Psychiatric disorders: diagnosis to therapy

Abstract

Recent findings in a range of scientific disciplines are challenging the conventional wisdom regarding the etiology, classification, and treatment of psychiatric disorders. This Review focuses on the current state of the psychiatric diagnostic nosology and recent progress in three areas: genomics, neuroimaging, and therapeutics development. The accelerating pace of novel and unexpected findings is transforming the understanding of mental illness and represents a hopeful sign that the approaches and models that have sustained the field for the past 40 years are yielding to a flood of new data and presaging the emergence of a new and more powerful scientific paradigm.

Figure 1

A) A standard model for the translation of genetic findings into therapeutics has been successful in Mendelian disorders and rare examples of common disorders showing Mendelian inheritance: A disease-related mutation is identified and modeled, typically via a constitutive or conditional knock-out; the molecular function and relevant pathways are identified and manipulated; a potential therapeutic target is identified. B) For many common psychiatric conditions, recent findings suggest additional complexities that confound this application of this approach: gene discovery has led to the identification of extreme locus heterogeneity; genes are often affected by heterozygous mutations and in non-coding regions; high-effect coding mutations may demonstrate biological pleiotropy with multiple roles that vary across cell types, brain region and developmental periods. C. An alternative emerging model for “bottom-up” translational work in psychiatric disorders. Multiple confirmed genes/mutations can be evaluated simultaneously, currently, for example, using protein-protein interaction and/or gene expression databases; the identification of points of overlap or intersection among genes and the identification of putative networks may be used to constrain key variables in the study of model systems, and convergent molecular, cellular and/or circuit level phenotype are sought as a prelude to target and assay development.

Figure 2

Recent examples of systems-biological approaches to interpreting genetic findings in common psychiatric disorders. A) (Adapted from Fromer et al. 2014) A recent exome sequencing study of schizophrenia used curated protein-protein databases to identify molecular pathways implicated by genes carrying non-synonymous rare de novo mutations. The proteins shown are present in the synaptic compartment of excitatory neurons and correspond in to NMDA (N-methyl-D-aspartate) and AMPA (α-Amino-3- hydroxy-5-methyl-4-isoxazolepropionic acid) receptor complexes and signaling pathways. B. (Adapted from Willsey et al 2013). An alternative recent approach restricted an initial analysis of exome data in ASD to only to those genes showing multiple de novo loss-of-function mutations in the same gene, consequently having the highest probability of being true ASD risk alleles. Expression data from the Brainspan project, capturing both anatomical and temporal dimensions of human brain development, were used to evaluate these nine seed genes. Co-expression networks representing discrete spatio-temporal windows were constructed for each of the seed genes and then evaluated for additional ASD mutations in order to identify pathology-associated networks. These in turn were found to correspond to human mid fetal prefrontal cortex and to implicate cortical excitatory neurons in deep layers (Layers 5–6). The inset illustrates that the proteins constructing these networks include genes in multiple cellular compartments and with varying functions. In this analysis, the temporal-spatial convergence was prioritized over identifying overlap in a specific cellular compartment or mapping to specific signaling pathways.

Figure 3

Examples of multiple paths through which new treatments may emerge in psychiatry. The archetypal rapid-acting antidepressant, the NMDA glutamate receptor (NMDA-R) antagonist ketamine, was first identified as a treatment for depression as a consequence of a clinical observation. The ability of D-cycloserine (DCS) to augment the efficacy of behavioral therapy for anxiety disorders developed from an animal model where this drug promoted fear extinction. Deep brain stimulation treatment for depression was inspired by neuroimaging studies describing dysfunction of the subgenual prefrontal cortex in depressed patients.