A very cool paper was published in JAMA yesterday that is a result of Google Research asking if machine learning and computer vision could improve retinal fundoscopic examinations of patients with diabetic retinopathy. The outcome of course is increased patient screening for physicians with limited resources.
Bruch’s membrane is a highly specialized and multi-laminar structure in our retinas that forms the basis for mediating interactions between the retinal pigment epithelium and blood flow from the choroid. I’ve not seen many good images online, so figured this image from mouse would be a good addition showing the relationship of the basal surface of the RPE with Bruch’s membrane.
This is a 58 year old white female with a retinal astrocytic hamartoma on her right optic nerve. Retinal astrocytic hamartomas are glial tumors of the retinal nerve fiber layer arising from retinal astrocytes.
This animated GIF file illustrates the height of the hamartoma and is another example of where animated gifs can be a fantastic teaching tool.
The left and right stereo images shown were taken with a Zeiss FF-4 Fundus camera by James Gilman of the Moran Eye Center.
We at Webvision would like to wish you the very best this holiday season. As in past years, we like to post an image from retinal science that is somehow evocative of the Holiday Season and this year, Gabe Luna from the Steve Fisher / Geoff Lewis laboratory delivers a stunning image of astrocytes in a retinal flat mount, but with a twist… We think you’ll be seeing more of Gabe’s beautiful imagery, but for now, here is his description of how he made this image:
“I used a GFAP-GFP mouse to identify all the astrocytes in the retina and manually (at the time it was manual) annotate their coordinates, then we used a probabilistic random-walk algorithm to go to each “cell center” and perform a segmentation result of that one astrocyte. Once all the 5,000 or so cells are segmented as a greyscale image of the individual cell, then they are assigned various hues that are spectrally distinct and the montage is re-assembled into one large image. The image there is a grossly down-sized image of the original. The original was a seamless mosaic of 412 individual z-stacks of about 15 planes at 1 micron intervals, using a 40x oil immersion lens.”
We’ve talked about jumping spiders before here on Webvision as they are an amazing animal with very well developed vision. However, their retinas and visual pathways are very different from the vertebrate retinas in that they use image defocusing for depth perception rather than parallax like humans and other vertebrates do. Figuring out spider vision has been a long standing effort by a small group of scientists and one of the problems of observing spiders is figuring out how they scan. The movie above however shows a transparent jumping spider with the pigment cells in its eyes/retinas moving while they scan an image. There is another pretty impressive movie here, showing a microscopic view into the retina of a living jumping spider.
The Paul Kayser International Award in Retina Research was created by the Directors of Retina Research Foundation and endowed by the Trustees of The Kayser Foundation to honor and perpetuate the memory of long-time friend and dedicated benefactor of RRF, Paul Kayser. Through this award both organizations are demonstrating the conviction they shared with Mr. Kayser that blindness caused by retinal disease is a global concern and must be addressed accordingly. It is thus the purpose of this award to foster greater awareness of the need for intensive study of the retina, its role in the visual process, and the retinal diseases that threaten and/or destroy eyesight by recognizing outstanding achievement and sustaining meritorious scientific investigations worldwide.
Dr. Marc was chosen as the recipient of this award for his lifetime body of work in retinal research, discovering the structure and function of the retina through novel technologies and approaches that have pushed our understanding of the retina forward.
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.
Purpose: To elucidate mechanisms underlying the dendrite developmental plasticity of retinal ganglion cells, we examined the role of glutamate receptors on retinal ganglion cell dendrite elongation and filopodia elimination.
Methods: We used the JamB genetically labeled subtype of RGCs as our working model. JamB-CreER:YFP ganglion cell dendritic arbors were imaged in whole mount retina using confocal microscopy. Dendrite length, area, branching, and filopodia number were traced and measured using Neurolucida. Visual inputs were blocked by dark-rearing pups after P5. Glutamatergic activity was blocked using daily intraocular injections of AP5 and CNQX from P9 to P13 or genetic ablation of the NMDA receptor in these RGCs.
Results: To test the role of visual inputs on dendrite development, we dark-reared mice from P5 to P30 and found a modest effect on filopodia elimination in JamB RGCs. Anticipating that spontaneous glutamatergic activity in the retina may also contribute to RGC filopodia elimination, we blocked spontaneous glutamatergic activity by daily intraocular injections of AP5 and CNQX from P9 to P13. This led to an increase in filopodia density due to decreased dendrite length but no change in filopodia number. We confirmed this result by examining NMDAR knockout JamB cells (JamB-CreER:YFP:Grin1-/-). As expected, Grin1-/- JamB RGCs have decreased dendrite outgrowth like the pharmacologic blockade. However, filopodia elimination in these cells was significantly decreased as well, suggesting that NMDA and non-NMDA glutamate receptors might regulate the RGC dendritic development in a differential manner. This effect was dramatic at P13. To test if this effect persists into adulthood, we examined Grin1-/- JamB RGCs at P30 and found that they are indistinguishable from wild-type JamB RGCs, suggesting that a compensatory mechanism exists to drive dendrite elongation and filopodia elimination in the absence of the NMDA receptor.
