Abstract
ABBREVIATIONS
anaerobic threshold
base excess
creatine kinase
electromyography
Exome Aggregation Consortium
fraction of end tidal CO2
forced vital capacity
limb-girdle muscular dystrophy
Miyoshi myopathy
rate of perceived exertion
gas exchange ratio V’CO2/V’O2
spin echo
inversion time
ventilatory equivalent for CO2 production
ventilatory equivalent for O2 uptake
maximal oxygen uptake
mean workload during the last 3 minutes of exercise
mechanical efficiency
INTRODUCTION
Anoctaminopathies are muscle diseases caused by recessively inherited ANO5 mutations first identified in proximal limb-girdle muscular dystrophy type 2L (LGMD2L) and distal non-dysferlin Miyoshi myopathy (MMD3) [1, 2]. The phenotypic heterogeneity of anoctaminopathy has since expanded, with the mildest cases presenting as the asymptomatic elevation of creatine kinase (‘hyperCK-emia’) [1, 3]. The symptoms and signs are restricted to skeletal muscle in most cases, but occasionally patients present with a suspicion of heart involvement [4, 5]. Using muscle MRI, we and others have shown that anoctaminopathy typically involves the posterior compartments of the thighs and calves rather than the anterior compartments [6–8].
Vissing et al. recently reported positive effects from a 10-week supervised aerobic exercise regimen in patients with anoctaminopathy [9]. However, little controlled data exist measuring the effects of the disease on baseline oxidative capacity. Spiroergometric testing (breathing gas analysis during exercise) with the simultaneous analysis of blood lactate and ammonia concentrations provides not only an objective means of evaluating pulmonary, skeletal, and heart muscle involvement, but can also underline biochemical abnormalities through the monitoring of blood metabolites [10].
This study aimed to characterize a series of anoctaminopathy patients specifically emphasizing oxidative performance, an aspect that has been little studied previously. Patients were examined in detail, including by muscle MRI, changes in muscle pathology, and spiroergometric testing.
MATERIALS AND METHODS
Patients
Patients were examined at Helsinki University Hospital. All patients and healthy controls subjects provided their informed consent to participate in the study, and investigations were conducted in accordance with the 1975 Declaration of Helsinki. Furthermore, the ethics committee of the Hospital District of Helsinki and Uusimaa approved the study protocol. Total DNA was extracted from peripheral blood, and the coding exons and flanking intron sequences of the ANO5 gene (NM_213599.2) were sequenced. Primer sequences are available upon request from the authors. Variants whose pathogenicity were previously unknown were scored using PolyPhen-2 v.2.2.5 (http://genetics.bwh.harvard.edu/pph2/) [11] and Mutation Taster 2 (http://www.mutationtaster.org/) [12]. The population frequencies for variants were taken from the Exome Aggregation Consortium (ExAC, http://exac.broadinstitute.org/, accessed August 2015). Novel variants were submitted to the ClinVar database (http://www.ncbi.nlm.nih.gov/clinvar/).
Spiroergometry, control subjects, laboratory specimens, and statistical methods
Spiroergometric exercise testing was completed for 12 ANO5 variant–positive patients and 24 age- and gender-adjusted controls using follow-up venous lactate and ammonium samples (Table 1).
Spiroergometry was performed as described earlier [13]. A cannula was inserted into the left cubital vein, and the blood specimens for venous ammonia and lactate were drawn at nine time points: at rest, low exercise, maximal exercise, and at 2, 4, 6, 10, 20, and 30 min after exercise. Blood specimens were collected using the vacuum technique. Following a waste specimen of 2 ml, the lactate and ammonia specimens were placed into fluoride oxalate and EDTA syringes, respectively, centrifuged, and analyzed with a Cobas Integra 400 + analyzer (Roche Diagnostics, Mannheim, Germany). Lactate and ammonia were assayed by enzymatic methods using lactate dehydrogenase and glutamate dehydrogenase, respectively. Next, 3-ml samples for the analysis of the venous blood gases were placed in Ca-titrated Lithium heparin syringes and analyzed using a Radiometer ABL800 analyzer (Radiometer Medical, Bronshoj, Danmark). The test started with a 40-W workload increased by 40 W in 3-min steps in women and increased by 50 W in 3-min steps in men. If the patient’s reported physical condition was low, we used 20-W increases in 2-min steps or 30 W in 3-min steps. We attained a rate of perceived exertion (RPE) of 17 to 19 on the Borg scale and a gas exchange ratio V’CO2/V’O2 (RQ) of more than 1.0 for all participants.
