Molecular neuroanatomy of anorexia nervosa

Abstract

Anorexia nervosa is a complex eating disorder with genetic, metabolic, and psychosocial underpinnings. Using genome-wide methods, recent studies have associated many genes with the disorder. We characterized these genes by projecting them into reference transcriptomic atlases of the prenatal and adult human brain to determine where these genes are expressed in fine detail. We found that genes from an induced stem cell study of anorexia nervosa cases are expressed at higher levels in the lateral parabrachial nucleus. Although weaker, expression enrichment of the adult lateral parabrachial is also found with genes from independent genetic studies. Candidate causal genes from the largest genetic study of anorexia nervosa to date were enriched for expression in the arcuate nucleus of the hypothalamus. We also found an enrichment of anorexia nervosa associated genes in the adult and fetal raphe and ventral tegmental areas. Motivated by enrichment of these feeding circuits, we tested if these genes respond to fasting in mice hypothalami, which highlighted the differential expression of Rps26 and Dalrd3. This work improves our understanding of the neurobiology of anorexia nervosa by suggesting disturbances in subcortical appetitive circuits.

Figure 1

Overview of the study. Genes associated with anorexia nervosa (top) are from several sources. For these studies, varying amounts of genes were associated: Negraes et al. (361 genes), Lutter et al. (186 genes), Duncan et al. (6 genes), and Watson et al. (107 genes). These studies highlighted neuropeptide signalling genes and genetic correlations with metabolic traits. After filtering, these genes are characterized in several genome-wide expression datasets (middle) to provide locations, cell types, and fasting expression differences (bottom). These resources are obtained from microarray and RNA sequencing assays of gene expression profiles obtained from the human and mouse brain. Images are from the cited publications, and Wikimedia Commons (Gray's Anatomy by Henry Vandyke Carter [public domain]; Sobotta's Human Anatomy 1908 [public domain]; users kaibara87 [changes: converted to grayscale, Creative Commons Attribution 2.0 Generic license, https://creativecommons.org/licenses/by/2.0/deed.en], Rama [ Creative Commons Attribution-Share Alike 2.0 France license, https://creativecommons.org/licenses/by-sa/2.0/fr/deed.en], and Konrad Förstner [ Creative Commons CC0 1.0 Universal Public Domain Dedication, https://creativecommons.org/publicdomain/zero/1.0/deed.en]).

Figure 2

Violin plots showing the distributions of gene expression ranks of the set of AN associated genes derived from Negraes et al. Each violin is split to show the fetal (white) and adult (light grey) brain data. Dashed vertical lines within the violins mark expression quartile borders.

Figure 3

Anatomical maps showing aggregate gene expression patterns of the anorexia nervosa associated genes from Negraes et al. in the adult brain. Regions of interest are highlighted with purple boxes. The inset thumbnail marks the current slice in pink. AUROC values range from depleted expression in dark blue to enriched in dark red with missing values in white. (PVH: paraventricular nucleus of hypothalamus; GPi: globus pallidus, internal segment; Put: putamen; VMH: ventromedial hypothalamic nucleus; SNpr: substantia nigra pars reticulata; SNpc: substantia nigra pars compacta).

Figure 4

Anatomical maps showing aggregate gene expression patterns of the AN associated genes from Negraes et al. in the fetal brain. Regions of interest are highlighted with purple boxes. AUROC values range from depleted expression in dark blue to enriched in dark red with missing values in white. (GPe: globus pallidus external segment; SNpr: substantia nigra pars reticulata).

Figure 5

Density plots of z-scored genome-wide expression within a brain region for either the adult or fetal brain reference atlases (in black). Coloured lines mark expression of the 6 genes near rs4622308.

Figure 6

Violin plots of Rps26 single-cell expression in ad libitum (red) and food-deprived (blue) conditions. The first two plots are from the Chen dataset, where expression is measured by the log of transcripts per million. The last plot is from the Campbell dataset, which measured expression by the natural log of the counts per million plus one (unique molecular identifier method). Horizontal lines mark the median expression. Connecting brackets at the top provide p-values from Mann-Whitney U tests.