Conclusions: Our study demonstrated that ganglion cell dendrite outgrowth and pruning of filopodia require glutamatergic activity and visual input that act via NMDA and possibly non-NMDA glutamate receptors.
Purpose: Arf-like protein 3 (Arl3) localizes predominantly in the photoreceptor inner segment. Germline Arl3 knockout mice do not survive beyond PN 21 and display multiple organ ciliary defects as well as retinal regeneration (Schrick et al., (2006). Am. J. Pathol. 168, 1288-1298). We therefore generated rod-specific Arl3 knockouts to elucidate the role of Arl3 in transport of photoreceptor membrane-associated proteins.
Methods: Knockouts containing a gene trap in intron 1 of the Arl3 gene were generated using a EUCOMM cell line. Breeding with Flp mice, followed by mating with iCre75+ mice, generated rod-specific knockouts. Photoreceptor function and retina morphology of wild-type (WT) and mutant mice were analyzed by confocal microscopy, ERG and immunohistochemistry. An Arl3-specific polyclonal antibody (Ab) was generated using a full-length recombinant Arl3 polypeptide expressed in bacteria.
Results: Western blot of WT retina with anti-Arl3-Ab identified a 20 kDa protein, which was significantly reduced in two month-old mutant (Arl3flox/flox;iCre75+) retina. Immunohistochemistry revealed Arl3 localization predominantly in the inner segments of WT photoreceptor cells. Arl3 immunoreactivity was absent in homozygous rod knockouts, but still present in cones and the inner retina. Scotopic and photopic ERGs of rod knockout and WT mice at PN15 had comparable amplitudes suggesting normal phototransduction. Retina histology of PN15 knockout mice was comparable to WT. One month-old Arl3flox/flox;iCre75+ mice showed reduced (80-90%) scotopic, but normal photopic ERG responses. In retinas of two month-old knockout mice, scotopic ERGs were extinguished, whereas cone ERGs were highly attenuated. Retinas of one month-old homozygous knockout mice had 4-5 rows of nuclei in the ONL, and only one row in two month-old mice. Immunohistochemistry of PN 15 and one month-old retina sections revealed that rhodopsin transport, as shown by rho1D4 labeling of ROS, is normal. Rhodopsin was undetectable in two month-old conditional knockout mice due to complete photoreceptor degeneration.
Conclusions: Rod-specific knockout of Arl3 revealed rapidly progressing photoreceptor degeneration, with knockout mice being completely blind at two months of age. Outer segment development appeared to be unimpaired by Arl3 deletion and rod photoreceptor function was normal at P14.
“The primary abnormality involves the choroidal circulation, and the characteristic lesion is an inner choroidal vascular network of vessels ending in an aneurysmal bulge or outward projection, visible clinically as a reddish orange, spheroid, polyp-like structure…. The natural course of the disease often follows a remitting-relapsing course, and clinically, it is associated with chronic, multiple, recurrent serosanguineous detachments of the retinal pigment epithelium and neurosensory retina with long-term preservation of good vision.”
Optic nerve head drusen (ONHD) or optic disc drusen (ODD) occurs rarely in the population, about 1% of the population, though there seems to be a genetic association as in families with a history of ONHD, it increases to almost 3.5% of those families.
In todays Grand Rounds on Webvision, we have a classic case of ONHD with typical fundus photographs, but also red-free, autofluorescence, IR and OCT captured by James Gilman of the Moran Eye Center.
A color fundus photograph (CF) shows discreet multiple yellowish calcium deposits in the optic nerve head. The Red-Free photograph (RF) reveals clearer outlines of the drusen. The Fundus AutoFluorescence photograph (FAF) shows the highly fluorescent drusen in this patient’s optic nerve. The infrared image (IR) shows small discreet reflective bodies in the optic nerve. The OCT image (OCT) shows very dense round inclusions in the optic nerve that shadow the OCT signal and indicate the shallow depth and geographic cluster of the drusen.
This image of ganglion cells, Müller cells and starburst amacrine cells in the human retina is from a patient suffering from retinitis pigmentosa (RP). This disease this patient suffered from slowly causes people affected with this disease to go blind and is a constant reminder to me of why we engage in our research.
For some, this is a pretty, though abstract image created through a set of technologies called computational molecular phenotyping (CMP). The colors in this image come from antibodies labeling taurine, glutamine and glutamate, all small molecular species that reveal metabolic states in these tissues.
For us, these images reveal variation in cell types as well as abnormalities in other kinds of cells that presage retinal stress and the cellular responses that alter the retina in ways that both cause blindness and make it difficult to rescue vision loss. We also see the beginnings of changes in the circuitry of the retina that forever will alter the way that diseased retinas process information.
Image courtesy of Bryan William Jones, Ph.D. and originally appeared here.