Monthly Archives: May 2014

A Synaptic Basis for Small World Network Design in the ON Inner Plexiform Layer of the Rabbit Retina

Leave a reply

Bipolar cells_

This abstract was presented today at the 2014 Association for Research in Vision and Opthalmology (ARVO) meetings in Orlando, Florida by J Scott Lauritzen, Noah T. Nelson, Crystal L. Sigulinsky, Nathan Sherbotie, John Hoang, Rebecca L. Pfeiffer, James R. Anderson, Carl B. Watt, Bryan W. Jones and Robert E. Marc.

Purpose: Converging evidence suggests that large- and intermediate-scale neural networks throughout the nervous system exhibit small world’ design characterized by high local clustering of connections yet short path length between neuronal modules (Watts & Strogatz 1998 Nature; Sporns et al.2004 Trends in Cog Sci). It is suspected that this organizing principle scales to local networks (Ganmor et al. 2011 J Neurosci; Sporns 2006 BioSystems) but direct observation of synapses and local network topologies mediating small world design has not been achieved in any neuronal tissue. We sought direct evidence for synaptic and topological substrates that instantiate small world network architectures in the ON inner plexiform layer (IPL) of the rabbit retina. To test this we mined ≈ 200 ON cone bipolar cells (BCs) and ≈ 500 inhibitory amacrine cell (AC) processes in the ultrastructural rabbit retinal connectome (RC1).

Methods: BC networks in RC1 were annotated with the Viking viewer and explored via graph visualization of connectivity and 3D rendering (Anderson et al. 2011 J Microscopy). Small molecule signals embedded in RC1 e.g. GABA glycine and L-glutamate combined with morphological reconstruction and connectivity analysis allow for robust cell classification. MacNeil et al. (2004 J Comp Neurol) BC classification scheme used for clarity.

Results: Homocellular BC coupling (CBb3::CBb3 CBb4::CBb4 CBb5::CBb5) and within-class BC inhibitory networks (CBb3 → AC –| CBb3 CBb4 → AC –| CBb4 CBb5 → AC –| CBb5) in each ON IPL strata form laminar-specific functional sheets with high clustering coefficients. Heterocellular BC coupling (CBb3::CBb4 CBb4::CBb5 CBb3::CBb5) and cross-class BC inhibitory networks (CBb3 → AC –| CBb4 CBb4 → AC –| CBb3 CBb4 → AC –| CBb5 CBb5 → AC –| CBb4 CBb3 → AC –| CBb5 CBb5 → AC –| CBb3) establish short synaptic path lengths across all ON IPL laminae.

Conclusions: The retina contains a greater than expected number of synaptic hubs that multiplex parallel channels presynaptic to ganglion cells. The results validate a synaptic basis (ie. direct synaptic connectivity) and local network topology for the small world architecture indicated at larger scales providing neuroanatomical plausibility of this organization for local networks and are consistent with small world design as a fundamental organizing principle of neural networks on multiple spatial scales.

Support:  NIH EY02576 (RM), NIH EY015128 (RM), NSF 0941717 (RM), NIH EY014800 Vision Core (RM), RPB CDA (BWJ), Thome AMD Grant (BWJ).

Metabolic Changes Associated With Müller Cells In A Transgenic Rabbit Model Of Retinal Degeneration

Leave a reply

Retina-RLP

This abstract was presented today at the 2014 Association for Research in Vision and Opthalmology (ARVO) meetings in Orlando, Florida by  Rebecca L. Pfeiffer, Bryan W. Jones and Robert E. Marc.

Purpose: Müller cells play a central role in retinal metabolism via the glutamate cycle. During retinal degeneration Müller cells are among the first to demonstrate changes, reflected in alterations of metabolic signatures and morphology. The timing, extent and regulation of these changes is not fully characterized. To address this issue, we evaluated Müller cell metabolic phenotypes at multiple stages of retinal remodeling.

Methods: Samples were collected post-mortem from both WT and P347L rabbits. The retinas were then divided into fragments, fixed in buffered aldehydes, and embedded in epoxy resins. Tissues were sectioned at 200nm followed by classification with computational molecular phenotyping (CMP) using an array of small and macromolecular signatures (aspartate (D), glutamate (E), glycine (G), glutamine (Q), glutathione (J), GABA (yy), taurine (T), CRALBP, Glutamine Synthetase (GS), and GFAP). Levels of amino acid or protein were quantified by selecting a region of interest either within the Müller cell population or surrounding neurons and evaluating the intensity of the signal within that region.

Results: CMP reveals overall decreases in GS levels over the course of degeneration. Of notable importance, we saw that in regions of near complete photoreceptor loss neighboring Müller cells may express independent variation in metabolic signatures of E, Q, and GS. Also observed in these Müller cells, ratios of GS:E and GS:Q are not consistent with the ratios seen in WT retina. These results are inconsistent with the current models of both E to Q metabolism and microenvironment regulation of Müller cell phenotypes.

Conclusions: These observations indicate two conclusions. First, although the degenerate state of the retina is the likely trigger inducing Müller cells to express altered metabolic signatures, the rate at which the metabolic state changes is not purely a product of the surrounding environment, but also a stochastic change within individual Müller cells. Second, although it is commonly accepted that GS is the primary enzyme which converts Q to E as part of the glutamate cycle, in degenerate retina alternative pathways may be utilized following decrease in GS.

Support: NIH EY02576 (RM), NIH EY015128 (RM), NSF 0941717 (RM), NIH EY014800 Vision Core (RM), RPB CDA (BWJ), Thome AMD Grant (BWJ).

Synapse Classification And Localization In Electron Micrographs

Leave a reply

Synapse-classification_

We have a new publication, Synapse Classification And Localization In Electron Micrographs in Pattern Recognition Letters.  Authors are: Vignesh Jagadeesh, James Anderson, Bryan W. Jones, Robert Marc, Steven Fisher and B.S. Manjunath.

Abstract:  Classification and detection of biological structures in Electron Micrographs (EM) is a relatively new large scale image analysis problem. The primary challenges are in modeling diverse visual characteristics and development of scalable techniques. In this paper we propose novel methods for synapse detection and localization, an important problem in connectomics. We first propose an attribute based descriptor for characterizing synaptic junctions. These descriptors are task specific, low dimensional and can be scaled across large image sizes. Subsequently, techniques for fast localization of these junctions are proposed. Experimental results on images acquired from a mammalian retinal tissue compare favorably with state of the art descriptors used for object detection.