The inner plexiform layer (IPL) is the second synaptic layer of the retina ( Fig. 23.1 ) and the final stage for processing visual information before it leaves the eye. Visual signals from rod and cone photoreceptors are first processed in the outer plexiform layer (OPL; Fig. 23.1 ), where horizontal cells modulate their signaling to bipolar cells. Bipolar cells then transmit these signals to the inner plexiform layer (IPL), where amacrine cells shape bipolar cells signaling to ganglion cells. The visual signals are first separated into separate signaling streams at the bipolar cells. Distinct classes of bipolar cells form parallel pathways that convey information about different aspects of the visual world. Synaptic connections between bipolar cells, amacrine cells and ganglion cells in the IPL produce complex processing of the visual signals. These synaptic interactions in the IPL process spatial, motion, and directional information within the visual scene. The outputs of the IPL are transmitted to distinct ganglion cell types that convey visual signals to different brain regions.
Bipolar cells form parallel pathways and provide excitatory input to the IPL
A variety of bipolar cell types form parallel signaling pathways that relay the processed photoreceptor output to the IPL. Excitatory inputs to the IPL originate from ON bipolar cells and OFF bipolar cells that depolarize in response to increasing or decreasing illumination within their receptive field centers, respectively. These distinct responses are determined by separate classes of dendritic glutamate receptors. , ON and OFF bipolar cells respond with opposite polarities because, in darkness, glutamate release from photoreceptors hyperpolarizes ON bipolar cells by activating mGluR6Rs that close cation channels and depolarizes OFF bipolar cells by activating AMPA/KA receptors that open cation channels. The axon terminals of ON and OFF bipolar cells stratify in two distinct sublaminae of the IPL, contacting the dendrites of ON and OFF ganglion cells and amacrine cells. The outer half of the IPL encodes responses to light OFF and the inner half encodes responses to light ON. , The ON and OFF sublaminae are further divided into ten bipolar cell types that have axon terminals in different morphological strata ( Fig. 23.2 ), encoding different aspects of the visual scene.
An additional major functional subdivision of both the ON and OFF sublaminae is a stratification of the processes of temporally distinct responses. Retinal neurons generate either transient or sustained responses to continuous visual stimulation that encode the temporal and spatial features of the visual scene, respectively. A transient response can signal the offset and/or onset of a light stimulus, while a sustained response can signal the duration of the light stimulus. Similar to the ON and OFF response separation, sustained and transient responsiveness is first observed at the bipolar cells. Transient and sustained ON and OFF signals recorded in bipolar cells have been attributed to variations in the types of mGluR6 and AMPA/KARs expressed on ON and OFF bipolar cell dendrites, respectively. , Transient bipolar cell terminals contact transient amacrine cell and ganglion cell processes in the mid IPL and sustained bipolar cell terminals contact the sustained amacrine cell and ganglion cell processes near the inner and outer margins of the IPL.
These functional divisions in the IPL illustrate a common organizational property of the retina. Distinct retinal neuron types have different morphological classifications. The morphology of retinal neurons determines their functional properties by constraining the IPL strata at which they receive inputs and make outputs. Because neuron morphology is so tightly correlated with neuron physiology in the retina, many studies make use of the morphological properties to help elucidate neuronal function. However, the organization of the IPL may be disrupted by degenerative photoreceptor diseases ( Box 23.1 ).
Many retinal diseases, such as macular degeneration and retinitis pigmentosa, result in the degeneration of photoreceptors. It was long thought that the inner retina was spared in photoreceptor degeneration. Recent studies , show that this is clearly not accurate. The inner nuclear layer and IPL of animal models of photoreceptor degeneration show aberrant organization and the formation of inappropriate synaptic contacts, strongly suggesting that the elimination of photoreceptors providing the sensory input to inner retina leads to the profound disorganization of the inner retina. This large-scale structural and presumably functional reorganization of the inner retina points out that caution must be exercised when using approaches to restore vision with photoreceptor transplantation or prosthetic devices.
