Amacrine cells are interneurons in the retina. Amacrine cells operate at the inner plexiform layer (IPL), the second synaptic retinal layer where bipolar cells and ganglion cells form synapses. There are about 40 different types of amacrine cells and are classified by the width of their field of connection, which layer(s) of the stratum in the IPL they are in, and by neurotransmitter type. No single type of amacrine cell predominates; the type with most frequency is observed only 13% of total population, and the remainders are distributed among many types of cell, each making up 5% or less of the total amacrine cell population. The average diameter of dendritic field for each type varies over 34 to 400 microns, and their overall shapes alone are enough to serve as criterion of classification.bipolar cells, however, their tasks are often more specialized. Each type of amacrine cell connects with a particular type of bipolar cell, and generally has a particular type of neurotransmitter. For example, one such population, AII, 'piggybacks' rod bipolar cells onto the cone bipolar circuitry. It connects rod bipolar cell output with cone bipolar cell input, and from there the signal can travel to the respective ganglion cells.
Most are inhibitory using either GABA or glycine as neurotransmitters.
Visual Response Properties
Amacrine cells are responsible for mediating "antagonistic inputs from bipolar cells in [a] ganglion cell's surround." Thus, amacrine cells take excitatory signals coming from bipolar cells and consequently mediate inhibitory signals to a postsynaptic ganglion cell in the 'center' of its respective receptive field. In terms of visual response properties, this means that an amacrine cell in this sort of lateral pathway located within the surround of an ON-center ganglion cell's receptive field will depolarize in the presence of light on the surround of the ganglion cell's receptive field, while one located in the surround of an OFF-center ganglion cell's receptive field will hyperpolarize under the same circumstances (and depolarize in the absence of light on the surround). ON-center ganglion cells are therefore inhibited by amacrine cells in their surrounds when light shines on the surround of their receptive fields, while OFF-center ganglion cells do not receive this inhibitory input under the same circumstances, but instead receive it in the absence of light on the ganglion cell receptive field surround.
As detailed below, starburst amacrine cells (SAC) exhibit very selective visual response properties that have to do with a stimulus' direction with respect to the SAC's dendrites.
Amacrine cells are interesting biophysically in that they operate using both sodium-mediated action potentials and sodium-independent graded potential changes. This has been shown in inhibitory (i.e., GABAergic/glycinergic) amacrine cells, which make up the majority of amacrine cells, though it is unclear if this holds for all amacrine cells.
Starburst amacrine cells exhibit very strange biophysics. Distinct SAC dendrites are selectively activated by visual stimuli moving centrifugally with respect to those distinct dendrites. SAC dendrite-specific direction selectivity is thought to underlie the direction selectivity of on/off direction-selective ganglion cells, but the mechanism by which this direction selectivity is generated in SAC dendrites remains unknown.
Amacrine cells have their cell bodies located in the inner nuclear layer of the retina and have projections in the inner plexiform layer. Different subtypes of amacrine cells project differently in the inner plexiform layer, as shown in the figure to the right depicting different types of narrow-field amacrine cells.
Amacrine cells send projections from their cell bodies into the inner plexiform layer. These projections arborize differently for different subtypes of amacrine cells. Amacrine cells have these projections distributed roughly circularly in the inner plexiform layer, though some subtypes arborize asymmetrically. Most amacrine cells can be classified according to the diameter of their projection arborization: "narrow-field" cells have arbors less than 125 µm in diameter, "medium-field" cell arbors range from 125 to 400 µm in diameter, and "wide-field" cell arbors are larger than 400 µm.
Their overall shapes alone are enough to serve as criterion for the classification.
Amacrine cells are postsynaptic targets of bipolar cells; these bipolar-to-amacrine cell synapses occur in the inner plexiform layer and are thought to be excitatory. Amacrine cells have their postsynaptic targets in the inner plexiform layer as well. Amacrine cell dendrites are known to synapse onto ganglion cell neurites in the IPL, mediating "antagonistic inputs from bipolar cells in the ganglion cell's surround."  These synapses are thought to be inhibitory.; this suggests that amacrine cells serve to regulate the output of bipolar cells in a negative-feedback loop fashion, and in fact it is thought that these amacrine-to-bipolar cell synapses are inhibitory. Further, "amacrine processes are also seen to contact other amacrine processes" in the IPL. These amacrine-to-amacrine cell synapses, interestingly enough, are thought to be excitatory. Amacrine cells are also known to form "reciprocal synapses" onto the bipolar cells that synapse onto them Thus, amacrine cells form synapses onto bipolar cells, ganglion cells, and other amacrine cells, all in the inner plexiform layer.
A few types of amacrine cells are associated with their respective functions and with corresponding ganglion cells. For example, starburst amacrine cells are known to make synapses onto on/off direction-selective ganglion cells (On/Off DSGCs) , and wide-field (WF) amacrine cells, also known as polyaxonal amacrine cells, are considered to be associated with object motion sensitive ganglion cells either directly or indirectly via bipolar cells .
There exist several molecular markers for amacrine cells, including Pax6, Tcfap2b, Gad1, and GlyT1. However, no markers exclusively expressed in amacrine cells are known to exist, and there exist "far fewer molecular markers [for amacrine cells] than known morphological types" of amacrine cells.
