No comprehensive metabolic profile of trait anxiety is to date available. To identify metabolic biosignatures for different anxiety states, we compared mice selectively inbred for ∼40 generations for high (HAB), normal (NAB) or low (LAB) anxiety-related behavior. Using a mass spectrometry-based targeted metabolomics approach, we quantified the levels of 257 unique metabolites in the cingulate cortex and plasma of HAB, NAB and LAB mice. We then pinpointed affected molecular systems in anxiety-related behavior by an in silico pathway and network prediction analysis followed by validation of in silico predicted alterations with molecular assays. We found distinct metabolic profiles for different trait anxiety states and detected metabolites with altered levels both in cingulate cortex and plasma. Metabolomics data revealed common candidate biomarkers in cingulate cortex and plasma for anxiety traits and in silico pathway analysis implicated amino acid metabolism, pyruvate metabolism, oxidative stress and apoptosis in the regulation of anxiety-related behavior. We report characteristic biosignatures for trait anxiety states and provide a network map of pathways involved in anxiety-related behavior. Pharmacological targeting of these pathways will enable a mechanism-based approach for identifying novel therapeutic targets for anxiety disorders.
The identification of molecular biosignatures for psychiatric disorders and their subsequent implementation in clinical settings will revolutionize the diagnostic and therapeutic options in the field of neuropsychiatry. Emerging systems biology approaches allow the interrogation of distinct disease/health states in a holistic and hypothesis-free manner. Data derived from omics technologies reveal alterations in pathways, systems and networks, casting light on underlying molecular disease mechanisms and potential therapeutic targets. In this regard, metabolomics captures the status of biochemical pathways at a given time point and is able to define distinct metabolic states in health and disease (Kaddurah-Daouk and Krishnan, 2009 and Suhre et al., 2011). Anxiety disorders encompass a wide spectrum of conditions, ranging from panic disorder to agoraphobia (American Psychiatric Association, 2000). Discovering molecular biosignatures for such disorders is of particular importance as they constitute the most common psychiatric disorders with up to one third of the patients not responding to existing treatments (Bystritsky, 2006 and Kessler et al., 2005) and no molecular biomarkers for diagnosis/therapeutics available (Filiou et al., 2011b).
To investigate the molecular underpinnings of trait anxiety, a mouse model of high (HAB), normal (NAB) or low (LAB) anxiety-related behavior was generated by selective inbreeding for ∼40 generations based on the % time spent on the elevated plus-maze (EPM) open arms (Kromer et al., 2005). Due to the selective enrichment of the genetic risk factors related to anxiety across generations, the HAB/NAB/LAB mouse model is a robust system for studying genetically inherited behavioral extremes of trait anxiety (Landgraf et al., 2007).
We have previously established a multi-omics biomarker discovery platform for animal models of disease (Filiou and Turck, 2012, Frank et al., 2009, Haegler et al., 2009 and Zhang et al., 2009). This platform was used to compare cingulate cortex (Filiou et al., 2011a) and plasma (Zhang et al., 2011) proteomes of HAB and LAB mice. We focused on cingulate cortex due to its involvement in the regulation of emotional behavior (Drevets and Savitz, 2008) and on plasma as the specimen of choice for translational applications. HAB mice exhibit energy metabolism and mitochondrial pathway alterations in the cingulate cortex and energy metabolism changes in the plasma compared to LAB mice (Filiou et al., 2011a and Zhang et al., 2011). To further study these differences in energy homeostasis, we analyzed cingulate cortex and plasma specimens of HAB, NAB and LAB mice with a metabolomics platform that is based on selected reaction monitoring (SRM) mass spectrometry and predominantly targets major metabolic pathways (Yuan et al., 2012). In the present study we report distinct metabolic profiles for different trait anxiety states both in cingulate cortex and plasma. We extended our analyses to predict in silico affected pathways in trait anxiety and validated selected pathway predictions with molecular methods, providing a map of affected pathways in anxiety-related behavior.