Synaptic mechanisms shape excitatory signals in the IPL
When ON and OFF bipolar cells are depolarized by increments or decrements of illumination, respectively, they release glutamate, which excites ganglion cells and amacrine cells. Bipolar cells do not spike or fire action potentials, but instead use slow-graded depolarizations to signal. , Their transmitter release machinery is optimized for sustained glutamate release ( Fig. 23.3A ). L-type calcium channels in bipolar cell terminals mediate a sustained calcium influx because they do not desensitize to prolonged depolarization, unlike other subtypes of calcium channels found at conventional CNS synapses. Similar to photoreceptor and hair cell synapses, bipolar cells possess specialized ribbons that are located at release zones, containing large numbers of tethered, glutamate-filled vesicles that mediate sustained signaling.
The excitatory output of each bipolar cell class is shaped by both pre- and postsynaptic processes ( Fig. 23.3B ). Response termination at bipolar cell to ganglion cell synapses , is attributed to the clearance of glutamate by presynaptic transporters, which limit the amplitude and time course of ganglion cell excitation. Blocking glutamate transporters, located in the IPL, increased the amplitude and duration of the excitatory signal from bipolar cells to ganglion cells. , This role of transporters is unique to the retina and some other specialized synapses in other parts of the CNS. Elsewhere in the CNS, responses are typically terminated by either the chemical degradation of transmitter or rapid diffusion of transmitter from the synapse. , Glutamate transporters may regulate excitatory responses in ganglion cells by limiting signaling attributable to glutamate spillover. Spillover signaling results from released glutamate activating postsynaptic receptors that are distant from the release sites.
Glutamate transporters may affect signaling in the IPL in several ways. Transporters may improve the temporal response properties of some classes of ganglion cells by speeding the decay of ganglion cell light responses. Another possible role of transporters in the IPL is to limit crosstalk between different sublaminae in the IPL to preserve the functional segregation of distinct layers. , The findings that glutamate transporters regulate the spillover activation of glutamate receptors suggests that a potential site for controlling IPL signaling is the regulation of glutamate transporters, which control the excitatory signal to ganglion cells.
The desensitizing AMPA receptor (AMPAR), a subtype of postsynpatic glutamate receptor on ganglion cell dendrites, also shapes ganglion cell responses to glutamate released from bipolar cells. Reducing AMPAR desensitization enhances excitatory light responses in ganglion cells, suggesting that AMPAR desensitization shapes ganglion cell responses. , Additionally, glutamate and protons that are released from bipolar cells can restrict glutamate release. Both glutamate and protons are contained in synaptic vesicles and act presynaptically upon metabotropic glutamate receptors and calcium channels, respectively, on bipolar cell terminals to reduce vesicular release.
Amacrine cells mediate inhibition in the IPL
Excitatory signaling in the IPL is also controlled by inhibitory amacrine cell inputs. Amacrine cells receive excitation from bipolar cells and mediate inhibition in the IPL by releasing either GABA or glycine onto ganglion cell dendrites, bipolar cell axon terminals and other amacrine cells ( Fig. 23.4 ). Approximately 50 percent of amacrine cells are GABAergic and 50 percent are glycinergic. Ganglion cells receive mainly amacrine cell input, underscoring the importance of inhibition in shaping the retinal output. Bipolar cell terminals are inhibited by presynaptic amacrine cell inputs, limiting excitatory signaling. Amacrine cells also inhibit other amacrine cells, resulting in serial, inhibitory synaptic interactions that shape the timing and spatial sensitivity of inhibition. Using ERG recordings, some of the synaptic processing in the IPL can be assessed non-invasively ( Box 23.2 ).
ERG recordings are routinely used clinically and in the laboratory for assessing photoreceptor and bipolar cell function. Retinal function attributed to synaptic interactions in the IPL can also be assessed with ERG recordings (see Chapter 24 for details). Oscillatory potentials (OPs) that are superimposed on the ERG B-wave (attributed to ON bipolar cell responses) reflect synaptic activity in the IPL. Photopic negative responses and scotopic negative responses, ERG response components observed when cone and rod pathways, respectively, are activated, is thought to reflect ganglion cell responses to light-elicited inputs. These more subtle components of the ERG may be helpful in assessing IPL function in the healthy and diseased retina.