We used an unpaired t-test to analyze the spiroergometric results as well as the results of the venous blood specimen between patients and controls; a nonparametric Mann-Whitney U test was used for parameters not normally distributed. Because of the multiple comparisons, we used a Bonferroni correction to the significance level applied. Due to the slight difference in the weight and BMI between the groups, we controlled the results of the spiroergometric variables for BMI.
The patient and control groups had slight differences in maximal level of exercise in the spiroergometric test. To control for this difference in the comparison of spiroergometric data, we normalized the data according to blood lactate levels. This is because the lactate level increases with the level of anaerobic metabolism and is therefore a measure of the level of exercise during the test.
Increase of lactate increases ventilation, which can be measured as increase of ventilatory equivalent for O2 and CO2 (VE/O2% and VE/CO2) and decrease of fraction of end tidal CO2 (FetCO2%). In the patients, VE/O2% and VE/CO2% were increased and FetCO2% decreased. This could have been interpreted as increased anaerobic metabolism, suggesting patients’ exercise tests to be more maximal than those of the controls. By controlling spiroergometric variables for maximal lactate levels in addition to BMI, we could see that the level of exercise of the patients was not greater than in the controls. Before these calculations, we had analyzed the behavior of lactate and ammonia associated with exercise, and no specific findings were seen. The respiratory quotient, RQ, could also have been used to control for the level of exercise. However, maximal lactate level serves here better as adjustment because, in contrast to RQ, lactate is not a spiroergometric variable itself.
Muscle MRI
MRI was performed using a 1.5 T Siemens system. Axial images were acquired for the pelvic, thigh, and leg muscles. The pulse sequences were T1-weighted spin echo (SE) with TR of 600 to 700 ms, TE of 15 ms, and a STIR fat suppression sequence with TR of 3200 to 4300 ms, TE of 33 ms, and an inversion time (TI) of 160 ms. Slices of 7 mm were used. The involvement of each muscle was scored on a scale of 0 to 3, where 0 indicated normal, 1 indicated minor involvement, 2 indicated moderate involvement, and 3 indicated maximal involvement. Aggregate scores were computed for the muscles of the thigh (quadriceps femoris, adductors, hamstrings, sartorius, and gracilis) and lower leg (tibialis posterior, peroneus, deep posterior compartment, soleus, and gastrocnemius). Based on five leg and thigh muscles, the maximum aggregated score was 15. Follow-up scans were performed using the same scanner and identical parameters for patients 1 (2 years between scans), 2 (4 years between scans), 4 (2 years between scans), 7 (2 years between scans), 9 (2 years between scans), and 13 (12 years between scans).
RESULTS
Screening of the ANO5 gene
We sequenced the ANO5 coding regions and flanking intron sequences in 111 patients with suspected myopathy of an undetermined cause based on symptoms and clinical findings, an elevated CK, myopathic electromyography (EMG), muscle biopsy, or muscle MRI. We identified 12 patients from 10 families as positive for ANO5 variants (Table 2). Clinical and genetic data from patients 13 and 14 were previously reported [1, 8].
The most common genetic finding was homozygous ANO5 c.2272C>T (p.Arg758Cys). Eight patients were homozygous for this variant, while the remaining four were compound heterozygous for p.Arg758Cys in combination with either c.191dupA (p.Asn62LysfsX15), c.1409T>G (p.Met470Arg), or c.1664G>T (p.Ser555Ile) (Table 2). The latter two variants have not previously been confirmed as pathogenic in anoctaminopathy. Both were predicted as ‘disease causing’ by Mutation Taster, whereas PolyPhen-2 rated p.Met470Arg as ‘possibly damaging’ and p.Ser555Ile as ‘probably damaging’. The population frequency of p.Ser555Ile was 1.2×10- 4 in European populations, whereas p.Met470Arg was not found in the ExAC database.
Clinical, neurophysiologic, and pathologic findings
Despite being relatively genetically homogeneous, our anoctaminopathy series showed some variability in clinical features (Table 2). Six of our 14 patients were female. The age of reported symptom onset ranged from 12 to 62 years, with an average age of 33 for female patients (SD 5.9 years) and 31 for male patients (SD 16 years). The disease duration ranged from 6 to 58 years. At onset, three patients complained of muscle pain, unpleasant muscle sensations in the calves, and cramps. Six patients initially experienced difficulty walking on their tiptoes, two patients complained of difficulty climbing stairs, and two patients complained of difficulty running and climbing stairs. One patient reported no muscle symptoms and an increased CK was found incidentally. During the course of the disease, muscle pain, stiffness, or cramps were common among all patients. Clinical examination showed distal lower limb muscle involvement in six patients, proximal involvement in two patients, and both distal and proximal involvement in six patients.