Antibodies against choline acetyltransferase (ChAT), the acetylcholine biosynthetic enzyme, are useful in staining amacrine cells in the retina, as it has been thought for some time that amacrine cells are the only cholinergic retinal neurons. However, at least one study has shown that there may exist cholinergic ganglion cells through staining against an alternative splice variant of ChAT mRNA in rat retina.
Staining the retina against tyrosine hydroxylase (TH), a key enzyme in the dopamine biosynthetic pathway, reveals the dopaminergic amacrine cells. It should be noted, though, that (nor)adrenergic cells also contain TH, and thus in order to isolate only dopaminergic cells, staining should also be carried out against dopamine β-hydroxylase and phenylethanolamine N-methyltransferase, enzymes found in the (nor)adrenaline biosynthetic pathways, but not in the dopamine pathway.
Most amacrine cells are inhibitory and secrete only GABA or glycine, though in total, amacrine cells as a class use eight different neurotransmitters, including acetylcholine, dopamine, and several neuropeptides, such as vasoactive intestinal peptide (VIP), substance P, and somatostatin. A particular class of amacrine cells—the starburst amacrine cell—has been found to be both cholinergic and GABAergic. According to retinal neuron expert Richard Masland, it appears that every amacrine cell is GABA- or glycinergic, with those amacrine cells that secrete other neurotransmitters secreting them concurrently with GABA or glycine.
The first characterization of amacrine cells is often attributed to Santiago Ramón y Cajal. Using the Golgi method of staining neurons, he first saw these cells in the avian retina in the late 1880s, naming them "amacrine" cells ("amacrine" meaning "without axon" in Greek). Though he was the first to call them amacrine cells, he built on the earlier work of J. Müller, who had previously described "spongioblasts" in the retina that were likely the very same cells Ramón y Cajal later named "amacrine."
Open questions/status/relevance to Eyewire
Though it is more or less well-established how inhibitory amacrine cells function, it is less clear what functions non-GABAergic/glycinergic amacrine cells have in the retina. In particular, it is not well understood for which functions starburst amacrine cells require acetylcholine secretion or how starburst amacrine cells might use both GABA and acetylcholine in concert to accomplish certain fucntions. Starburst amacrine cells also exhibit very curious biophysics in that any given individual SAC dendrite is selectively activated by visual stimuli centrifugal with respect to that particular dendrite. The mechanism for this selectivity remains unknown.
Also, as stated above, no amacrine cell-exclusive molecular markers are known to exist; the discovery of such a marker would be incredibly beneficial to further amacrine cell research.
Status/relevance to Eyewire
Thus far, dozens of starburst amacrine cells have been reconstructed through Eyewire, as well as at least one AII amacrine cell.
- Tessier-Lavigne, M. "Visual Processing by the Retina." Principles of Neural Science. New York: McGraw-Hill Medical, 2000. 507-522.
- M. C. Bieda & D. R. Copenhagen (1999) Sodium Action Potentials Are Not Required for Light-Evoked Release of GABA or Glycine From Retinal Amacrine Cells J. Neurophysiol. 81 (6): 3092-3095
- Thomas Euler, Peter B. Detwiler & Winfried Denk (2002). Directionally selective calcium signals in dendrites of starburst amacrine cells Nature 418: 845-852
- M. A. MacNeil & R. H. Masland (1998) Extreme Diversity among Amacrine Cells: Implications for Function Neuron 20: 971-982
- J. E. Dowling & B. B. Boycott (1996) title=Organization of the Primate Retina: Electron Microscopy Proc. R. Soc. A 166 (1002): 80–111
- K. L. Briggman, M. Helmstaedter & W. Denk (2011)Wiring specificity in the direction-selectivity circuit of the retina Nature 471: 183-188. doi:10.1038/nature09818
- S. A. Baccus et al. (2008) A Retinal Circuit That Computes Object Motion J. Neurosci. 28 (27): 6807-6817 doi:10.1523/JNEUROSCI.4206-07.2008
- T. J. Cherry, J. M. Trimarchi, M. B. Stadler & C. L. Cepko (2009)Development and diversification of retinal amacrine interneurons at single cell resolution Proc. Natl. Acad. Sci. USA 106 (23): 9495–9500
- O. Yasuhara et al. (2003) Demonstration of Cholinergic Ganglion Cells in Rat Retina: Expression of an Alternative Splice Variant of Choline Acetyltransferase J. Neurosci. 23 (7): 2872–2881
- Marc, Robert E. "Retinal Neurotransmitters." The Visual Neurosciences. Vol. 1. Cambridge: The MIT Press, 2003. 304-319.
- Masland, R. H. Personal communication. April 6, 2012.
- D. M. O'Malley, J. H. Sandell & R. H. Masland (1992) Co-release of Acetylcholine and GABA by the Starburst Amacrine Cells J. Neurosci. 12 (4): 1394–1408
- H. Uchiyama & W. K. Stell (2005) Association amacrine cells of Ramón y Cajal: Rediscovery and reinterpretation Visual Neuroscience 22: 881—891