. Dogs with DMD were housed in a raised platform kennel, whereas normal dogs were housed in regular floor kennels. Dogs were kept in groups where possible, to promote socialization. Normal and carrier dogs were fed dry Purina Laboratory Canine Diet 5006 (LabDiet, St Louis, MO), whereas dogs with DMD were fed wet Purina Pro Plan Puppy food (LabDiet, St Louis, MO) as instructed by the veterinarian. Dogs were monitored daily by the care-givers for overall health condition and activity and given ad libitum access to clean drinking water and toys for enrichment. None of the experimental dogs were euthanized at the end of the study. All experiments were approved by the Animal Care and Use Committee of the University of Missouri and were performed in accordance with National Institutes of Health guidelines. Samples were processed at the University of British Columbia affiliated, St Paul’s Hospital, (Vancouver, Canada) under Ethics number H19–01573, as well as University of Missouri, Veterinary Diagnostic Laboratory (Missouri, USA).

Serology and lipid profile measurements

All biobanked samples mentioned previously were collected according to standard clinical operating procedures under nonfasting conditions to minimize patient discomfort. Plasma lipoprotein levels are shown to change minimally in response to food intake, although triglycerides (TGs) show the most variation (maximum 20% or ± 0.3 mmol/L).^16,17 Briefly, venous blood from nonfasted patients was collected in EDTA vacutainers. Whole blood was centrifuged at 2850 g for 10 min, and the supernatant plasma removed and stored at −80°C pending use. Anonymized human plasma samples were processed using the Siemens ADVIA 1800 system for total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C) and TGs as previously reported.^8,9 Low-density lipoprotein cholesterol (LDL-C) was calculated using the formula: TC–HDL-C - (TG/2.2). Non–high-density lipoprotein cholesterol (non–HDL-C) was calculated using the formula: TC–HDL-C.

For dog samples, plasma was collected in heparinized tubes and spun down at 4000 RPM for 10 min at 4°C and was stored at −80°C. Levels of CK, aspartate aminotransferase (AST), alanine aminotransferase (ALT) as well as TC, HDL-C, and TG were measured using the Siemens ADVIA 1800 system. Calculated values of LDL and non–HDL-C were calculated as described previously. For time-course studies, TC (only) was performed at the UMC Vet Med Diagnostic Lab.

Lipid parameters

Patient lipid levels were plotted relative to the 2018 American College of Cardiology/American Association Task Force on Clinical Practice Guidelines.¹⁸ Nonfasting ranges for children and adolescents aged, <19 years were defined as follows; normal: TC (<4.4 mmol/L); LDL-C (<2.8 mmol/L); HDL-C (>1.2 mmol/L); and non–HDL-C (<3.1 mmol/L); borderline high/low: TC (4.4–5.1 mmol/L); LDL-C (2.8–3.3 mmol/L); HDL-C (1.2–1 mmol/L), and non–HDL-C (3.1–3.7 mmol/L); high/low: TC (≥5.2 mmol/L); LDL-C (≥3.4 mmol/L); HDL-C (<1 mmol/L); and non–HDL-C (>3.7 mmol/L). Normal ranges for TG levels in pediatric patients aged 0–9 years were set at <0.85 mmol/L, borderline high between 0.85 and >1.1 mmol/L, and high >1.1 mmol/L. Normal ranges for TG levels in pediatric patients aged 10–19 years were set at, 1 mmol/L, borderline high between 1 and 1.45 mmol/L, and high >1.45 mmol/L.¹⁸ Cutoffs for high, and borderline high parameters represent approximately the 75th and 95th percentiles, respectively, whereas the low HDL-C cutoff represents approximately the 10% percentile.¹⁸

Although there are no clear percentile cutoffs for lipid abnormalities in adult patients, optimal/healthy lipid values for men and women aged >20 years were defined as follows: TC (<5.2 mmol/L); LDL-C (<3.4 mmol/L); HDL-C (>1 mmol/L in men and >1.3 mmol/L in women); non–HDL-C (<4.1 mmol/L); and TG (<1.7 mmol/L).¹⁸

Histopathology and immunostaining

Immunohistopathology was performed on representative sections using standard hematoxylin–eosin staining as previously described.^19–21 Similarly, dystrophin was examined with a mouse monoclonal antibody raised against the repeat 17 of human dystrophin (Mannex44A) (1:100, a gift from Dr Glenn Morris, Wolfson Center for Inherited Neuromuscular Disease, RJAH Orthopedic Hospital, Oswestry, and Keele University, Keele, Staffordshire, UK).