Plasma CK ranged from 240 to 12,290 IU/l. EMG findings were myopathic in 7 out of 12 patients tested. Four patients exhibited both neurogenic and myopathic features, whereas one patient had a normal EMG (Table 2). Muscle biopsies were obtained from 13 patients. The findings ranged from necrotizing myopathy or strong dystrophic changes to mild atrophy of type 2 fibers. Three out of nine patients who underwent cardiac ultrasound had slightly abnormal findings primarily attributed to hypertension (Table 2). ECGs showed normal findings for all patients (Table 4).
Skeletal muscle imaging
Table 3 summarizes the muscle MRI findings. Muscle T1-weighted MRI scans showed fatty replacement predominantly in the adductors, hamstrings, gastrocnemius, and soleus. Muscle MRI STIR-weighted scans showed hyperintensities suggesting myoedema primarily in the hamstrings, quadriceps, adductors, gastrocnemius, and soleus. These findings were more evident at the early stages of disease.
In six patients who completed two follow-up MRI scans, the total T1 score increased by an average of 0.6 (SD 0.7) points per year in the thighs and 0.6 (SD 0.3) points per year in the legs based on the 0 to 15 scoring scale. The STIR score does not show any significant change over time. Thus, the progression of the disease is reflected in any sequential increases in fatty degeneration observable in T1 imaging. Representative images from one patient are shown in Figs. 1A–F.
Spiroergometric testing
Table 1 shows the anthropometric and spirometric characteristics for patients and controls. During the exercise test, patients with anoctaminopathy exhibited a significantly lower maximal working capacity (Wmax/3 min), maximum oxygen uptake (V’O2max), oxygen pulse (V’O2/HR), and mechanical efficiency (Wmax/V’O2max) than those for controls (Table 4). However, the respiratory quotient (V’CO2/V’O2) tended to be higher in patients, and signs of increased ventilation demonstrated by a significantly increased ventilatory equivalent for O2 (VE/V’O2) and CO2 (VE/V’CO2) in addition to a decreased fraction of end-tidal CO2 (FetCO2) were seen in patients compared to controls.
In the blood gas analyses, patients exhibited the lowest blood pH of 7.3 at 2 to 6 min and the lowest base excess (BE) –5.5 mmol/l at 6 min after exercise, while the corresponding values in controls were pH of 7.26 at 2 min and BE –7.77 mmol/l at 4 min after exercise. For the BE values, we found a significant difference between patients and controls at 2 and 4 min after exercise (p = 0.039 and 0.013, respectively), and a significant difference in the pH values at 2 min after exercise (p = 0.048). In addition, we found significantly lower lactate readings in patients compared to controls (p = 0.049) at 2 and 4 min after exercise. However, we found no difference between patients and controls in the levels of ammonia (Fig. 2).
DISCUSSION
We studied a cohort of patients with ANO5-related dystrophy, providing a detailed clinical analysis. Anoctaminopathy has been reported in several countries, including Canada [2], the UK and Germany [6, 14], Finland [1, 3], the Netherlands [2, 15], Denmark [5], the Czech Republic [16], and Italy and other countries in Europe and worldwide [17–19]. We found that 11% of our suspected LGMD patients had ANO5 mutations, supporting that they are a common cause of myopathy in northern Europe.
We found the variant c.2272C>T (p.Arg758Cys) in all of our patients, appearing as homozygous in 10 patients from 8 families and as compound heterozygous with another variant in 4 patients from 3 families. This finding is not surprising since c.2272C>T has a carrier frequency of 0.35% in Finland, thus explaining most known Finnish cases [3]. Among the compound heterozygous patients, one had the c.191dupA (p.Asn62Lysfs15X) variant, which is common in other Northern European populations [14]. One patient had the c.1664G>T (p.Ser555Ile) variant, which was recently found in the heterozygous state in a patient sequenced as part of a cohort of patients with LGMD or unspecified myopathy [17]. Bioinformatic predictions support the pathogenicity of this variant. However, since carriers are found in European populations, it is likely that a second hit is necessary on the other allele for this variant to cause disease, as was the case in our patient. Finally, two siblings (Family J, patients 11 and 12) presented with previously unknown variant c.1409T>G (p.Met470Arg). This variant received high scores for pathogenicity using bioinformatics tools, and was not found in public exome databases.
