Lower novelty-related locus coeruleus function is associated with Aβ-related cognitive decline in clinically healthy individuals

Animal and human imaging research reported that the presence of cortical Alzheimer’s Disease’s (AD) neuropathology, beta-amyloid and neurofibrillary tau, is associated with altered neuronal activity and circuitry failure, together facilitating clinical progression. The locus coeruleus (LC), one of the initial subcortical regions harboring pretangle hyperphosphorylated tau, has widespread connections to the cortex modulating cognition. Here we investigate whether LC’s in-vivo neuronal activity and functional connectivity (FC) are associated with cognitive decline in conjunction with beta-amyloid. We combined functional MRI of a novel versus repeated face-name paradigm, beta-amyloid-PET and longitudinal cognitive data of 128 cognitively unimpaired older individuals. We show that LC activity and LC-FC with amygdala and hippocampus was higher during novelty. We also demonstrated that lower novelty-related LC activity and LC-FC with hippocampus and parahippocampus were associated with steeper beta-amyloid-related cognitive decline. Our results demonstrate the potential of LC’s functional properties as a gauge to identify individuals at-risk for AD-related cognitive decline.


Reporting for specific materials, systems and methods
We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material, system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response. Methods n/a Involved in the study ChIP-seq Flow cytometry

MRI-based neuroimaging
There is no justification of sample size. This is a longitudinal cohort (panel) study and this could not be predicted beforehand. Only participants that engaged in 6 functional MRI runs of the task and had a PiB-PET scan at baseline were included (N=128 at baseline and up to 10 years follow-up measurements). However, taking into account previous work in the literature, we were convinced that our sample size would be adequate to achieve the aims of our study. Previous studies: -Sperling et al., 2018 (PMID: 30549303) investigated the association and interaction of A! and tau on prospective cognitive decline in normal aging and preclinical Alzheimer's disease (N=137).
- Pihlajamäki et al., 2011 (PMID: 21161449) using the same task showed that greater MTL repeated activity was correlated with worse wordlist delayed recall performance (N=60).
No data points were excluded. For the calculation of PACC5, we allowed at most one missing subtest. Missing subtests were excluded from the calculation.
To validate our imaging results, we analyzed two different fMRI datasets. The first one, the Replication Dataset, consisted of fMRI data acquired from forty-one older individuals using an alternative version of the face-name associative task. Twenty-four individuals overlapped with the main cohort but were scanned four years later using an alternative version of the face-name associative task. The other seventeen participants joined HABS later in the study and their baseline imaging and cognitive measurements were not within one year from each other and were therefore excluded from the main sample. The characteristics of the Replication Dataset are provided in Table S1. The other dataset, the Matched Dataset, consisted of a subset of 36 A!-individuals who were matched to the 36 A!+ individuals based on the age, sex and years of education distributions using propensity-based matching. The characteristics of the Matched Dataset are provided in Table S2. Our results were further validated using additional sensitivity analyses using unsmoothed data (providing the highest spatial resolution), an eroded version of the LC ROI (for the gPPI analyses), as well as grey matter density as a covariate. Each of these validation and sensitivity analyses reproduced our imaging results. N/A. There was no intervention.
All investigators (clinicians) were blind to biomarker /genetics status and imagers were blind to biomarker/APOE status and cognitive performance. One hundred twenty-eight older individuals from the Harvard Aging Brain Study underwent imaging, as well as longitudinal neuropsychological evaluations over up to 10 years. Seventy-one participants (55.46%) were female. At baseline, the mean age of the participants was 70.07 ± 8.86 (SD) and the mean education level was 15.74 ± 2.67 (SD) years. In addition, all participants had no history of medical or psychiatric disorders and were clinically unimpaired at baseline: Mini-Mental State Examination (MMSE) > 25 and Clinical Dementia Rating (CDR) = 0. Thirty-six participants (28.1%) were classified as A! positive. This information is also stated in Table 1, which also provides an overview of the differences between A! positive and negative individuals.
Participants were recruited from a longitudinal cohort followed at the Alzheimer Disease Research Center (ADRC) at Massachusetts General Hospital (MGH). In addition, participants were recruited through advertisements in local newspapers, internet sites and community-based outreach events. The Harvard Aging Brain Study has recruited more white and highly educated individuals than expected based on the New England population, therefore results may be less generalizable to other communities.
The study complied with all ethical regulations and was approved by the MGB/Partners Human Research Committee at Massachusetts General Hospital. All participants provided written informed consent following the MGB/Partners Human Research Committee regulations, and received monetary compensation after each visit.
Task (encoding of face-name associations); Mixed block/event-related design.
The task comprised events of unfamiliar and familiar face-name pairs organized within blocks of novelty and repetition, respectively. The novelty blocks consisted of 7 face-name pairs (Ni, i=1,..7). The repetition blocks consisted of 7 trials during which two face-name pairs were alternated, one male and one female. (Rj, j=1,2). The novelty, repetition and visual fixation (+) blocks, as well as the events within the blocks (Ni, i=1,…7; Rj, j=1,2; +) are depicted along with their corresponding duration. Each block was shown twice and alternated with visual fixation blocks. One functional run lasted for 4 minutes and 5 seconds, and a total of 6 functional runs were presented to each participant. Whole brain acquisition FSL version 5.0.7: brain extraction, slice timing correction, motion correction via volume realignment, normalization to the 2 mm isotropic MNI-152 EPI template. Spatial smoothing was performed using a custom ellipsoid Gaussian kernel.
We initially aligned the BOLD images to the high resolution 1 mm -T1 structural image of each subject using boundary-based registration (BBR). Subsequently, the T1 structural image was aligned with the MNI-152 template using a 3-step registration procedure: in the first step the T1 structural image was registered to the MNI-152 template using an affine, linear registration with 12 degrees of freedom. In the second step, this affine registration was refined using cost-function weighting input and reference volumes. The input and reference weighting images are provided in the Supplementary Material. In the third step, a non-linear registration was performed, which was initialized using the transformation matrix obtained from the previous step.