Statistical methods

Statistical analyses are described in figure legends. Data were analyzed using GraphPad PRISM v6. Data are mean ± SEM. P < .05 was considered statistically significant.

Results

Patient characteristics

Cross-sectional plasma samples were obtained from pediatric patients with DMD (mean 8.4 ± 2.6 years, male) and BMD (mean 11.4 ± 1.53 years, male) aged <18 years (N = 36 and 11, respectively). Sixteen patients with DMD were also longitudinally assessed (up to 2 years follow-up) involving blood collection at varying time intervals. To obtain a cross-sectional value for these 16 patients, longitudinal samples were averaged and plotted accordingly. Cross-sectional plasma samples were also obtained from adult patients aged >18 years were as follows: N = 20 males with BMD (mean 35.1 ± 3.31 years), and N = 13 female carriers, with a heterozygous mutation of the DMD gene (mean 45.8 ± 4.04 years). Of 13 carrier patients, 5 were also longitudinally assessed (up to 2 years of follow-up) involving blood collection at varying time intervals. Dystrophin mutations for all patients (where applicable) are listed in Supplement Table 1.

Prevalence of plasma lipid abnormalities in pediatric populations

To investigate plasma lipid abnormalities in pediatric dystrophinopathies, we first analyzed the plasma lipid profiles in 36 patients with DMD (mean 8.4 ± 2.6 years, male) and 11 pediatric patients with BMD (mean 11.4 ± 1.53 years, male) aged <18 years. Of 36 patients with DMD, 33 were ambulatory at the time of sampling (with 2 requiring the use of walking aids) and 3 were wheelchair bound. In addition, 29 patients with DMD were medicated with corticosteroids (either prednisolone or deflazacort), 2 were unmedicated, and 5 patients were unmedicated on their first visit but received steroids at follow-up. All 11 patients with BMD were ambulatory (3 requiring the use of walking aids), 1 was taking corticosteroids, and 2 were unmedicated. Medication usage was not specified for the 8 remaining patients with BMD.

On average, DMD samples showed 27% higher TC levels (P < .0001) than BMD samples, with 30% and 27% of these changes attributed to comparative increases in non–HDL-C (P = .0011) and LDL-C–associated lipoprotein fractions (P < .05), respectively (Fig. 1A). Further distribution of patients into normal and above-normal subcategories highlighted that 64% of boys with DMD displayed irregularities in total plasma cholesterol levels compared with only 9% in those with BMD (Fig. 1B). Interestingly, despite the heightened state of LDL-C and non–HDL-C levels observed in patients with DMD (Fig. 1A), only 25% and 47% of the DMD samples were found to have above-normal levels of LDL-C and non–HDL-C, respectively (Fig. 1C and D). Average plasma HDL-C levels were similar between patients with BMD and DMD (Fig. 1A) although only 36% and 17% of them showed abnormally low HDL-C levels, respectively (Fig. 1E). Most strikingly, 100% of boys with BMD and 91% of boys with DMD aged 0–9 years, as well as 20% of boys with BMD and 85% of boys with DMD aged between 10 and 19 years exhibited above-normal TG levels for their age group (Fig. 1F). On an individual basis, our data show that 92% of all pediatric patients with DMD (33 of 36) and 82% of pediatric patients with BMD (9 of 11) show at least one above-normal plasma lipoprotein parameter. Frequencies of patients who display multiple lipid abnormalities in both patient populations are highlighted in Supplement Figure 1A and B.

Figure 1. Comparative analysis and normal/above normal distribution of plasma lipoprotein and triglyceride levels in pediatric Becker/BMD and Duchenne/DMD patients <18 years of age.

A: Plasma Total cholesterol (TC), low density lipoprotein cholesterol (LDL-C), non high density lipoprotein cholesterol nonHDL-C (which includes very low, and intermediate density lipoprotein cholesterol, VLDL-C and IDL-C), high density lipoprotein cholesterol (HDL-C) and triglycerides (TG) in Becker and Duchenne patients. B–F: scatter plots of Becker and Duchenne lipoprotein plasma lipoproteins and TG showing values that fall within and outside of normal pediatric levels. Panel A was analyzed using a two way ANOVA, and Bonferroni’s post hoc testing was used to analyze direct mean comparisons. (**) P<0.01; (****) P<0.0001. Non-shaded areas indicate lipid values falling within normal range; light grey denotes borderline high (75^th percentile) ranges; and dark grey denotes either high (95^th percentile), or low (10^th percentile) ranges for pediatric patients.

