Genetically timed, activity-sensor and rainbow transsynaptic viral tools.

We developed retrograde, transsynaptic pseudorabies viruses (PRVs) with genetically encoded activity sensors that optically report the activity of connected neurons among spatially intermingled neurons in the brain. Next we engineered PRVs to express two differentially colored fluorescent proteins in a time-shifted manner to define a time period early after infection to investigate neural activity. Finally we used multiple-colored PRVs to differentiate and dissect the complex architecture of brain regions.

Light-activated channels targeted to ON bipolar cells restore visual function in retinal degeneration.

Genetically encoded optical neuromodulators create an opportunity for circuit-specific intervention in neurological diseases. One of the diseases most amenable to this approach is retinal degeneration, where the loss of photoreceptors leads to complete blindness. To restore photosensitivity, we genetically targeted a light-activated cation channel, channelrhodopsin-2, to second-order neurons, ON bipolar cells, of degenerated retinas in vivo in the Pde6b(rd1) (also known as rd1) mouse model. In the absence of 'classical' photoreceptors, we found that ON bipolar cells that were engineered to be photosensitive induced light-evoked spiking activity in ganglion cells. The rescue of light sensitivity was selective to the ON circuits that would naturally respond to increases in brightness. Despite degeneration of the outer retina, our intervention restored transient responses and center-surround organization of ganglion cells. The resulting signals were relayed to the visual cortex and were sufficient for the animals to successfully perform optomotor behavioral tasks.

CNN universal machine as classificaton platform: an art-like clustering algorithm.

Fast and robust classification of feature vectors is a crucial task in a number of real-time systems. A cellular neural/nonlinear network universal machine (CNN-UM) can be very efficient as a feature detector. The next step is to post-process the results for object recognition. This paper shows how a robust classification scheme based on adaptive resonance theory (ART) can be mapped to the CNN-UM. Moreover, this mapping is general enough to include different types of feed-forward neural networks. The designed analogic CNN algorithm is capable of classifying the extracted feature vectors keeping the advantages of the ART networks, such as robust, plastic and fault-tolerant behaviors. An analogic algorithm is presented for unsupervised classification with tunable sensitivity and automatic new class creation. The algorithm is extended for supervised classification. The presented binary feature vector classification is implemented on the existing standard CNN-UM chips for fast classification. The experimental evaluation shows promising performance after 100% accuracy on the training set.

Genetic reactivation of cone photoreceptors restores visual responses in retinitis pigmentosa.

Retinitis pigmentosa refers to a diverse group of hereditary diseases that lead to incurable blindness, affecting two million people worldwide. As a common pathology, rod photoreceptors die early, whereas light-insensitive, morphologically altered cone photoreceptors persist longer. It is unknown if these cones are accessible for therapeutic intervention. Here, we show that expression of archaebacterial halorhodopsin in light-insensitive cones can substitute for the native phototransduction cascade and restore light sensitivity in mouse models of retinitis pigmentosa. Resensitized photoreceptors activate all retinal cone pathways, drive sophisticated retinal circuit functions (including directional selectivity), activate cortical circuits, and mediate visually guided behaviors. Using human ex vivo retinas, we show that halorhodopsin can reactivate light-insensitive human photoreceptors. Finally, we identified blind patients with persisting, light-insensitive cones for potential halorhodopsin-based therapy.