The clinical findings in our anoctaminopathy patients correspond to previous descriptions, featuring asymmetric proximal or distal muscle involvement and myalgia [1–3, 14–20]. We found a tendency towards asymmetric muscle involvement, a feature of anoctaminopathy not typically reported for other forms of LGMD [2, 21] with the exception of facioscapulohumeral muscle dystrophy. Echocardiographic abnormalities were documented in only three patients, all associated with hypertension.
Our findings provide further confirmation of the variable age of onset of anoctaminopathy. Even in the presence of the homozygous p.Arg758Cys variant, initial symptoms were reported between the ages of 12 and 62 years. The variable age of onset suggests the presence of genetic or environmental disease modifiers. Previous reports have well-documented an association between being male and anoctaminopathy [3, 22]. However, we found no significant difference in the average age of disease onset based on gender, although the variance for the age of onset was greater among the men in our study.
Earlier radiologic studies by us and others demonstrated that anoctaminopathy affects lower leg muscles, suggesting predominantly posterior compartment involvement in both the thighs and the legs. The tibialis anterior, gracilis, and sartorius are better spared, similar to other LGMDs, indicating that MRI may not be the best diagnostic tool [1, 6–8]. The results presented here further demonstrate that fatty degeneration tends to increase as the disease progresses. Therefore, T1-weighted MRI may be useful as a follow-up tool or as an assessment tool for potential treatment. The STIR technique is a sensitive tool in early diagnostics, but not in the later stages of disease. Furthermore, EMG results remain variable, featuring both myopathic and neurogenic findings.
To our knowledge, this is the first study to explore the oxidative capacity of anoctaminopathy patients compared to an anthropometrically matched control population. Patients with anoctaminopathy showed significantly lower oxygen uptake than controls, which may serve as a primary feature of the disease or secondary to deconditioning provoked by muscle weakness and pains. Additionally, the patients’ oxygen pulse (V’O2/HR) was lower than controls, most likely associated with the diminished capacity of the peripheral muscles [23, 24]. The level of maximal oxygen uptake reported here corresponds to that measured by Vissing et al. [9] before regular aerobic training.
In patients, we found an excess of ventilation accompanied an increase in the ventilatory equivalents for O2 and CO2. In addition, more than half (7/12 patients) had FetCO2 lower than 4.5% during maximal exercise suggesting slight hyperventilation. By contrast, controls showed normoventilation, although we expected them to display a stronger respiratory compensation due to stronger metabolic acidosis associated with exercise compared to patients. The spirometric findings did not explain this inconsistency. The forced vital capacity (FVC) was slightly lower in patients than in healthy controls, but only one patient (patient 9) developed restrictive ventilatory impairment (FVC < 80% of the predicted value), suggesting that ventilatory function was not the cause of the increased ventilation. The increased ventilation might be explained by the lowered aerobic exercise capacity of the muscles leading to secondary hyperventilation during exercise. In addition, the reduced mechanical efficiency (Wmax/V’O2max) most likely reflected this finding, indicating an increased need for oxygen uptake related to the attained workload.
Patients’ lactate and ammonia levels tended to be lower than those among controls. This might be explained by the lower exercise capacity due to muscle disease. However, the RQ levels of patients tended to be higher than those among controls indicating at least the same level of aerobic capacity among both patients and healthy controls. We should note that an increase in ventilation may partly explain the elevated RQ levels we observed in patients. We found no specific findings or indications of metabolic muscular myopathy.
In conclusion, our report confirms the importance of the recessive ANO5 variants as causes of muscle disease with significant variability in clinical severity. In addition, we found a new disease-associated variant, and provide an initial and detailed case–control comparison suggesting a decreased oxidative capacity. Our results support aerobic training as a useful intervention to mitigate the symptoms of anoctaminopathy. Further studies featuring a larger sample of patients are needed to confirm our findings.
CONFLICTS OF INTEREST
The authors have no conflicts of interest to report.
Footnotes
ACKNOWLEDGMENTS
We thank all individuals who participated in this study. The authors wish to thank the following for their financial support of this research: Hospital District of Helsinki and Uusimaa (for S.K-E., M.A., and E.Y.), the Academy of Finland (for E.Y.), and the Folkhälsan Research Foundation (for A.-E.L.).