To confirm that age is not a factor regulating DMD-associated changes in lipoprotein levels, all data from pediatric patients with DMD were plotted longitudinally across the following age groups (<5, 6, 7, 8, 9, 10, 11, and 12+ years). Neither TC, LDL-C, non–HDL-C, HDL-C, nor TG levels were significantly affected by age in children with DMD (Supplement Fig. 2A and E). Moreover, of the 16 patients with DMD with up to 2 years of longitudinal measures and no significant elevations in plasma lipoprotein or TG content were observed when compared with baseline levels (Supplement Fig. 2F). Given that changes in body weight, as a result of steroid use, diet, or physical activity can also contribute to elevations in plasma lipoprotein and TG levels, age- and sex-standardized body mass index (BMIz) scores were also comparatively analyzed between pediatric patients with DMD and BMD. As expected, patients with DMD had significantly higher BMIz scores than patients with BMD (mean 1.62 ± 0.11 and 0.35 6 0.44, respectively; P = .0002, unpaired students t-test). BMIz was shown to be a significant (but weak) factor affecting TC levels in patients with DMD (P < .05), explaining up to 12% of the observed variation among samples (R² = 0.12) (Supplement Table 2). In patients with BMD, BMIz was shown to be a strong factor affecting non–HDL-C levels (P < .05) and to a lesser extent TC (P = .06), explaining up to 61% of the observed variation (R² ≥ 0.47) (Supplement Table 2). Interestingly, BMIz was not a significant predictor of plasma TG changes among patients with DMD or BMD (Supplement Table 2).

Given that steroid use is a potential confounding factor affecting lipoprotein and TG levels, plasma lipid profiles of 5 pediatric patients with DMD who were not receiving steroids at their baseline visit, but were administered, and received steroids for 1 year were assessed at follow-up (Supplement Fig. 3A–F). Patients were on average aged 5 years (±0.55 years) at their baseline measurement. In this small patient population, no differences in TC, LDL-C, non–HDL-C, or TG levels were observed between preand post-steroid values (Supplement Fig. 3A). While a varied distribution of TC and lipoprotein fractions (LDL-C, non–HDL-C, and HDL-C) was observed both before and after steroid administration, interestingly, elevations in plasma TGs occurred independent to medication.

Prevalence of plasma lipid abnormalities in adult BMD and female carrier populations

To investigate whether plasma lipid abnormalities were also present in adult dystrophinopathic patients (>18 years), plasma samples from N = 20 BMD patients (mean 35.1 ± 3.31 years, male) and N = 13 carrier samples (mean 45.8 ± 4.04 years, female) were analyzed and values plotted against normal lipid parameters for adults. Of these BMD patients, 18 were ambulatory (with 5 requiring the use of walking aids), and 2 were listed as non-ambulatory. Of the 13 carrier samples, 8 were ambulatory (with 1 requiring the use of walking aids), 1 was non-ambulatory, and ambulatory status was not available for 4 carrier patients. Medication status or BMI for these patients was not available retrospectively on NHS records.

In adult males with BMD and carrier females, 30% and 31% of patients exhibited suboptimal TC levels, respectively (Fig. 2A). Minimal abnormalities, however, were observed in LDL-C, HDL-C, and non–HDL-C lipoprotein fractions (Fig. 2B–D). Of patients with BMD, 50% exhibited suboptimal plasma TG levels, whereas heightened TG was only observed in 15% of carrier females (Fig. 2E). From a longitudinal perspective, 5 carrier patients were assessed for age-related plasma lipid and TG changes up to 2 years. No significant changes in TC, LDL-C, non–HDL-C, HDL-C, or TGs could be detected between baseline and follow-up samples in this patient population (Fig. 2F). Based on the aforementioned data, 60% of adult patients with BMD (12 of 20) and 46% of adult female carriers (6 of 13) showed at least one a abnormality, indicating the presence of dyslipidemia in less-severe adult BMD populations.

Figure 2. Optimal and suboptimal distribution of plasma lipoprotein and triglyceride levels in adult Becker/BMD and Carrier patients >20 years of age.

A–E: scatter plots of Becker and Carrier lipoprotein plasma lipoproteins and TG showing values that fall within and outside of normal adult levels. No statistics were conducted for these data sets. F: Percentage (%) change in lipoprotein and TG levels in carrier patients with longitudinal measurement (with up to 2 years follow up). Panel F: Paired samples t.tests were used to compare baseline and follow-up measurements between patients. Grey areas indicate suboptimal values.

