Long-range dysconnectivity in frontal and midline structures is associated to psychosis in 22q11.2 deletion syndrome

Abstract

Patients affected by 22q11.2 deletion syndrome (22q11DS) present a characteristic cognitive and psychiatric profile and have a genetic predisposition to develop schizophrenia. Although brain morphological alterations have been shown in the syndrome, they do not entirely account for the complex clinical picture of the patients with 22q11DS and for their high risk of psychotic symptoms. Since Friston proposed the “disconnection hypothesis” in 1998, schizophrenia is commonly considered as a disorder of brain connectivity. In this study, we review existing evidence pointing to altered brain structural and functional connectivity in 22q11DS, with a specific focus on the role of dysconnectivity in the emergence of psychotic symptoms. We show that widespread alterations of structural and functional connectivity have been described in association with 22q11DS. Moreover, alterations involving long-range association tracts as well as midline structures, such as the corpus callosum and the cingulate gyrus, have been associated with psychotic symptoms in this population. These results suggest common mechanisms for schizophrenia in syndromic and non-syndromic populations. Future directions for investigations are also discussed.

Appendices

Appendix

BOX 1: Summary of methods used to measure connectivity in 22q11DS:

Methods for structural connectivity analysis using diffusion tensor imaging (DTI):

• Voxel-based measures Measures that describe the properties of water diffusion. In white matter, water molecules are constrained by the cell’s structures to diffuse preferentially along the axon axis. Such an environment is called anisotropic, as the probability of water diffusion is not equal in all directions (Basser et al. 1994).

  • Fractional anisotropy (FA) FA is a scalar value between 0 and 1 and reflects the degree of anisotropy in an environment (Armitage and Bastin 2000). It gives information about the microarchitecture of the white matter (Mukherjee and McKinstry 2006).

  • Axial diffusivity (AD) AD measures the amplitude of water diffusion along the main diffusion direction. Animal studies suggest that AD is a measure of axonal integrity (Budde et al. 2009; Song et al. 2003; Song et al. 2005).

  • Radial diffusivity (RD) RD measures the amplitude of water diffusion perpendicular to the main diffusion direction. Studies in mouse models suggest that increased RD reflects reduced myelination (Song et al. 2003; Song et al. 2005).

• Tractography Tractography algorithms are used to reconstruct the anatomy of the white matter bundles. With this method, virtual fibres (also called streamlines) are initiated in each point of the image and are grown in both directions, point by point, following the main diffusion direction. Thus, the pathway of the major white matter tracts can be reconstructed and the presence of a white matter bundle between two brain regions can be measured by the number of streamlines that connect them (Bammer et al. 2003; Hagmann et al. 2003).

Methods for functional connectivity analysis using resting-state fMRI:

Resting-state fMRI records spontaneous brain activity that takes place during rest, providing information about the intrinsic structure of the functional brain network through the measure of synchrony among brain regions (Biswal et al. 1995; Greicius 2008). Two methods have principally been used in studies about 22q11DS:

• Independent Component Analysis (ICA) ICA is a data-driven approach retrieving a set of networks (components) that include regions with a coherent temporal activity (Calhoun et al. 2009; Raichle 2009). These networks are highly consistent across populations and sustain different brain functions, example of such resting state networks include the saliency, auditory, central executive or default mode networks (Rosazza and Minati 2011).

• Correlation analysis Correlation analyses are based on the computation of the Pearson’s correlation coefficient between time series of predefined regions of interest (ROIs). It can be computed from one ROI to all the points on the brain surface resulting in a complete brain map representing the connectivity of a particular region. Alternatively, the correlation can be computed for a set of ROIs that can eventually cover the whole brain.

Glossary

Connectome :

Complete map of brain connections where the brain network is represented as a set of nodes that correspond to brain regions (Hagmann 2005; Sporns et al. 2005). These nodes are connected by edges that can be either structural connectivity (for example tractography measures between a pair of nodes) or functional connectivity measurements (for example correlation analysis between the nodes’ time series)

Degree :

of a node is the number of edges connecting that node

Hubs :

Highly connected nodes providing necessary shortcuts between distant network nodes (Sporns et al. 2007)

Local clustering coefficient :

Number of connections between the neighbours of a node (Watts and Strogatz 1998)

Path length :

Shortest path connecting pair of regions in the network (Sporns 2006)

Global efficiency :

Inverse of the path length (Latora and Marchiori 2001)