Carl B. Watt

Neural remodeling in retinal degeneration

Mammalian retinal degenerations initiated by gene defects in rods, cones or the retinal pigmented epithelium (RPE) often trigger loss of the sensory retina, effectively leaving the neural retina deafferented. The neural retina responds to this challenge by remodeling, first by subtle changes in neuronal structure and later by large-scale reorganization. Retinal degenerations in the mammalian retina generally progress through three phases. Phase 1 initiates with expression of a primary insult, followed by phase 2 photoreceptor death that ablates the sensory retina via initial photoreceptor stress, phenotype deconstruction, irreversible stress and cell death, including bystander effects or loss of trophic support. The loss of cones heralds phase 3: a protracted period of global remodeling of the remnant neural retina. Remodeling resembles the responses of many CNS assemblies to deafferentation or trauma, and includes neuronal cell death, neuronal and glial migration, elaboration of new neurites and synapses, rewiring of retinal circuits, glial hypertrophy and the evolution of a fibrotic glial seal that isolates the remnant neural retina from the surviving RPE and choroid. In early phase 2, stressed photoreceptors sprout anomalous neurites that often reach the inner plexiform and ganglion cell layers. As death of rods and cones progresses, bipolar and horizontal cells are deafferented and retract most of their dendrites. Horizontal cells develop anomalous axonal processes and dendritic stalks that enter the inner plexiform layer. Dendrite truncation in rod bipolar cells is accompanied by revision of their macromolecular phenotype, including the loss of functioning mGluR6 transduction. After ablation of the sensory retina, Müller cells increase intermediate filament synthesis, forming a dense fibrotic layer in the remnant subretinal space. This layer invests the remnant retina and seals it from access via the choroidal route. Evidence of bipolar cell death begins in phase 1 or 2 in some animal models, but depletion of all neuronal classes is evident in phase 3. As remodeling progresses over months and years, more neurons are lost and patches of the ganglion cell layer can become depleted. Some survivor neurons of all classes elaborate new neurites, many of which form fascicles that travel hundreds of microns through the retina, often beneath the distal glial seal. These and other processes form new synaptic microneuromas in the remnant inner nuclear layer as well as cryptic connections throughout the retina. Remodeling activity peaks at mid-phase 3, where neuronal somas actively migrate on glial surfaces. Some amacrine and bipolar cells move into the former ganglion cell layer while other amacrine cells are everted through the inner nuclear layer to the glial seal. Remodeled retinas engage in anomalous self-signaling via rewired circuits that might not support vision even if they could be driven anew by cellular or bionic agents. We propose that survivor neurons actively seek excitation as sources of homeostatic Ca(2+) fluxes. In late phase 3, neuron loss continues and the retina becomes increasingly glial in composition. Retinal remodeling is not plasticity, but represents the invocation of mechanisms resembling developmental and CNS plasticities. Together, neuronal remodeling and the formation of the glial seal may abrogate many cellular and bionic rescue strategies. However, survivor neurons appear to be stable, healthy, active cells and given the evidence of their reactivity to deafferentation, it may be possible to influence their emergent rewiring and migration habits.

Retinal remodeling triggered by photoreceptor degenerations

Many photoreceptor degenerations initially affect rods, secondarily leading to cone death. It has long been assumed that the surviving neural retina is largely resistant to this sensory deafferentation. New evidence from fast retinal degenerations reveals that subtle plasticities in neuronal form and connectivity emerge early in disease. By screening mature natural, transgenic, and knockout retinal degeneration models with computational molecular phenotyping, we have found an extended late phase of negative remodeling that radically changes retinal structure. Three major transformations emerge: 1) Müller cell hypertrophy and elaboration of a distal glial seal between retina and the choroid/retinal pigmented epithelium; 2) apparent neuronal migration along glial surfaces to ectopic sites; and 3) rewiring through evolution of complex neurite fascicles, new synaptic foci in the remnant inner nuclear layer, and new connections throughout the retina. Although some neurons die, survivors express molecular signatures characteristic of normal bipolar, amacrine, and ganglion cells. Remodeling in human and rodent retinas is independent of the initial molecular targets of retinal degenerations, including defects in the retinal pigmented epithelium, rhodopsin, or downstream phototransduction elements. Although remodeling may constrain therapeutic intervals for molecular, cellular, or bionic rescue, it suggests that the neural retina may be more plastic than previously believed. J. Comp. Neurol. 464:1–16, 2003.

