We characterized structural disease load using volumetric me

We characterized structural disease load using volumetric measures of the caudate, putamen, global gray and global white matter that show sensitivity towards HD-related changes (Tabrizi et al., 2012). These structural markers of disease load were included in a systematic examination of both phalloidin activity and connectivity. Brain activity was measured during performance of a motor or cognitive task using task-fMRI and brain connectivity within cognitive and motor networks using resting-state fMRI (rsfMRI). This allowed us to perform a comprehensive, unbiased whole-brain assessment of potential markers of compensation for neurodegeneration in preHD.
Materials and Methods
Results
Discussion In contrast, brain activity and connectivity centered in the left hemisphere showed no evidence of compensatory changes in activity. A functionally more resilient right hemisphere is consistent with findings from previous studies (Lambrecq et al., 2013; Muhlau et al., 2007), which have already indicated that HD-pathology is in a subtle, but robust fashion, leftward biased. Use and stress-related neuronal demands, potentially including excitotoxic mechanisms, may be more pronounced in the dominant left hemisphere and may underlie asymmetry (Jenkins et al., 1998). The notion of a subtle but reproducible larger left-hemispheric deterioration in HD requires further investigation in terms of the pathophysiological underpinnings of this asymmetry. For example, these changes could be related to a more metabolically active left hemisphere with higher energy demands (Mochel et al., 2012), thus making it more susceptible in HD, in which bioenergetic defects are well documented (Ross and Tabrizi, 2011). Alternatively, it may reflect a use-dependent possible prion-like spread of mutant huntingtin resulting in more subtle, but extensive damage in the left hemisphere in a functional connectivity-dependent fashion (Ross et al., 2014). It is surprising that despite evidence of compensation in cognitive networks associated with working memory performance, neither the task-based nor the rsfMRI-based effective connectivity analyses provided evidence of a compensatory mechanism in the motor system. For task-based fMRI, this could potentially be due to the use of a motor task that was insufficiently challenging to participants to engage compensatory processes. Of note, a SFM task previously studied in preHD with more demanding difficulty levels than those used in the current study revealed some evidence of compensation Klöppel et al. (2009). In this study, participants were required to memorize an irregular 10-item sequence of finger movements. Here, we aimed to remove the working memory component from the SFM task and replace it with an independent working memory task. Importantly the compensation model used in the previous study did not consider differing levels of performance within preHD and healthy participants, as did the current compensation model. Differences between the studies could therefore relate to the precise compensation model used, emphasizing the importance of operationally and explicitly defining neural compensation. Our rsfMRI-based analysis of functional coupling between key regions of the motor system did not provide any additional insights, as no regions were sufficiently correlated with the seed region to enter compensation analyses. Based on previous evidence, we did expect to see correlations between activity within the M1 and that of other regions of the brain, particularly regions within the motor network. We did identify significant correlations, but these were present at lower thresholds. Given that our thresholds were defined a priori and that the connectivity parameters were simply extracted for inclusion in our compensation model, we were unable to report lower threshold findings. For future longitudinal analyses, we will modify our approach to include regions of the motor network that are temporally correlated.