Presynaptic GABAergic inhibition occurs at the terminals of all classes of bipolar cells and is mediated by GABA receptors (GABARs). , The retina is unique because it possesses two types of ionotropic GABA receptors, the GABA A and GABA C receptors. Both receptor subtypes gate chloride channels, but they are composed of distinct molecular subunits, have distinct pharmacological profiles and have different biophysical properties. GABA C receptors are more sensitive to GABA than GABA A receptors and as a result have much slower activation and decay kinetics in response to GABA application compared to GABA A Rs. These distinct receptor properties shape GABA-mediated synaptic responses in bipolar cells. GABA A R-mediated light-evoked inhibitory postsynaptic currents (L-IPSCs) rise and decay rapidly and determine the rise time and peak amplitude of bipolar cell inhibition ( Fig. 23.5B ). In contrast, GABA C R-mediated L-IPSCs rise and decay slowly and determine the duration of bipolar cell inhibition ( Fig. 23.5B ). These two forms of presynaptic inhibition distinctly limit bipolar cell outputs. Slow GABA C Rs limit the extent of glutamate release and the duration of excitatory responses in amacrine and ganglion cells, while fast GABA A Rs limit the initial glutamate release and the initial postsynaptic excitatory responses, as detailed below.
In different bipolar cell classes, the L-IPSC time course depended on distinct contributions of GABA A Rs and GABA C Rs. GABAergic L-IPSCs in rod bipolar cells decay slowly because they are dominated by GABA C Rs, while in OFF cone bipolar cells L-IPSCs decayed rapidly, reflecting a larger GABA A R contribution ( Fig. 23.5C ). These different contributions influence bipolar cell outputs. Removing GABA C R-mediated presynaptic inhibition in bipolar cells enhances ON, but not OFF pathway signaling in the IPL, suggesting that presynaptic inhibition is asymmetric ( Fig. 23.6 ). Different complements of GABA A Rs and GABA C Rs appear to match the time course of inhibition with the time course of rod and cone photoreceptor input to different bipolar cells.
There are two forms of presynaptic inhibition of bipolar cell terminals by amacrine cells. Local presynaptic or feedback inhibition shapes the time course and extent of glutamate release. Lateral inhibition, mediated by long-distance presynaptic inhibitory signaling, contributes to the antagonistic receptive field surround of ganglion cells.
GABAergic feedback inhibition changes the timecourse of bipolar cell signaling
GABA receptors on bipolar cell axon terminals mediate inhibitory feedback from amacrine cells that reduces and shortens bipolar cell transmitter release. In the mouse, these distinct GABA receptors temporally tune inhibitory input to bipolar cells. Recordings of light-evoked inhibitory inputs from rod bipolar cells show that slowly responding GABA C receptors prolong the response decay ( Fig. 23.5B ) and rapidly activating and decaying GABA A receptors ( Fig. 23.5B ) determine the rise time and peak amplitude of the response. Because bipolar cell light-evoked excitatory responses are slow, GABA C receptors are thought to be better suited than GABA A receptors for mediating inhibitory feedback signals. Consistent with this, feedback inhibition to salamander bipolar cells , is more sensitive to GABA C receptor blockade than GABA A receptor blockade.
The GABA C receptor-mediated feedback inhibition of bipolar cell release also determines the nature of the excitatory signal to amacrine and ganglion cells. , When feedback to bipolar cell terminals is reduced with GABA C receptor blockers, glutamate release is increased and the NMDA receptor-mediated component of the ganglion cell light responses are enhanced, attributable to the spillover activation of NMDA receptors. , So GABA C receptor-mediated feedback can control the excitatory ganglion cell response to light by limiting the spillover activation of NMDA receptors.
Both ON and OFF bipolar cells receive presynaptic input at their axon terminals from amacrine cells, suggesting that signaling to ON and OFF ganglion cells is shaped by presynaptic inhibition. However, presynaptic inhibition differentially shapes ON and OFF pathway signaling in the IPL. When GABA C receptor-mediated presynaptic inhibition was eliminated, ON, but not OFF, ganglion cell responses were augmented, consistent with asymmetric presynaptic inhibition ( Fig. 23.6 ). Electrophysiological measurements showed that presynaptic inhibition alters the response range in ON ganglion cells by limiting glutamate release from ON, but not OFF bipolar cells.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