Canine characteristics

Because the unexpected changes in lipid profiles described previously can be affected by medication use and lifestyle, we investigated plasma lipoprotein levels in unmedicated and ambulant canine patients with DMD, carriers, and normal controls. Affected dogs are widely considered the gold standard preclinical model of human DMD. Studying canine samples also minimizes the presence of potential confounders. Plasma samples from 33 dogs were obtained: 11 were normal; 8 were carrier female; and 14 were DMD-affected, mix of male and female. The age range of normal dogs was between 11.1 and 25.2 months (mean 19.1 ± 1.6 months); carrier dogs between 19.6 and 27.9 months (mean 20.2 ± 1.24 months); and dogs with DMD between 9.8 and 27.9 months (mean 17.0 ± 1.5 months). Groupings were shown to be statistically similar in age (P < .09; one-way ANOVA, Tukey’s post hoc test). For additional time-course studies, TC was determined for both normal and affected dogs in a separate cohort of dogs across the following age groups: 2 months (normal N = 14, mean age 2.2 months; affected N = 28; mean age 2 months); 7–8 months (normal N = 23, mean age 7.8 months; affected N = 46, mean age 7.2 months); 14–15 months (normal N = 7, mean age 13.8 months; affected N = 16, mean age 14.8 months); and 28–34 months (normal N = 2, mean age 33.7 months; affected N = 12, mean age 28.8 months).

Prevalence of plasma lipoprotein abnormalities in a canine model of DMD

To confirm DMD-associated muscle pathology after PCR-based genotyping (not shown), muscle histology and CK level measurements were performed. Dystrophin protein was robustly detected by immunohistochemistry (IHC) in skeletal muscles from normal and unaffected carrier dogs but undetected in DMD-affected dogs (Fig. 3, right, red channel). Moreover, the muscle of dogs with DMD displayed profound degeneration, inflammation, and fibrosis, clear features of muscle wasting. Except for the occasional presence of centrally nucleated myofibers (purple), carrier dog tissues were mostly similar to normal muscle (Fig. 3 left). Furthermore, DMD-affected animals showed CK elevation compared with normal samples, whereas unaffected carrier samples showed no CK elevation, (Fig. 4A), indicating absence of damage in the latter samples. When lipoproteins levels were analyzed, we observed that affected dogs with DMD showed significant 75%, 440%, 272%, 34%, and 65% increases in plasma TC (P < .0001), LDL-C (P = .0002), non–HDL-C (P < .0001), HDL-C (P = .0002), and TGs (P = .02) compared with normal dogs, respectively (Fig. 4B–E). Most strikingly, unaffected carrier dogs with normal CK also displayed a statistically significant 55% increase in plasma TC (P = .002) and a 46% increase in HDL-C (P < .0001) compared with normal samples (Fig. 4B and E).

Figure 3. Representative images of normal, carrier and affected Duchenne dog skeletal muscle stained with HE and dystrophin.

Asterisks (*) follow the same myofibre across serial sections of dogs with the same genotype. Arrows (↗) indicate the presence of centralized nuclei. Scale bar (100μm) applies to all images in the same panel.

Figure 4. Plasma Creatine kinase (A), Total Cholesterol (B), Low Density Lipoprotein (LDL-C) (C), nonHDL-C (D), High Density Lipoprotein (HDL-C) (E) and Triglyceride (F) levels from Normal, Duchenne and Carrier GRMD dogs.

Mean±SEM; A one way ANOVA with Tukey’s post-hoc test was used to compare means, P<0.05 (*), P<0.01 (**), P<0.001 (***), P< 0.0001 (****).

When age was analyzed, only TC levels were tested in a separate cohort of normal and affected dogs across the following age ranges: 2, 7–8, 14–15, and 28–34 months of age (Table 1). In normal dogs, TC decreased between 7 and 14 months of age (28%; P < .05), whereas in affected dogs, a significant increase was observed at 7–8 months (18%; P < .01) and 14–15 months (30%; P < .0001) when compared with 2 months levels (Table 1). Compared with normal samples, DMD was associated with a significant increase in plasma TC levels starting from the 7-month time point (P < .05) and remained significantly elevated at 14–15 and 28–34 months (P < .01), suggesting a chronically elevated state of dyslipidemia.

Table 1.

