Altered somatosensory neurovascular response in patients with Becker muscular dystrophy

Abstract Introduction Patients with dystrophinopathies show low levels of neuronal nitric oxide synthase (nNOS), due to reduced or absent dystrophin expression, as nNOS is attached to the dystrophin‐associated protein complex. Deficient nNOS function leads to functional ischemia during muscle activity. Dystrophin‐like proteins with nNOS attached have also been identified in the brain. This suggests that a mechanism of cerebral functional ischemia with attenuation of normal activation‐related vascular response may cause changes in brain function. Methods The aim of this study was to investigate whether the brain response of patients with Becker muscular dystrophy (BMD) is dysfunctional compared to that of healthy controls. To investigate a potential change in brain activation response in patients with BMD, median nerve somatosensory evoked stimulation, with stimulation durations of 2, 4, and 10 s, was performed while recording electroencephalography and blood oxygen level‐dependent (BOLD) functional magnetic resonance imaging. Results Results in 14 male patients with BMD (36.2 ± 9.9 years) were compared with those of 10 healthy controls (34.4 ± 10.9 years). Compared to controls, the patients with BMD showed sustained cortical electrical activity and a significant smaller BOLD activation in contralateral primary somatosensory cortex and bilaterally in secondary somatosensory cortex. In addition, significant activation differences were found after long duration (10 s) stimuli in thalamus. Conclusion An altered neurovascular response in patients with BMD may increase our understanding of neurovascular coupling and the pathogenesis related to dystrophinopathy and nNOS.


| INTRODUC TI ON
and Becker (BMD) muscular dystrophies, collectively called dystrophinopathies, are X-linked and caused by mutations in the dystrophin gene. DMD and BMD are characterized by progressive muscle wasting, cardiomyopathy, and a reduced life expectancy (Bushby et al., 1993(Bushby et al., , 2010aEmery, 2002). However, a slower progression and less severe phenotype are seen in patients with BMD compared to DMD. Patients with BMD, unlike patients with DMD, typically carry in-frame mutations of the dystrophin gene preserving some residual dystrophin function (Anthony et al., 2011). In addition to muscle and cardiac involvement, dystrophinopathies also involve cognitive symptoms. Patients with DMD score one standard deviation below normal in IQ tests (Cotton, Voudouris, & Greenwood, 2001), and behavioral and psychiatric abnormalities have also been reported (Haenggi & Fritschy, 2006;Waite, Brown, & Blake, 2012;Waite, Tinsley, Locke, & Blake, 2009;Young et al., 2007). Patients with BMD show similar deficits, though with less severity and more variability (Chamova et al., 2013;Melo et al., 1995).
The pathophysiology of brain involvement in dystrophinopathies is not yet understood.
Reduced or absent dystrophin expression causes the membranebound dystrophin-associated glycoprotein complex (DGC), present in skeletal and cardiac muscle cells (Hoffman & Kunkel, 1989;Matsumura & Campbell, 1994) as well as the neuromuscular junction and the brain (Nichols, Takeda, & Yokota, 2015) to be dysfunctional.
This in turn may affect the regulation and distribution of brain blood flow due to NO's vasodilatory effect exerted on the smooth muscle cells of the brain arterioles (Attwell et al., 2010).
In patients with DMD, compared with matched controls, structural abnormalities such as smaller total brain volume, reduced cerebral gray matter volume, and perfusion deficits were recently reported (Doorenweerd et al., 2014(Doorenweerd et al., , 2017. Functional connectivity studies showed an altered neurovascular activation in DMD patients with blood oxygen level-dependent (BOLD) changes in the motor cortex measured by resting state functional magnetic resonance imaging (fMRI) (Lv et al., 2011). Such studies suggest that dystrophin or proteins in the DGC play an important functional role in the brain.
Using positron emission tomography (PET), reduced brain tissue glucose metabolism in sensorimotor areas was reported in patients with DMD (Lee et al., 2002). Functional brain MRI studies in patients with DMD are few, and no studies have been undertaken with regard to BMD. Reports from a different muscular disease, myotonic dystrophy, show some evidence of abnormalities in the BOLD response, both resting state and task-induced responses (Okkersen et al., 2017).
We hypothesized that a deficiency in dystrophin, most likely associated with a reduced nNOS expression in the brain similar to that shown in dystrophin-deficient muscle, leads to an altered neurovascular response in patients with BMD. As a result, the neurovascular coupling would be altered in BMD patients compared with healthy subjects. An attenuation of NO production by reduced nNOS levels would cause decreased cerebrovascular responsiveness to neuronal stimulation causing a nonsufficient supply of oxygen and glucose, in particular during high demanding periods of repeated brain activation. In this study, we apply both somatosensory evoked potentials (SEP) and fMRI allowing a comparison of the neurovascular coupling with median nerve stimulation in the sensorimotor brain area.

