Cognition, Aryl Hydrocarbon Receptor Repressor Methylation, and Abstinence Duration-Associated Multimodal Brain Networks in Smoking and Long-Term Smoking Cessation

Shile Qi, Zening Fu, Lei Wu, Vince D. Calhoun, Daoqiang Zhang, Stacey B. Daughters, Ping-Ching Hsu, Rongtao Jiang, Victor M. Vergara, Jing Sui, Merideth A. AddicottFrontiers in Neuroscience 2022 · DOI

TL;DR

This study uses a novel multimodal fusion approach to identify brain networks associated with cognitive function, AHRR gene methylation, and smoking abstinence duration in smokers and long-term ex-smokers. The findings suggest that long-term smoking cessation leads to improved cognition and structural/functional brain changes, with AHRR methylation serving as a promising biomarker for smoking recency that complements self-reported abstinence.

What Problem Does This Paper Address?

Cigarette smoking has well-documented negative impacts on cognition and brain structure/function. While smoking cessation offers health benefits, the specific neurobiological changes associated with long-term abstinence and their relationship with cognitive recovery are not fully understood. Furthermore, while DNA methylation, particularly of the AHRR gene, is a known biomarker for smoking exposure, its precise relationship with brain changes and cognitive function in ex-smokers remains to be elucidated. This paper aims to bridge this gap by investigating the multimodal brain network correlates of cognition, AHRR methylation, and abstinence duration in a cohort of smokers and long-term ex-smokers.

Methods

The researchers employed a supervised multimodal fusion analysis technique. This method integrates data from three different MRI modalities:

  • Functional MRI (fMRI): Specifically, fractional amplitude of low-frequency fluctuations (fALFF) from resting-state fMRI (rs-fMRI) to capture brain activity.
  • Structural MRI (sMRI): Gray matter volume (GMV) to assess brain tissue structure.
  • Diffusion MRI (dMRI): Fractional anisotropy (FA) values to evaluate white matter integrity.

The "supervised" aspect means that clinical variables of interest – cognitive performance (measured by the Paced Auditory Serial Addition Test, PASAT), AHRR gene methylation levels, and abstinence duration – were used as references to guide the fusion process. This allows for the identification of brain networks that not only show group differences (smokers vs. ex-smokers) but also covary with these specific clinical measures. This approach is novel in its simultaneous integration of these three MRI modalities and its direct linkage to these key clinical variables.

Datasets

The study utilized data from N=58 participants: 29 current smokers and 30 ex-smokers (abstinent for >2 years). Participants were recruited from the Little Rock, AR community.

  • MRI Data: Functional, structural, and diffusion MRI data were collected for all participants.
  • DNA Methylation Data: Saliva samples were collected from a subset of participants (21 smokers, 26 ex-smokers) for methylation analysis of the AHRR gene (specifically the cg05575921 loci).
  • Cognitive Performance Data: Assessed using the Paced Auditory Serial Addition Test (PASAT).
  • Smoking History Data: Included cigarettes per day, pack years, and duration of the most recent quit attempt.

No new benchmark datasets were introduced.

Key Insights and Ideas

  • Multimodal Brain Networks Discriminate Smokers and Ex-smokers: The fusion approach successfully identified distinct brain networks associated with cognition, AHRR methylation, and abstinence duration that differentiated current smokers from long-term ex-smokers. This highlights that these factors are reflected in a coordinated pattern of structural and functional brain alterations.
  • Long-Term Cessation Benefits Cognition and Brain Structure: Ex-smokers demonstrated more accurate cognitive performance and showed specific patterns of brain changes, including lower fALFF and higher GMV in the hippocampus complex, compared to smokers. This suggests that long-term abstinence can lead to a recovery or improvement in cognitive function and associated brain regions.
  • AHRR Methylation as a Complementary Biomarker: The study found that AHRR methylation levels were significantly different between smokers and ex-smokers and positively correlated with abstinence duration. This suggests that AHRR methylation can serve as a valuable, objective biomarker for smoking recency, complementing subjective self-reports.
  • Structural and Functional Co-alterations: The identified networks revealed co-occurring changes in brain structure (GMV, FA) and function (fALFF) in regions like the insula, frontal gyri, and lingual gyri, linking these changes to smoking abstinence and cognitive performance.