Time course of total plasma cholesterol levels in both normal and affected dogs

Age group (months)	Genotype	Sample size	Mean age (months)	Cholesterol (mg/dL)
2	Normal	14	2.2 ± 0.1	201.5 ± 7.5
	Affected	28	2.0 ± 0.1	231.5 ± 9.8
7–8	Normal	23	7.8 ± 0.4	232.9 ± 12.1
	Affected	46	7.2 ± 0.2	274.3 ± 7.7†,ǁ
14–15	Normal	7	13.8 ± 1.0	167.7 ± 9.5§
	Affected	16	14.8 ± 0.5	301.9 ± 10.4‡,ǁ
28–34	Normal	2	33.7 ± 8.7	141.5 ± 8.5
	Affected	12	28.8 ± 2.2	272.4 ± 18.1†

Data are presented as mean ± SEM.

^†P < .01.

^‡P < .0001.

^§Different from 7 months in normal dogs.

^ǁDifferent from 2 months in affected dogs.

Assessment of liver function enzymes in canine DMD

Because the liver is the main regulator of lipoprotein plasma concentrations and metabolism, the levels of plasma ALT and aspartate transaminase (AST), 2 enzymes associated with either hepatocellular injury or muscle wasting, were analyzed (Fig. 5). As expected, higher levels of both ALTand AST were observed in the plasma from DMD-affected dogs (P < .0001). In contrast, carrier samples presented with normal plasma transaminase levels (Fig. 5A and B).

Figure 5. Plasma levels of ALT (A) and AST (B) liver enzymes are elevated in in age-matched Normal, Duchenne and unaffected Carrier GRMD dogs.

Mean±SEM, A one way ANOVA with Tukey’s post-hoc test was used to compare means, P< 0.0001 (****).

Discussion

Our study is the first to document an unexpected chronic state of plasma dyslipidemia in DMD and BMD using routine, clinically certified approaches. Importantly, comparative assessment of the plasma lipid profiles isolated from a large, unmedicated cohort of DMD-affected or carrier dogs to minimize confounders confirmed dystrophin deficiency–associated lipid abnormalities at early and later stages of disease. The observation that unaffected heterozygous carrier dogs present with elevated TC and HDL-C in the absence of muscle wasting or elevated CK suggests the presence of a primary state of dyslipidemia that is not secondary to muscle degeneration.

Although the abnormal lipid profiles of the patients with DMD studied herein are of interest, these data must be taken in the context of conventional management of patients with DMD as many environmental, dietary, medical, and genetic factors, as well as pharmacological treatments, could contribute to dyslipidemia. Corticosteroid use has been shown to increase plasma TC, very-low-density lipoprotein cholesterol, LDL-C, and TGs and have variable effects on HDL-C levels.^24–28 However, others report that corticosteroid treatment (prednisone or deflazacort continued for 4 years) in patients with DMD had no effect on pre-established rates of dyslipidemia in fasted samples²⁹ although we herein report nonfasting lipid values; furthermore, in other disease settings, deflazacort is reported to lower plasma TC and LDL-C levels.³⁰ Angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, beta-blockers, and/or aldosterone antagonists are often part of pharmacotherapy of a patient with DMD, as they have been shown to improve or preserve left ventricular systolic function and delay cardiomyopathy progression,^31,32 but can also have modest effects on plasma lipid levels.³³ These findings are further compounded by transient nutritional interventions in patients with DMD, obvious inactivity and weight gain, as more than half are classified as obese by the age of 13 years.³⁴ Another limitation is the fact that puberty and its effect on TC and HDL are both delayed in boys with DMD; hence, these factors make it difficult to confirm causality between dystrophin deficiency and plasma lipid abnormalities in a patient population. On the other hand, if plasma lipoprotein abnormalities can exacerbate muscle wasting as previously shown by our group,^8,9 DMD severity could potentially be exacerbated by lipid metabolism–related disease-modifying genes. It is of interest to note that statins have been shown to attenuate the rate of muscle wasting in models of DMD,³⁵ but because no change in plasma cholesterol levels were noticed, these data might indicate abnormal regulation of the LDL receptor system in DMD in response to statins, which might also explain the high rate of dyslipidemia we report herein. Having already shown that elevated plasma lipids exacerbate DMD-associated muscle pathology in mice, our data suggest that the primary state of dyslipidemia we report may play a direct role in disease severity and DMD-associated muscle wasting. Indeed, mdx mice exhibit worsened muscle wasting and fibroadipogenic infiltration when their plasma lipoprotein levels were elevated to an atherogenic state, through either genetic manipulations or switch to high-fat diets.⁸ One can therefore speculate that dyslipidemia could enhance direct lipoprotein toxicity in weakened muscles. These lipoproteins could originate from a leaky or defective vasculature as previously reported³⁶ and exacerbate muscle damage, as shown by our group.⁸ In addition, preclinical DMD biopsies show evidence of vascular dysfunction (platelet embolism, impaired angiogenesis and endothelial cell enlargement),^36,37 and dystrophic myofibers are more susceptible to oxidative stress and free-radical–mediated injury, known to be markedly increased during hyperlipidemia.^38–41 Because phospholipids are a major constituent of the sarcolemmal membrane and thus tightly regulated, changes to phospholipid composition could also affect membrane integrity and calcium signaling and further contribute to myofiber damage, as is reported for a number of muscle wasting diseases.^42,43 Hence the new data presented in this report could improve the management of dystrophinopathies because lipid levels are routinely measured and may be targeted by pharmacotherapy, which could enhance muscle stability.

