Polygenic Analyses Show Important Differences Between Major Depressive Disorder Symptoms Measured Using Various Instruments.

BACKGROUND: Symptoms of major depressive disorder (MDD) are commonly assessed using self-rating instruments like the Patient Health Questionnaire-9 (PHQ-9) (current symptoms) and the Composite International Diagnostic Interview Short-Form (CIDI-SF) (worst-episode symptoms). We performed a systematic comparison between them for their genetic architecture and utility in investigating MDD heterogeneity. METHODS: Using data from the UK Biobank (n = 41,948-109,417), we assessed the single nucleotide polymorphism heritability and genetic correlation (rg) of both sets of MDD symptoms. We further compared their rg with non-MDD traits and used Mendelian randomization to assess whether either set of symptoms has more genetic sharing with non-MDD traits. We also assessed how specific each set of symptoms is to MDD using the metric polygenic risk score pleiotropy. Finally, we used genomic structural equation modeling to identify factors that explain the genetic covariance between each set of symptoms. RESULTS: Corresponding symptoms reported through the PHQ-9 and CIDI-SF have low to moderate genetic correlations (rg = 0.43-0.87), and this cannot be fully attributed to different severity thresholds or the use of a skip structure in the CIDI-SF. Both Mendelian randomization and polygenic risk score pleiotropy analyses showed that PHQ-9 symptoms are more associated with traits that reflect general dysphoria, whereas the skip structure in the CIDI-SF allows for the identification of heterogeneity among likely MDD cases. Finally, the 2 sets of symptoms showed different factor structures in genomic structural equation modeling, reflective of their genetic differences. CONCLUSIONS: MDD symptoms assessed using the PHQ-9 and CIDI-SF are not interchangeable; the former better indexes general dysphoria, while the latter is more informative about within-MDD heterogeneity.

Deep learning-based phenotype imputation on population-scale biobank data increases genetic discoveries.

Biobanks that collect deep phenotypic and genomic data across many individuals have emerged as a key resource in human genetics. However, phenotypes in biobanks are often missing across many individuals, limiting their utility. We propose AutoComplete, a deep learning-based imputation method to impute or 'fill-in' missing phenotypes in population-scale biobank datasets. When applied to collections of phenotypes measured across ~300,000 individuals from the UK Biobank, AutoComplete substantially improved imputation accuracy over existing methods. On three traits with notable amounts of missingness, we show that AutoComplete yields imputed phenotypes that are genetically similar to the originally observed phenotypes while increasing the effective sample size by about twofold on average. Further, genome-wide association analyses on the resulting imputed phenotypes led to a substantial increase in the number of associated loci. Our results demonstrate the utility of deep learning-based phenotype imputation to increase power for genetic discoveries in existing biobank datasets.

Polygenic analyses show important differences between MDD symptoms collected using PHQ9 and CIDI-SF.

Symptoms of Major Depressive Disorder (MDD) are commonly assessed using self-rating instruments like the Patient Health Questionnaire 9 (PHQ9, for current symptoms), and the Composite International Diagnostic Interview Short-Form (CIDI-SF, for lifetime worst-episode symptoms). Using data from the UKBiobank, we show that corresponding symptoms endorsed through PHQ9 and CIDI-SF have low to moderate genetic correlations (rG=0.43-0.87), and this cannot be fully attributed to different severity thresholds or the use of a skip-structure in CIDI-SF. Through a combination of Mendelian Randomization (MR) and polygenic prediction analyses, we find that PHQ9 symptoms are more associated with traits which reflect general dysphoria, while the skip-structure in CIDI-SF allows for the identification of heterogeneity among likely MDD cases. This has important implications on factor analyses performed on their respective genetic covariance matrices for the purpose of identification of genetic factors behind MDD symptom dimensions and heterogeneity.

Tripartite extended amygdala-basal ganglia CRH circuit drives locomotor activation and avoidance behavior.

An adaptive stress response involves various mediators and circuits orchestrating a complex interplay of physiological, emotional, and behavioral adjustments. We identified a population of corticotropin-releasing hormone (CRH) neurons in the lateral part of the interstitial nucleus of the anterior commissure (IPACL), a subdivision of the extended amygdala, which exclusively innervate the substantia nigra (SN). Specific stimulation of this circuit elicits hyperactivation of the hypothalamic-pituitary-adrenal axis, locomotor activation, and avoidance behavior contingent on CRH receptor type 1 (CRHR1) located at axon terminals in the SN, which originate from external globus pallidus (GPe) neurons. The neuronal activity prompting the observed behavior is shaped by IPACLCRH and GPeCRHR1 neurons coalescing in the SN. These results delineate a previously unidentified tripartite CRH circuit functionally connecting extended amygdala and basal ganglia nuclei to drive locomotor activation and avoidance behavior.

MicroRNA-101 inhibits cadmium-induced angiogenesis by targeting cyclooxygenase-2 in primary human umbilical vein endothelial cells.

Exposure to toxic metal contaminants, such as cadmium compounds (Cd2+), has been shown to induce adverse effects on various organs and tissues. In particular, blood vessels are severely impacted by Cd2+ exposure, which may lead to cardiovascular diseases (CVDs). According to previous studies, CVDs are associated with increased cyclooxygenase 2 (COX-2) levels. However, the mechanisms by which CdCl2-induced COX-2 overexpression leads to cardiovascular dysfunction remain unclear. Herein, we show that the relative gene expressions of VEGF and PTGS2 (COX-2 encoding gene) are positively correlated in CVDs patients. Moreover, we demonstrate that the in vitro administration of CdCl2 induces cytotoxicity and endoplasmic reticulum (ER) stress in primary human umbilical vein endothelial cells (HUVECs). The induction of ER stress and the overexpression of COX-2 in CdCl2-treated cells alters the protein level of vascular endothelial growth factor (VEGF), resulting in abnormal angiogenesis and increased cytotoxicity. At the pre-transcription level, the inhibition of ER stress by siGRP78 (a key mediator of ER stress) can restore normal angiogenesis in the CdCl2-exposed cells. Meanwhile, at the transcription level, the adverse effects of CdCl2 exposure may be reversed via genetic modification with siRNA (siPTGS2) or by using phytochemical inhibitors (parthenolide, PN) of COX-2. Finally, at the post-transcription level, COX-2 expression may be restricted by the binding of microRNA-101 (miR-101) to the 3'-UTR of PTGS2 mRNA. The use of mimic miR-101 (mi101) to induce the expression of miR-101 eventually leads to reduced COX-2 protein levels, relieved ER stress, and less abnormal angiogenesis and cytotoxicity of CdCl2-exposed primary HUVECs. Overall, our results suggest that CdCl2-induced abnormal angiogenesis is mediated by miR-101/COX-2/VEGF-axis-dependent ER stress, and that cardiovascular dysfunction may be controlled by manipulating COX-2 at the pre-transcription, transcription, and post-transcription levels.