24. Differential Expression Of Salivary Microrna In Anorexia Nervosa And Anxiety Disorders

Abstract

Micro ribonucleic acids (miRNAs) regulate protein translation throughout the human body, influencing diverse processes such as brain development and metabolism. Brain to saliva transfer of miRNAs may occur through the cranial nerves, glymphatic drainage or perfusion of the salivary glands. They are altered in psychiatric conditions, such as depression, but have not been studied in anorexia nervosa (AN), a disorder with both psychiatric and metabolic features. The purpose of this study was to identify salivary miRNAs “altered” in AN that may offer diagnostic and therapeutic targets. Participants were females, age 11-21 years, with: 1) restrictive-type AN, at the outset of partial hospitalization treatment; 2) anxiety (AX); and 3) healthy controls (HC). Morning pre- and post-prandial salivary samples were collected. RNA was quantified using high throughput sequencing. Partek Flow was used to align mature and premature miRNAs with the human genome. Significance analysis of microarray (SAM) identified individual miRNAs with differences between 20 AN, 20 AX and 18 HC samples. Variable importance projection (VIP) scores were used to quantify miRNA contributions to group separation on partial least squares discriminant analysis (PLSDA), and hierarchical clustering was used to visualize expression patterns across individual samples. Fasting and post-prandial miRNA levels were compared across AN and non-AN samples with 2-way analysis of variance (ANOVA). MiRNAs of interest were functionally interrogated in DIANA miRPath software. Relationships between miRNAs and medical/neuropsychiatric characteristics were investigated with Pearson correlation analyses. PLSDA and hierarchical clustering analyses revealed distinct distribution patterns for AN, AX and HC samples. Ten individual miRNAs had significant VIP scores (≥2.0) that were critical for group separation. 5 miRNAs were downregulated in both AN and AX, which may be anxiety related. 5 were upregulated in AN only, suggesting that these changes are related to malnutrition. MiRNAs upregulated in AN identified in DIANA miRPath showed enrichment (FDR < 0.05) for several KEGG pathways including fatty acid biosynthesis (p=2.06E-19, 2 miRNAs), fatty acid metabolism (p=4.00E-07, 8 miRNAs), regulation of pluripotency of stem cells (p=3.83E-06, 14 miRNAs), steroid hormone biosynthesis (p=9.43E-06, 8 miRNAs), biotin metabolism (p=3.66E-05, 2 miRNAs) and GABAergic signaling (p=0.037232, 13 miRNAs) pathways. Three miRNAs demonstrated a dynamic response to food intake by 2-way ANOVA Simultaneous Components Analysis (ASCA) of fasting and post-prandial samples in those with AN and those without AN. All 3 miRNAs tended to be lower in AN during the fasting state and higher in the post-prandial state. Individual miRNA levels were associated with both medical characteristics and neuropsychiatric measures. Salivary miRNA profiles are disrupted in adolescent females with AN, demonstrating potential utility for differentiating adolescent females with AN from peers. These miRNAs are influenced by prandial-status, target genes related to fatty acid metabolism and GABAergic signaling, and are related to neuropsychiatric measures. They yield potential insights into the pathophysiologic milieu underlying AN and may provide an objective measure for tracking AN severity, or treatment response.