In contrast, the characterization of an unmedicated, DMD-affected cohort of dogs that present with severe muscle wasting at an early age allowed the testing of plasma lipoprotein levels in the absence of several confounders, mainly the lack of steroid use. The present study is the first to report rates of dyslipidemia (inclusive of TC and TG elevation) in dogs with DMD, an effect we demonstrate to be further exacerbated by age and that presents months before disease manifestation (by 7 months of age). Nonetheless, the most unexpected finding of the present study was the presence of severe and ubiquitous dyslipidemia in heterozygous, carrier dogs in the absence of CK elevation and overt muscle pathology. Collectively, these data are the first to signify that there may be a distinct primary, genetic link between dystrophin and the plasma accretion of lipoproteins. Whether dystrophin participates in liver and plasma lipoprotein homeostasis is unknown, although Dystrophin Dp71 is expressed in hepatocytes, and full-length dystrophin mRNA and low levels of protein have also been detected in human liver tissues.⁴⁴ As in skeletal muscle, Dp71 associates with dystroglycans, sarcogly-cans, dystrobrevins, syntrophins, and other accessory proteins in nonmuscle tissues forming the dystrophin-associated protein complex.⁴⁵ Hence, key aspects of lipoprotein sensing could be impaired in DMD, although how this occurs warrants further investigation. Indeed, studies have reported reduced expression of very-low-density lipoprotein receptor and LDL receptor–associated protein in skeletal muscles of dysferlin-null LGMD2B mice.⁴⁶ As other types of MD (including dysferlin) also express their candidate genes in the liver^47,48; this raises the possibility that other types of MDs could be part of a family of MD-related primary dyslipidemias.

The high plasma AST and ALT transaminases we report in DMD-affected samples are generally signs of hepatocellular damage, which if severe enough could rationalize the altered lipoprotein metabolism we observed in these samples. While higher plasma AST/ALT have been reported in patients with DMD,⁴⁹ muscle damage can also result in higher AST/ALT levels, making the DMD-affected results more difficult to interpret. Nonetheless, higher lipids in nonaffected carriers in the absence of muscle wasting or transaminase elevation argues against liver damage as the primary cause of dyslipidemia. However, whether loss of dystrophin causes fundamental defects in lipoprotein production, clearance, or signaling—particularly in the liver, the major organ involved in lipoprotein production and clearance—is unknown. It must be noted that others have reported that levels of fatty acid–binding protein 5 (Fabp5) and apolipoprotein AII (Apoa2) drastically fluctuate in the liver of rodents with DMD,²² which supports the concept that DMD and BMD could be considered as lipometabolic diseases, as suggested by others^10,11,23 and that abnormal lipoprotein metabolism might play a causal role in muscle wasting.^8,9 With the data presented herein, it is also of interest to note that mdx and mdx/ApoE mice also show abnormal lipid parameters compared with their non-mdx controls on a similar diet, although mice and humans show major differences in their respective LDL vs HDL distribution.⁵⁰ Meanwhile, others have shown that mice with DMD display higher plasma LDL/very-low-density lipoprotein cholesterol levels than age-matched wild-type (WT) controls.³⁵ In the event of causality between dyslipidemia causing muscle injury in DMD and other types of MD,^8,9 how plasma cholesterol might influence sarcolemma homeostasis, membrane integrity, and calcium signaling needs further investigation.

Conclusion

In conclusion, our data suggest that aberrant circulating lipid regulation is part of DMD and could act as a primary contributor to the development and progression of the muscle pathology. Because clinical and preclinical research has implicated lipid abnormalities in DMD, it may be time to consider lipid-lowering therapy as a component of DMD management.