The role of the locus coeruleus noradrenergic system in inhibitory control

Response inhibition is an important function underpinning the appropriate control of everyday behavior. The inability to prevent or stop inappropriate responses is a symptom in many neurological and psychiatric disorders (Passamonti et al., 2018). There is converging evidence from animal models (Eagle and Baunez, 2010; Bari et al., 2011; Bari and Robbins, 2013) to human psychopharmacology (Chamberlain, 2006; Robinson et al., 2008; Chamberlain et al., 2009), showing that the noradrenergic system facilitates the inhibitory processes underpinning rapid action cancellation. Specific cortical areas are also implicated in response inhibition, including the inferior frontal gyrus (rIFG) and pre-supplementary motor area (pSMA; Chambers et al., 2006; Frank et al., 2007; Duann et al., 2009; Forstmann et al., 2012; Aron et al., 2014; Rae et al., 2015). Cerebral noradrenergic innervation arises from the locus coeruleus, in the brainstem. By exploiting the accumulation of iron-rich neuromelanin, magnetization-transfer sensitive magnetic resonance imaging sequences can be used to image the locus coeruleus (Betts et al 2019; Ye et al 2020). Such studies have associated locus coeruleus signal intensity with diverse cognitive function in health (Liu et al., 2020) and disease (O’Callaghan et al., 2020). We hypothesize that variations in the locus coeruleus integrity (as quantified by changes in signal intensity using a magnetization transfer weighted MRI sequence) reflect variation in changes in noradrenergic-dependent processes of inhibitory control. We focus on healthy adults over the age of 50 years, because of the robustness of locus coeruleus imaging signals in older adults (Zecca et al., 2004). Variation in performance might also be expressed relative to the degree to which a participant has “successfully aged” cognitively and maintained cognitive ability on par with early adult life - in contrast to “unsuccessfully aged” cognitively with decline in cognitive ability. To estimate this change in cognitive ability, one can compare current fluid intelligence to crystallized intelligence. This difference approximates the degree to which a participant has sustained or changed their cognitive ability (McDonough et al., 2016). We predict that “unsuccessful ageing” (with lower fluid than crystallised intelligence) is accompanied by reduced locus coeruleus signal and poor inhibitory control. There is inconsistent evidence for an association between cognitive functions and locus coeruleus subregions. In Parkinson’s disease, where the integrity of locus coeruleus predicts the improvement of response inhibition following noradrenergic treatment, the caudal subregion of the locus coeruleus is more impaired compared to central and rostral subregions (O’Callaghan et al., 2020). In healthy ageing, however, age-related neural loss predominates in the rostral locus coeruleus (Betts et al., 2017) and has been associated with decline in cognitive functions (Dahl et al., 2019). Here, we will estimate the association between the integrity of rostral, central, and caudal locus coeruleus subregions with response inhibition performance. In this cross-sectional study we will use neuromelanin-sensitive MRI to investigate the relationship between locus coeruleus integrity and inhibitory control in cognitively normal healthy adults from the Cambridge Centre for Ageing and Neuroscience cohort (Cam-CAN; Shafto et al., 2014), an open-access, population-based dataset. We will focus on a subset of the cohort where inhibitory control was assessed with a Stop-Signal Task (SST), in which participants perform a reaction time task but must occasionally cancel an action after it has been initiated. Our study has five main differences with respect to an earlier analysis of the Cam-CAN cohort (Liu et al., 2019): (i) we will extract the locus coeruleus signal in an atlas-based segmentation pipeline using a probabilistic atlas generated with high-resolution 7T images, which provides unbiased estimation, with superior accuracy and reliability compared to manual and semi-automatic segmentation approaches (Ye et al., 2020). This is especially important with the low resolution of the Cam-CAN MRI MT-weighted images, that limits the precision of manual segmentation approaches. (ii) we will focus on healthy adults aged above 50 years. Neuromelanin accumulates with age (Zecca et al., 2004). Below 50 years the locus coeruleus neurons may not yet be sufficiently pigmented to allow reliable inference on structural integrity by neuromelanin-sensitive MRI. (iii) We will adopt recent consensus recommendations (Verbruggen et al.,2019) to estimate individual efficiency of response inhibition in terms of the stop-signal reaction times (SSRT), and use a hierarchical Bayesian estimation of an ex-Gaussian race model (Matzke et al., 2013). (iv) To assess the specificity of the predicted relationship between locus coeruleus integrity and response inhibition, we will examine whether changes in inhibitory control are related to changes in cortical functional connectivity as well as to regional variations in grey matter volume. For this purpose, we will extract fMRI functional connectivity and voxel-based morphometry measures from functional masks of the preSMA and rIFG regions defined in a previous fMRI study employing a stop-signal task (Tsvetanov et al. 2018). (v) We will adopt age-related changes in participants’ fluid intelligence relative to fluid intelligence in their early adult life (as reflected by crystallized intelligence) as a measure of successful versus unsuccessful aging (McDonough et al., 2016).