| Participants
Fourteen males affected by BMD, carrying previously reported mutations in the dystrophin gene, were included (Witting et al., 2014).
Ten healthy males with comparable handedness and age were studied for comparison (Table 1). The Ethics Committee of the Capital Region of Denmark approved the study, and all subjects gave written informed consent according to the Helsinki Declaration before participating. This study was part of a larger setup in which treatment with sildenafil was investigated and compared in the patients (Lindberg et al., 2017;Witting et al., 2014). All subjects underwent two identical somatosensory electrical stimulation tasks: one outside the scanner with recorded EEG and one inside the scanner during BOLD-fMRI recording. All parameters were kept constant between the two runs.

| Electrical stimulation
The delivery of the electrical impulses was carried out using a DS7A & DS7AH high-voltage constant current electrical stimulator (Digitimer Ltd, United Kingdom). Median nerve stimulation was performed on the dominant hand, as determined by the Edinburgh Handedness Scale (Oldfield, 1971  The stimulation paradigm consisted of three different stimulation block lengths (2, 4, and 10 s). All blocks were repeated 10 times in a random order with a stimulation frequency of 5 Hz and an interstimulus block interval of 30 s to allow the BOLD response to settle down ( Figure 1). This adds to a total of 30 stimulation blocks with a total of 800 stimulations during the run.

| Magnetic resonance imaging
Magnetic resonance imaging was performed on a 3.0T Philips Achieva scanner (Philips Medical Systems, Best, The Netherlands) using a 32-channel head coil. Anatomical images were acquired using a MPRAGE three-dimensional turbo field-echo sequence (137 sagittal slices of 1.1 mm thickness; in-plane resolution 1 × 1 mm 2 ; TR 6,900 ms; TE 2.8 ms; TI 900 ms; and flip angle 9°).
Acquisition of the functional images was performed using an

| Somatosensory evoked potentials
The recording of the somatosensory evoked potentials (SEP) outside the scanner was done using a 32-channel actiCap system (Brain Products GmbH, Munich, Germany) placed according to the international 10-20 standard electrode placement system. The participants were placed on the same table as used inside the MRI scanner and the SEP data from channels C3 and C4 were acquired at 5 kHz using the BrainVision Recorder software (Brain Amps GmbH, Germany).
In total, 800 electrical median nerve stimulations were acquired.
For each subject, an average of all 800 somatosensory evoked potentials (SEP) from 1 to 200 ms was computed, followed by averaging the subjects within the two groups individually.
For all the stimulation blocks, each stimulation number in the train of stimulations, s n , was averaged individually to investigate the time effect of each stimulus in the train ( Figure 2).

| Postprocessing
Postprocessing and statistical analysis of the MRI data were per-

| RE SULTS
Age and handedness were not statistically different between the two groups ( Table 1). The intensity of the applied stimulus to the dominant median nerve was (mean ± SD) 4.29 ± 1.07 mA in the BMD group and 4.81 ± 1.07 mA in the control group, which was not significantly different when compared using a Mann-Whitney U test (p = .26).
The voxelwise group statistics are reported in Figure 3, with the list of significant clusters summarized in Table 2.

| Blood oxygen level-dependent response
The two-second median nerve stimulation elicited a significant posi- Regression analysis of the atlas-based regions showed a significant decrease with stimulus duration in contralateral SI in both groups (p < .001), but with no group difference when comparing slopes. Neither the SII nor thalamus showed any significant decrease in BOLD amplitude or a significant group difference as the effect of stimulus duration.

| Somatosensory Evoked Potentials
No significant difference (p = .14) was found comparing the N20-P26 SEP amplitude of the patient and control groups.
The RMS value of each stimulus number calculated from equation 1 showed that patients with BMD had a significantly (p < .02) increased power when compared to healthy controls from stimulus 9 and onwards ( Figure 4).

