ReviewEvaluating retinal ganglion cell loss and dysfunction
Introduction
Retinal ganglion cells (RGC) project their axons along the optic nerve and are responsible for the propagation of visual stimuli to the brain. In humans, 50% of RGC decussate within the chiasm with 80% of projections terminating in the lateral geniculate nucleus (LGN). In contrast, 96–99% of RGC in the rodent retina decussate within the optic chiasm (Thuen et al., 2005) with the vast majority synapsing in the superior colliculus (SC) (Berry et al., 2008, Nadal-Nicolas et al., 2012, You et al., 2013). The importance of these cells is fast realised when they die or become dysfunctional following injury and disease. The mechanism for RGC apoptosis following damage to the optic nerve has been reviewed previously (You et al., 2013). The development of models of eye disease along with the testing of potential therapies requires accurate and reliable assessment of RGC numbers along with appropriate testing of RGC function. Techniques for calculating RGC loss include: counting using either phenotypic markers of RGC, retrograde labelling, expression of fluorophores under a RGC specific promoter, and retinal nerve fibre layer thickness (RNFL) measurements. The counting method also varies between counting in histological retinal sections, whole mounts, and in vivo (with a laser ophthalmoscope). Determining RGC function is done through electrophysiological testing, such an electroretinography (ERG) and visual evoked potentials (VEP).
Although researchers have employed all of the above techniques, the comparable accuracy and reliability, and the relationship between RGC death and RGC dysfunction are poorly understood. It is also often overlooked the large number of different RGC subtypes (Sanes and Masland, 2015) and the inevitable difference in the sensitivity of these subtypes to the various quantitative techniques. This is particularly important when considering that different RGC subtypes are more resilient to optic nerve damage (Muller et al., 2014, Duan et al., 2015, Nadal-Nicolas et al., 2015b), ocular hypertension (Filippopoulos et al., 2006, Li et al., 2006, Zhou et al., 2008), photoreceptor degeneration induced-RGC loss (Lin and Peng, 2013), more resistant to particular neuroprotective strategies (Valiente-Soriano et al., 2015) and equally, may be under or over represented in certain functional assessments. This review aims to provide a discussion on the above techniques, their strengths and weaknesses and accurate quantitative comparisons between RGC counting and RGC functional assessments. For quantifying RGC axonal loss and dysfunction in their axonal transport properties, see recent review (Nuschke et al., 2015).
Section snippets
Quantification of RGC – phenotypic markers and tracers
To accurately count RGC, good phenotypic markers are required that leave no doubt to the identity of the cell in question. The ganglion cell layer (GCL) is occupied not only by RGC but also astrocytes and displaced amacrine cells, potentially leaving RGC at only 40–50% of the total GCL population (Bunt and Lund, 1974, Perry, 1981, Schlamp et al., 2013, Nadal-Nicolas et al., 2015b). Previously used markers such as βIII tubulin (Chou et al., 2013, Leibinger et al., 2013, Jiang et al., 2015) and
RNFL assessment by optical coherence tomography (OCT)
OCT can be used to produce a cross sectional image of the retina, from which the thickness of various retinal layers can be measured. RNFL, which is comprised of the RGC axons, is an easily distinguishable layer whose thickness can be measured (Fig. 2). Although this technique does not measure RGC numbers directly, axonal loss precedes RGC loss (Buckingham et al., 2008, Soto et al., 2011) and so it is predicted to be a reliable surrogate marker of RGC numbers due to the approximate 1:1
Counting techniques
Whereas accuracy relies on the marker used, reliability is more dependent on the technique the investigator uses to count the cells. While power calculations are informative of the number of animals required (Mead et al., 2014), the ideal sampling technique would be one that is neither time nor equipment intensive and represents the total number of RGC.
Functional assessments
As opposed to quantifying the number of living RGC, electrophysiological techniques are also employed to give a readout of the function of the retina and in particular, the function of the remaining RGC. This has been demonstrated to not be a simple 1:1 ratio with RGC surviving injury without functional restitution. Equally, in many instances, dysfunction is an early marker and preceding factor to RGC death and thus, neuroprotective treatments may merely prevent a “sick” cell from dying but
Behavioral testing- optokinetic reflex
The use of behavioral testing and in particular, the optokinetic reflex (OKR) to assess retinal function has been recently reviewed (Grillo and Koulen, 2015). The OKR allows the eyes to follow an object in motion while maintaining a stabilized image. Detection of the velocity of the moving object in reference to the eye tracking velocity is reliant on a subtype of RGC known as direction-selective RGC (DS-RGC) as well as their inputs into the accessory optic system (Pinto and Enroth-Cugell, 2000
Conclusions
The determination of RGC function and numbers is important for research into retinal disease and a multitude of techniques can be employed. Disadvantages however must be carefully considered along with the understanding that retinal function and RGC numbers are not always correlated. Currently the most appropriate RGC marker is RBPMS, which provides an accurate measure of the entire RGC population. RGC functional assessments (pSTR, PERG) are equally important, as a surviving RGC are not
Author contributions
Ben Mead: Conception and design; Collection and/or assembly of data; Data analysis and interpretation; Manuscript writing.
Stanislav Tomarev: Manuscript writing.
Grant information
This work was funded by a grant from the National Eye Institute Intramural Research Program.
References (113)
- et al.
Tracer coupling of neurons in the rat retina inner nuclear layer labeled by Fluorogold
Brain Res.
(2005) - et al.
Anatomical and functional damage in experimental glaucoma
Curr. Opin. Pharmacol.
(2013) - et al.
ERG changes in albino and pigmented mice after optic nerve transection
Vis. Res.
(2010) - et al.
Vinblastine-induced blockage of orthograde and retrograde axonal-transport of protein in retinal ganglion-cells
Exp. Neurol.
(1974) - et al.
Subtype-specific regeneration of retinal ganglion cells following axotomy: effects of osteopontin and mTOR signaling
Neuron
(2015) - et al.
Axonal transport along retinal ganglion cells is grossly intact during reduced function post-injury
Exp. Eye Res.
(May 2016) - et al.
Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP
Neuron
(2000) - et al.
The ERG responses to light stimuli of melanopsin-expressing retinal ganglion cells that are independent of rods and cones
Neurosci. Lett.
(2010) - et al.
Fluoro-golds toxicity makes it inferior to true blue for long-term studies of dorsal-root ganglion neurons and motoneurons
Neurosci. Lett.
(1991) - et al.
Psychophysical testing in rodent models of glaucomatous optic neuropathy
Exp. Eye Res.
(2015)