Molecular senescence is associated with white matter microstructural damage in late-life depression

Abstract

Objectives

We aim to investigate the association between white matter integrity and accelerated brain aging in late-life depression.

Methods

We measured senescence-associated secretory phenotype (SASP) index proteins, cognitive performance, and MRI diffusion tensor imaging (DTI) measures of fractional anisotropy and mean diffusivity based indices of white-matter microstructure measures in 56 older adults with remitted late-life depression (LLD).

Results

Higher SASP index was significantly correlated with older age (r= 0.42, p=0.001) and worse executive function performance (r=−0.27, p=0.04). After controlling for the effect of age, overall cognitive performance, and white matter hyperintensities, the association between SASP and left and right cingulate bundle mean diffusivity remained statistically significant.

Conclusions

Our data suggests that, in the context of LLD, SASP proteins are associated with microstructural abnormalities in white matter tracts in brain and worse executive function performance.

Introduction

Late-Life Depression (LLD) is common and associated with cognitive impairment and higher risk for dementia (1, 2). The mediators of risk are not well understood, but probably involve the interaction of biological mechanisms related to aging processes, like inflammatory activation, loss of proteostasis control, increased tissue remodeling markers and reduced neurotrophic support (3).

In a recent study, we developed the Senescent Associated Secretory Phenotype (SASP) index (4). This index includes signaling proteins related to inflammation, growth factors, tissue remodeling, and control of cell fate, and can be viewed as an indicator of more intense background molecular senescence changes(5, 6). In our previous study, we showed that individuals with LLD had higher SASP indexes compared to neverdepressed older adults. We also demonstrated that a higher SASP index was associated with worse cognitive global performance, executive function and processing speed as well as with whole brain gray matter atrophy in LLD. However, this study did not address whether changes in SASP index were associated with structural changes in the tracts related to the neuronal circuitry subserving executive and episodic memory (EEM) performance.

Diffusion tensor imaging (DTI) can quantify and characterize in vivo microstructural properties of the white matter (WM) tracts in the central nervous system (CNS) (7). Previous studies have demonstrated microstructural abnormalities in the WM in LLD and healthy comparators, suggesting that there is a disruption of large-scale brain tracts in neuronal networks contributing to EEM performance (8, 9). However, no previous study has examined the association between WM changes and molecular or cellular senescence. Therefore, our study aimed to evaluate the association between DTI measures, focusing on fractional anisotropy and mean diffusivity, and SASP index in older adults with major depression.

Methods

Sample

For this analysis, we included data from 56 older adults (82% female), with mean age and years of education of 73.1±6.2 and 14.3±3.0, respectively. All of the participants were enrolled in a research clinic based at the University of Pittsburgh’s NIMH-sponsored Advanced Center for Intervention and Services Research for Late-Life Mood Disorders. All LLD participants had previously met DSM-IV criteria for current unipolar MDD without psychotic features and were successfully treated to remission (i.e., Hamilton Depression Rating of 10 or less for two consecutive weeks) in pharmacotherapy and/or interpersonal psychotherapy intervention trials. The mean score on the Hamilton Depression Scale 17- item scale was 4.4±3.2. After remission of the depressive episode, the participants underwent neuropsychological testing and brain MRI scans. The neuropsychological battery included at least two tests per domain, assessing language, visuoconstructional ability, attention/information processing speed, episodic delayed memory, and executive functions. The score on the neuropsychological tests were corrected for age and education. Details of the tests included in this battery can be found in previous publications from our group (10).

Magnetic resonance acquisition and DTI analysis

Magnetic resonance (MR) scanning was performed with a Siemens MAGNETOM Trio 3-Tesla Scanner (Siemens Medical Solutions USA, Malverne, PA). T1-weighted scans were acquired using a magnetization prepared rapid gradient (MPRAGE) protocol with the following parameters. Repetition time (TR) = 2300 ms, echo time (TE) = 3.43 ms, flip angle = 9°, slice thickness = 1 mm, field of view (FOV)= 256 mm × 224 mm voxel size = 1 mm × 1 mm and 176 slices. DTI scans were acquired using a single-short spin echo sequence protocol with the following parameters. Echo time = 88 ms, repetition time = 5300 ms, flip angle = 90°, FOV = 256 mm × 256 mm, 12 diffusion directions, b-value = 1000 s/mm2, 40 slices, voxel size = 2 mm × 2 mm, and slice thickness = 3 mm.

We used a fully automated method for localizing and quantifying voxels as white matter hyperintensity on the fluid-attenuated inversion recovery images, and then converted these values to a volume (1 voxel = 4.2 mm3). Individual regions of white matter changes were summed to create a variable representing total white matter hyperintensity burden for each participant, and then were expressed as the ratio of total white matter hyperintensity volume by total white matter volume.

We used DTI based tractography methods to quantify microstructural characteristics of major white-matter fiber bundles such as the corpus callosum, cingulum, thalamic radiations, uncinate fasciculus, corticospinal tracts, inferior longitudinal fasciculus, and superior longitudinal fasciculus. Specifically, we used the Tracts Constrained by Underlying Anatomy (TRACULA) software routines implemented as part of the freesurfer software library version 5.3.0 (http://surfer.nmr.mgh.harvard.edu/). TRACULA performs automatic reconstruction of major white matter tracts or pathways using whole-brain tractography and constraining white matter tracts using anatomical priors from T1-weighted MRI scans. For the current analyses, we focused on the FA and MD measures in the uncinate fasciculus and the cingulate bundle because they are key tracts related to episodic memory and executive function performance. We calculated the average fractional anisotropy (FA) and mean diffusivity (MD) for the tracts of interest.

