Self-reported Sleep Problems Related to Amyloid Deposition in Cortical Regions with High HOMER1 Gene Expression

Abstract

Sleep problems are related to the elevated levels of the Alzheimer's disease (AD) biomarker β-amyloid (Aβ). Hypotheses about the causes of this relationship can be generated from molecular markers of sleep problems identified in rodents. A major marker of sleep deprivation is Homer1a, a neural protein coded by the HOMER1 gene, which has also been implicated in brain Aβ accumulation. Here, we tested whether the relationship between cortical Aβ accumulation and self-reported sleep quality, as well as changes in sleep quality over 3 years, was stronger in cortical regions with high HOMER1 mRNA expression levels. In a sample of 154 cognitively healthy older adults, Aβ correlated with poorer sleep quality cross-sectionally and longitudinally (n = 62), but more strongly in the younger than in older individuals. Effects were mainly found in regions with high expression of HOMER1. The anatomical distribution of the sleep-Aβ relationship followed closely the Aβ accumulation pattern in 69 patients with mild cognitive impairment or AD. Thus, the results indicate that the relationship between sleep problems and Aβ accumulation may involve Homer1 activity in the cortical regions, where harbor Aβ deposits in AD. The findings may advance our understanding of the relationship between sleep problems and AD risk.

Keywords: HOMER1; Alzheimer’s disease; amyloid; gene expression; sleep.

Figure 1

Regional Aβ levels and sleep problems. Left: The surface plots show regions where self-reported sleep problems and Aβ accumulation are significantly stronger correlated in the younger (50–67 years) compared with the older (68–81 years) age group. Right: Bubble plots of the relationship between sleep problems (PSQI global score) and global cortical Aβ levels (amyloid factor expressed in Z-scores) within each age group.

Figure 2

Validation sample. The different PSQI-Aβ relationship in the younger versus the older part of the sample (left column) was tested in an independent sample of cognitively normal older adults, where Aβ was measured by Aβ1–42 in CSF (right column). As can be seen, a nominal positive relationship between sleep problems and amyloid levels was seen in the young–old in both samples. In the old–old, there was a negative relationship in the PET sample and a lack of relationship on the CSF sample.

Figure 3

Relationships between Aβ levels and sleep problems in the youngest group. Relationship between self-reported sleep problems and Aβ accumulation in the youngest participants (50–67 years) was tested vertex-wise across the cortical surface and corrected for multiple comparisons across space. Red–yellow colors represent regions demonstrating a positive relationship between self-reported sleep problems and amyloid accumulation.

Figure 4

Aβ levels in AD and gene expression. Left: Regions with significantly higher levels of Aβ in MCI/AD patients compared with cognitively normal controls. Right: Bubble plot of the relationship between the patients versus controls differences in Aβ accumulation across 34 cortical regions (gamma values) and regional HOMER1 expression levels (top) and the strength of the sleep-Aβ relationship in the youngest group (gamma values) versus the patients-controls differences (bottom).

Figure 5

Relationship between sleep problems related Aβ accumulation and HOMER1 expression. Top: Regional expression of HOMER1 in 34 cortical regions in the left hemisphere. Bottom: Bubble plot of the relationship between regional HOMER1 expression levels and the strength of the sleep-Aβ relationship in the youngest group (gamma values). The bubbles are scaled by the group difference in Aβ accumulation between controls and MCI/AD patients. The clustering of large bubbles to the right and to the top of the plot illustrates that regions with high levels of HOMER1 expression and sleep-related Aβ accumulation also show more Aβ accumulation in AD patients.