The Function of Guanylate Cyclase 1 and Guanylate Cyclase 2 in Rod and Cone Photoreceptors

Retinal guanylate cyclases 1 and 2 (GC1 and GC2) are responsible for synthesis of cyclic GMP in rods and cones, but their individual contributions to phototransduction are unknown. We report here that the deletion of both GC1 and GC2 rendered rod and cone photoreceptors nonfunctional and unstable. In the rod outer segments of GC double knock-out mice, guanylate cyclase-activating proteins 1 and 2, and cyclic GMP phosphodiesterase were undetectable, although rhodopsin and transducin α-subunit were mostly unaffected. Outer segment membranes of GC1–/– and GC double knock-out cones were destabilized and devoid of cone transducin (α- and γ-subunits), cone phosphodiesterase, and G protein-coupled receptor kinase 1, whereas cone pigments were present at reduced levels. Real time reverse transcription-PCR analyses demonstrated normal RNA transcript levels for the down-regulated proteins, indicating that down-regulation is posttranslational. We interpret these results to demonstrate an intrinsic requirement of GCs for stability and/or transport of a set of membrane-associated phototransduction proteins.

Exploring the retinal connectome

The synthesis and evaluation of 5α-reductase inhibitory activity of some 4-azasteroid-20-ones and 20-oximes and 3β-hydroxy-, 3β-acetoxy-, or epoxy-substituted C21 steroidal 20-ones and 20-oximes having double bonds in the A and/or B ring are described. Inhibitory activity of synthesized compounds was assessed using 5α-reductase enzyme and [1,2,6,7-3H]testosterone as substrate. All synthesized compounds were less active than finasteride (IC50: 1.2 nM). Three 4-azasteroid-2-oximes (compounds 4, 6 and 8) showed good inhibitory activity (IC50: 26, 10 and 11 nM) and were more active than corresponding 4-azasteroid 20-ones (compounds 3, 5 and 7). 3β-Hydroxy-, 3β-acetoxy- and 1α,2α-, 5α,6α- or 6α,7α-epoxysteroid-20-one and -20-oxime derivatives having double bonds in the A and/or B ring showed no inhibition of 5α-reductase enzyme.

A computational framework for ultrastructural mapping of neural circuitry.

Circuitry mapping of metazoan neural systems is difficult because canonical neural regions (regions containing one or more copies of all components) are large, regional borders are uncertain, neuronal diversity is high, and potential network topologies so numerous that only anatomical ground truth can resolve them. Complete mapping of a specific network requires synaptic resolution, canonical region coverage, and robust neuronal classification. Though transmission electron microscopy (TEM) remains the optimal tool for network mapping, the process of building large serial section TEM (ssTEM) image volumes is rendered difficult by the need to precisely mosaic distorted image tiles and register distorted mosaics. Moreover, most molecular neuronal class markers are poorly compatible with optimal TEM imaging. Our objective was to build a complete framework for ultrastructural circuitry mapping. This framework combines strong TEM-compliant small molecule profiling with automated image tile mosaicking, automated slice-to-slice image registration, and gigabyte-scale image browsing for volume annotation. Specifically we show how ultrathin molecular profiling datasets and their resultant classification maps can be embedded into ssTEM datasets and how scripted acquisition tools (SerialEM), mosaicking and registration (ir-tools), and large slice viewers (MosaicBuilder, Viking) can be used to manage terabyte-scale volumes. These methods enable large-scale connectivity analyses of new and legacy data. In well-posed tasks (e.g., complete network mapping in retina), terabyte-scale image volumes that previously would require decades of assembly can now be completed in months. Perhaps more importantly, the fusion of molecular profiling, image acquisition by SerialEM, ir-tools volume assembly, and data viewers/annotators also allow ssTEM to be used as a prospective tool for discovery in nonneural systems and a practical screening methodology for neurogenetics. Finally, this framework provides a mechanism for parallelization of ssTEM imaging, volume assembly, and data analysis across an international user base, enhancing the productivity of a large cohort of electron microscopists.

Retinal remodeling in human retinitis pigmentosa.

Retinitis Pigmentosa (RP) in the human is a progressive, currently irreversible neural degenerative disease usually caused by gene defects that disrupt the function or architecture of the photoreceptors. While RP can initially be a disease of photoreceptors, there is increasing evidence that the inner retina becomes progressively disorganized as the outer retina degenerates. These alterations have been extensively described in animal models, but remodeling in humans has not been as well characterized. This study, using computational molecular phenotyping (CMP) seeks to advance our understanding of the retinal remodeling process in humans. We describe cone mediated preservation of overall topology, retinal reprogramming in the earliest stages of the disease in retinal bipolar cells, and alterations in both small molecule and protein signatures of neurons and glia. Furthermore, while Müller glia appear to be some of the last cells left in the degenerate retina, they are also one of the first cell classes in the neural retina to respond to stress which may reveal mechanisms related to remodeling and cell death in other retinal cell classes. Also fundamentally important is the finding that retinal network topologies are altered. Our results suggest interventions that presume substantial preservation of the neural retina will likely fail in late stages of the disease. Even early intervention offers no guarantee that the interventions will be immune to progressive remodeling. Fundamental work in the biology and mechanisms of disease progression are needed to support vision rescue strategies.