| D ISCUSS I ON
Investigating the neurovascular response in patients with BMD, we found a decreased BOLD response contralateral to the stimulation in the primary somatosensory cortex (SI), as well as bilaterally in the secondary somatosensory cortex (SII), and thalamus, compared with healthy controls. In addition, a sustained EEG activity was This interpretation is supported by reports of decreased expression of nNOS in neurons that lack dystrophin (Sogos, Reali, Fanni, Curto, & Gremo, 2003) and corresponds well with the decrease in muscle nNOS found in our patient population (Witting et al., 2014).
With dystrophin present in neurons and the microvasculature of the brain, a cerebral functional ischemia may exist similar to that of skeletal muscles in the patients with BMD (Rando, 2001), though with a less remarkable clinical impact. Insufficient NO in the microvasculature of skeletal muscle is known to increase the susceptibility to functional ischemia, which leads to muscle injury during prolonged hypoxic muscle strain (Martin et al., 2012;Thomas, 2013). In the present study of patients with BMD, treatment with a phosphodiesterase 5 (PDE5) inhibitor resulted in a significant increase in the BOLD signal in somatosensory and visual areas, which indicates that the NO-cGMP cascade could be involved (Lindberg et al., 2017).
As previously reported in these patients, blood flow response in the brachial artery was reduced during exercise (Witting et al., 2014), though without improved response after treatment with sildenafil, in contrast to the current findings.
The blood flow changes represent a large artery response, but are still in line with reports on dysfunctional blood flow in the dystrophinopathies. Baseline cerebral perfusion is regulated by both large artery and small artery responses, and influences the size of the BOLD response due to possible vascular ceiling effects (Birn & Bandettini, 2005;Buxton, Uludağ, Dubowitz, & Liu, 2004) which could affect brain activation BOLD data.
A lower CBF has been reported in young DMD boys compared to healthy age-matched controls (Doorenweerd et al., 2017), which indicates that the patients could suffer from a dysfunctional stimulus-induced CBF regulation leading to a decreased BOLD response.
The neuronal output to the vascular system during the median nerve stimulation was increased in the BMD group (Figure 4)

Patients Controls
Stimulus number neural activity and vascular response is often considered nearly linear in the healthy brain (Li & Freeman, 2007;Logothetis, Pauls, Augath, Trinath, & Oeltermann, 2001), but a change in the coupling pattern was observed in the present study. A similar uncoupling of neuronal activity and metabolism has been reported in patients with mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes (MELAS) where episodes of a high paroxystic activity occur concurrently with a diminished metabolism and cerebral flow (Yeh et al., 2013).
A decreased activation was seen bilaterally in thalamus for the 4-and 10-s stimulation periods in the patients with BMD vs. controls. The thalamus, in addition to passively relaying information to the cortex, appears to be involved in response magnitude, firing mode, and synchrony of neurons (Saalmann & Kastner, 2011). The changes in neuronal activation discovered in this study could suggest changes in magnitude modulation in BMD patients, with altered sensory inputs in the thalamus leading to an altered adaptation measured in the cortex. A thalamic involvement in higher order functions has previously been described (Saalmann & Kastner, 2011). Our results, however, suggest that the thalamic activation represents the stimulation of the somatosensory cortical areas and that this is merely a consequence of a general change in neurovascular coupling, leading to a lower BOLD response in thalamus.
Stimulation strength was kept constant between the two groups, and variability in stimulation intensity could therefore not explain the intergroup differences in the results. A previous study on cortical excitability in patients with DMD did not show any difference when compared to healthy controls (Yayla et al., 2008).
In our study, the group comparison showed more prominent differences in the secondary somatosensory cortex in the control group compared to the patients with BMD ( Figure 3). According to Backes et al. (Backes, Mess, van Kranen-Mastenbroek, & Reulen, 2000), the BOLD response remains constant regardless of the applied stimulus intensity between 50 and 100% of motor threshold.
Stimulus strength is therefore unlikely to have caused any group differences in the SII.
The current study showed a change in neurovascular response in BMD patients compared with controls, which could represent a change in the neurovascular coupling. A further investigation of the functional impact of the insufficient neurovascular coupling in patients with BMD needs to be established. Further studies are needed to determine whether a deficient neurovascular coupling could impact the cognitive function.

ACK N OWLED G M ENTS
Thanks to Bryan T. Haddock for proofreading and commenting the presented findings. Funding was received from LUCENS, The Gangsted Foundation, The Arup and Hvidt Foundation, The Jascha Foundation, The AP Møller Foundation, and The Lundbeck Foundation.

CO N FLI C T O F I NTE R E S T
None of the authors have any conflict to declare.