SASP biomarkers

The SASP biomarkers were analyzed in the plasma using a Luminex-based multiplex platform. Details of the laboratory analysis can be found in a previous publication (6). We calculated a SASP index for each participant based on the regression analysis of individual weights of biomarkers included in the SASP panel (4).

Statistical analyses and Results

The demographic and clinical characteristics of the sample are shown in the supplementary table 1. Pearson correlation analysis indicated that higher SASP index was significantly correlated with older age (r=0.42, d.f. = 54, p=0.001) and worse executive function performance (r=−0.27, d.f. = 54, p=0.04). There were no significant correlations between SASP, HDRS scores, years of education, medical comorbidity burden (i.e., measured by the cumulative illness rating scale scores), time since the first depressive episode, and the Z-scores in other cognitive domains (processing speed, language, episodic memory, visuoconstructional abilities). The SASP index was not significantly different between women and men (t= 1.43, d.f.= 54, p=0.15).

We carried out univariate linear regression analysis to evaluate the association between SASP index and DTI measures in the subjects. We used the p-value < 0.0125 as a threshold to statistical significance, based on Bonferroni correction, to reduce the risk of type I error. We showed that SASP index was significantly associated with the left cingulate bundle MD (β=0.47, F(1,54)=15.0, p<0.001, r²=0.22), right cingulate bundle MD (β=0.5, F(1,54)=17.6, p<0.001, r²=0.25), left uncinate fasciculus FA (β= −0.30, F(1,54)=5.4, p=0.02, r²=0.09), left uncinate fasciculus MD (β= 0.39, F(1,54)=, p=0.003, r²=0.15), right uncinate fasciculus MD (β= 0.40, F(1,54)=10.3, p=0.002, r²=0.16).

As age and executive dysfunction were correlated with the SASP index scores, we carried out multivariate linear analyses to control for their effects on the association between SASP index and DTI measures. After controlling for the effect of age and cognitive performance measures, only the association between SASP and left and right cingulate bundle MD remained statistically significant (β=0.29, t(3)=2.37, p=0.02, r²=0.097; and β=0.34, t(3)=2.82, p=007, r²= 0.132, respectively).

Discussion

In this study, we report the association between higher SASP index (indicating more intense molecular senescence) and higher MD in the cingulate bundle (indicating more intense loss of cellularity) in LLD, after controlling for age and overall cognitive performance. The association between the SASP index and mean diffusivity measures in the cingulate bundle had a moderate effect size after controlling for age and executive dysfunction. The current analysis extends our previous finding that SASP index was related to more intense gray matter atrophy in late-life depression. These findings suggest that more intense molecular senescence is associated with significant microstructural abnormalities in WM tracts that subserve executive function in older adults with depression.

Mean diffusivity is a proxy measure of regional cellular density in diffusion MRI. Glial cells (e.g., microglia, astrocytes, and oligodendrocytes) are the most abundant cells types in the WM tracts. Therefore, our results indicate that enhanced molecular senescence, evaluated by the SASP index, is associated with reduced glial density in specific WM tracts in the brain in older adults with depression. This finding is in line with post-mortem studies showing a significant reduction in the glial cell density in individuals with major depression(11). Moreover, glial cells can actively secrete SASP-related proteins under physiological and pathological conditions (12, 13). Therefore, we hypothesize that glial cells change their secretory phenotype in depression towards a senescent profile, increasing SASP related proteins. Such cellular and secretory senescence changes can lead to structural and functional abnormalities in circuits subserving cognitive function, and the consequent cognitive impairment observed in these individuals.

Our results should be viewed in light of the study limitations. First, this is a crosssectional study, and we did not include non-depressed comparison individuals. Therefore, we cannot establish any causal link between SASP, DTI abnormalities, and cognitive impairment in our analysis. Also, all individuals were assessed and had blood draws very soon after achieving remission from the most recent depressive episode. Thus, our findings can be viewed as a biological scar of a recent depressive episode and not necessarily as linked mechanistically to the biology of depression. Ongoing cohort studies with sequential MRI and SASP measurements will be able to disentangle the relationship between cellular and molecular senescence changes with cognitive impairment and structural brain changes in late-life depression. We found no statistically significant differences on SASP index between women and men in this sample. However, we cannot fully exclude a potential sex effect on our results since the majority of subjects included in this analysis were females (82%).

In conclusion, we showed that enhanced molecular senescence is associated with microstructural abnormalities in a key white matter tract subserving executive function performance (e.g., the cingulate bundle) in older adults with depression. Together with evidence from past studies, molecular senescence can be a potential intervention target to improve cognition and restore brain structure in these individuals.

Figure 1 –

Association between cingulate bundle mean diffusivity and SASP index.

Highlights.

• Depression is associated with enhanced molecular senescence in older adults.

• Higher SASP index leads to greater mean diffusivity in late-life depression.

• Enhanced molecular senescence can be a putative mechanism of microstructural damage in white matter tracts in brain, and worse executive function performance in older adults.