Retinal remodeling in the Tg P347L rabbit, a large-eye model of retinal degeneration.

Retinitis pigmentosa (RP) is an inherited blinding disease characterized by progressive loss of retinal photoreceptors. There are numerous rodent models of retinal degeneration, but most are poor platforms for interventions that will translate into clinical practice. The rabbit possesses a number of desirable qualities for a model of retinal disease including a large eye and an existing and substantial knowledge base in retinal circuitry, anatomy, and ophthalmology. We have analyzed degeneration, remodeling, and reprogramming in a rabbit model of retinal degeneration, expressing a rhodopsin proline 347 to leucine transgene in a TgP347L rabbit as a powerful model to study the pathophysiology and treatment of retinal degeneration. We show that disease progression in the TgP347L rabbit closely tracks human cone‐sparing RP, including the cone‐associated preservation of bipolar cell signaling and triggering of reprogramming. The relatively fast disease progression makes the TgP347L rabbit an excellent model for gene therapy, cell biological intervention, progenitor cell transplantation, surgical interventions, and bionic prosthetic studies. J. Comp. Neurol. 519:2713–2733, 2011.

Trafficking of Membrane Proteins to Cone But Not Rod Outer Segments Is Dependent on Heterotrimeric Kinesin-II

Heterotrimeric kinesin-II is a molecular motor localized to the inner segment, connecting cilium and axoneme of mammalian photoreceptors. Our purpose was to identify the role of kinesin-II in anterograde intraflagellar transport by photoreceptor-specific deletions of kinesin family member 3A (KIF3A), its obligatory motor subunit. In cones lacking KIF3A, membrane proteins involved in phototransduction did not traffic to the outer segments resulting in complete absence of a photopic electroretinogram and progressive cone degeneration. Rod photoreceptors lacking KIF3A degenerated rapidly between 2 and 4 weeks postnatally, but the phototransduction components including rhodopsin trafficked to the outer segments during the course of degeneration. Furthermore, KIF3A deletion did not affect synaptic anterograde trafficking. The results indicate that trafficking of membrane proteins to the outer segment is dependent on kinesin-II in cone, but not rod photoreceptors, even though rods and cones share similar structures, and closely related phototransduction polypeptides.

Retinal connectomics: Towards complete, accurate networks

Connectomics is a strategy for mapping complex neural networks based on high-speed automated electron optical imaging, computational assembly of neural data volumes, web-based navigational tools to explore 10(12)-10(15) byte (terabyte to petabyte) image volumes, and annotation and markup tools to convert images into rich networks with cellular metadata. These collections of network data and associated metadata, analyzed using tools from graph theory and classification theory, can be merged with classical systems theory, giving a more completely parameterized view of how biologic information processing systems are implemented in retina and brain. Networks have two separable features: topology and connection attributes. The first findings from connectomics strongly validate the idea that the topologies of complete retinal networks are far more complex than the simple schematics that emerged from classical anatomy. In particular, connectomics has permitted an aggressive refactoring of the retinal inner plexiform layer, demonstrating that network function cannot be simply inferred from stratification; exposing the complex geometric rules for inserting different cells into a shared network; revealing unexpected bidirectional signaling pathways between mammalian rod and cone systems; documenting selective feedforward systems, novel candidate signaling architectures, new coupling motifs, and the highly complex architecture of the mammalian AII amacrine cell. This is but the beginning, as the underlying principles of connectomics are readily transferrable to non-neural cell complexes and provide new contexts for assessing intercellular communication.

Automatic mosaicking and volume assembly for high-throughput serial-section transmission electron microscopy

We describe a computationally efficient and robust, fully-automatic method for large-scale electron microscopy image registration. The proposed method is able to construct large image mosaics from thousands of smaller, overlapping tiles with unknown or uncertain positions, and to align sections from a serial section capture into a common coordinate system. The method also accounts for nonlinear deformations both in constructing sections and in aligning sections to each other. The underlying algorithms are based on the Fourier shift property which allows for a computationally efficient and robust method. We demonstrate results on two electron microscopy datasets. We also quantify the accuracy of the algorithm through a simulated image capture experiment. The publicly available software tools include the algorithms and a Graphical User Interface for easy access to the algorithms.