Key Results

  • Cognition-Associated Network: This network discriminated between smokers and ex-smokers and correlated with PASAT performance. Ex-smokers showed higher fALFF in the posterior cingulate cortex and middle temporal gyrus, and lower fALFF in the superior/middle frontal gyrus, parahippocampal gyrus, lingual gyrus, and anterior cingulate cortex. They also had more GMV in the lingual gyrus, fusiform gyrus, superior/middle temporal gyrus, and parahippocampal gyrus, and higher FA in the bilateral forceps major.
  • Methylation-Associated Network: This network correlated with AHRR methylation and discriminated between groups. Ex-smokers had higher fALFF in the lingual gyrus, posterior cingulate cortex, and fusiform gyrus, and lower fALFF in the superior/middle temporal gyrus, anterior cingulate cortex, and insula. They also had more GMV in the bilateral lingual gyrus, fusiform gyrus, and superior/middle temporal gyrus, and less GMV in the frontal gyrus, anterior cingulate cortex, and insula. Higher FA was observed in the middle part of the bilateral forceps major.
  • Abstinence Duration-Associated Network: This network correlated with abstinence duration and discriminated between groups. Similar patterns of fALFF and GMV differences were observed between smokers and ex-smokers as in the other networks, with ex-smokers generally showing patterns indicative of recovery.
  • Correlations: The abstinence duration-associated network positively correlated with AHRR methylation, and vice versa, reinforcing their complementary nature. The cognition-associated network showed a positive correlation with abstinence duration.

Limitations and Future Work

  • Cross-Sectional Design: The study is cross-sectional, meaning it cannot establish causality or definitively determine if observed group differences pre-dated smoking initiation. Longitudinal studies are needed to track changes over time.
  • Sample Size: While comparable to similar studies, the sample size is relatively small, particularly for the DNA methylation analysis.
  • Lack of Non-Smoking Control Group: The study only compared smokers and ex-smokers, lacking a true non-smoking control group for a more comprehensive comparison.
  • Generalizability: The sample was recruited from a specific community, which might limit the generalizability of findings.
  • Racial Differences: While groups were matched on most demographic and smoking history variables, they differed in racial distribution, which was accounted for in analyses but remains a consideration.

Future Work: The authors suggest future studies should investigate the rate and pattern of abstinence-related brain changes in ex-smokers compared to smokers and non-smokers to better understand the extent of recovery. Further research could also explore other methylation loci that may have more direct influence on brain structure and function, and investigate the similarity of methylation patterns in blood/buccal cells to those in brain cells.

Summary updated

Abstract

Cigarette smoking and smoking cessation are associated with changes in cognition and DNA methylation; however, the neurobiological correlates of these effects have not been fully elucidated, especially in long-term cessation. Cognitive performance, percent methylation of the aryl hydrocarbon receptor repressor (<i>AHRR</i>) gene, and abstinence duration were used as references to supervise a multimodal fusion analysis of functional, structural, and diffusion magnetic resonance imaging (MRI) data, in order to identify associated brain networks in smokers and ex-smokers. Correlations among these networks and with smoking-related measures were performed. Cognition-, methylation-, and abstinence duration-associated networks discriminated between smokers and ex-smokers and correlated with differences in fractional amplitude of low frequency fluctuations (fALFF) values, gray matter volume (GMV), and fractional anisotropy (FA) values. Long-term smoking cessation was associated with more accurate cognitive performance, as well as lower fALFF and more GMV in the hippocampus complex. The methylation- and abstinence duration-associated networks positively correlated with smoking-related measures of abstinence duration and percent methylation, respectively, suggesting they are complementary measures. This analysis revealed structural and functional co-alterations linked to smoking abstinence and cognitive performance in brain regions including the insula, frontal gyri, and lingual gyri. Furthermore, <i>AHRR</i> methylation, a promising epigenetic biomarker of smoking recency, may provide an important complement to self-reported abstinence duration.
BibTeX
@inproceedings{qi2022cognition,
  author = {Shile Qi and Zening Fu and Lei Wu and Vince D. Calhoun and Daoqiang Zhang and Stacey B. Daughters and Ping-Ching Hsu and Rongtao Jiang and Victor M. Vergara and Jing Sui and Merideth A. Addicott},
  title = {{Cognition, Aryl Hydrocarbon Receptor Repressor Methylation, and Abstinence Duration-Associated Multimodal Brain Networks in Smoking and Long-Term Smoking Cessation}},
  booktitle = {Frontiers in Neuroscience},
  year = {2022},
  doi = {10.3389/fnins.2022.923065},
}
Figure