Impaired Light Adaptation of ON-Sustained Ganglion Cells in Early Diabetes Is Attributable to Diminished Response to Dopamine D4 Receptor Activation.

Purpose: Retinal neuronal signaling is disrupted early in diabetes, before the onset of the vascular pathologies associated with diabetic retinopathy. There is also growing evidence that retinal dopamine, a neuromodulator that mediates light adaptation, is reduced in early diabetes. Previously, we have shown that after 6 weeks of diabetes, light adaptation is impaired in ON-sustained (ON-s) ganglion cells in the mouse retina. The purpose of this study was to determine whether changes in the response to dopamine receptor activation contribute to this dysfunction.
Methods: Single-cell retinal patch-clamp recordings from the mouse retina were used to determine how activating dopamine type D4 receptors (D4Rs) changes the light-evoked and spontaneous excitatory inputs to ON-s ganglion cells, in both control and 6-week diabetic (STZ-injected) animals. Fluorescence in situ hybridization was also used to assess whether D4R expression was affected by diabetes.
Results: D4R activation decreased light-evoked and spontaneous inputs to ON-s ganglion cells in control and diabetic retinas. However, D4R activation caused a smaller reduction in light-evoked excitatory inputs to ON-s ganglion cells in diabetic retinas compared to controls. This impaired D4R signaling is not attributable to a decline in D4R expression, as there was no change in D4R mRNA density in the diabetic retinas. Conclusions: These results suggest that the cellular response to dopamine signaling is disrupted in early diabetes and may be amenable to chronic dopamine supplementation therapy.

By 2050, it is estimated that 16 million Americans will be afflicted with diabetic retinopathy, with one-fifth having vision-threatening complications. This presents a major challenge to health care in the coming decades. Current clinical interventions largely target the vascular changes that occur with the onset of diabetic retinopathy. However, it has become increasingly clear that diabetes affects the neural retina long before functional changes in the retinal vasculature can be observed. Studies using electroretinograms (ERGs) have identified measurable changes in retinal activity in human diabetic patients– who do not exhibit any clinical signs of retinopathy. Similar deficits in ERGs are detected in rodent models of early diabetes, and single-cell studies in these models have further identified electrical dysfunction at the single-cell level.– Interestingly, multifocal ERG studies have demonstrated that these changes in electrical activity develop asymmetrically across the retina and that localized sites of disrupted neural activity precede and predict the sites of future vascular pathology.,– Thus, it is highly likely that the neuronal changes in retinal activity that are occurring early on in diabetes are involved in the progression of diabetic retinopathy.

In a recent study, we showed that 6 weeks of diabetes reduces the ability of ON-sustained (ON-s) ganglion cells to adapt to increased background light. The retina's ability to adapt to background light over a wide range of lighting conditions, a process known as light adaptation, is crucial to normal visual function.– In the healthy retina, light adaptation is mediated in part by dopamine release from dopaminergic amacrine cells.,, Dopamine functions in a paracrine manner by binding to D1, D2, and D4 receptors (R) located on retinal neurons.– There is growing evidence that dopaminergic signaling is affected in early diabetes– and could contribute to the observed changes in light adaptation. However, it is unknown whether this disruption is attributable to changes in downstream targets of dopamine, diminished release of dopamine by dopaminergic amacrine cells, or both.

Here, we sought to determine whether retinal responses to dopamine are affected in early diabetes. This was examined in ON-s ganglion cells that have previously shown impaired light adaptation in diabetes. The modulation of light-evoked and spontaneous excitatory currents by a D4R agonist in control and diabetic cells was measured to identify any impairment in dopaminergic signaling. In addition, the levels of D4R mRNA in control and diabetic retinas were quantified to assess whether changes in receptor expression were responsible for dopaminergic dysfunction.

Methods

Animals

Animal protocols followed the ARVO Statement for the Use of Animals in Ophthalmic and Visual Research and were approved by the University of Arizona Institutional Animal Care and Use Committee. C57BL/6J male mice (Jackson Laboratories, Bar Harbor, ME, USA) were housed in the University of Arizona animal facility and given the National Institutes of Health–31 rodent diet food and water ad libitum. Five-week-old mice were fasted for 4 hours and injected intraperitoneally with either streptozotocin (STZ; 75 mg/kg body weight) dissolved in 0.01 M (pH 4.5) citrate buffer or citrate buffer vehicle for 3 consecutive days. Six weeks after injections, fasted (4 hours) blood glucose was measured (OneTouch UltraMini; LifeScan, Milpitas, CA, USA). STZ-injected animals with blood glucose ≤200 mg/dL and control animals with blood glucose ≥200 mg/dL were eliminated from the study. Fasting blood glucose was 135 ± 12 mg/dL (control, n = 12 mice) or 376 ± 28 mg/dL (STZ, n = 13 mice; P < 0.001), and body weights were 23.1 ± 2.0 g (control) and 20.9 ± 0.7 g (STZ).

Whole-Cell Recordings

As previously described, six weeks after injections, mice were euthanized using carbon dioxide, the eyes were enucleated, and the cornea and lens were removed to form an eyecup. The eyecup was incubated in cold extracellular solution with hyaluronidase (800 U/mL, 20 minutes) and washed with cold extracellular solution, and the retina was removed. For slice preparations, the retina was trimmed into a rectangle, mounted onto nitrocellulose filter paper (0.45-µm; Millipore, Billerica, MA, USA), transferred to a hand chopper and sliced (250 µm), rotated 90°, and mounted onto glass coverslips using vacuum grease. Whole mounts were prepared according to previous methods., The retina was cut into four equal quadrants and mounted photoreceptor side down onto a trimmed cell culture insert (1 mm height; Millipore Sigma, Burlington, MA, USA). All dissections and light response recording procedures were performed under infrared illumination to preserve the light sensitivity.

Extracellular solution was bubbled with a mixture of 95% O2–5% CO2 (pH to ∼7.4) and contained (in mM) the following: 125.00 NaCl, 2.50 KCl, 1.00 MgCl2, 1.25 NaH2PO4, 20.00 glucose, 26.00 NaHCO3, and 2.00 CaCl2. Intracellular solution in the recording pipette contained (in mM) the following: 120.00 CsOH, 120.00 gluconic acid, 1.00 MgCl2, 10.00 HEPES, 10.00 EGTA, 10.00 tetraethylammonium-Cl, 10.00 phosphocreatine-Na2, 4.00 Mg-ATP, 0.50 Na-GTP, and 0.1% sulforhodamine-B dissolved in water (pH 7.2 with CsOH). To selectively activate D4Rs, the D4R agonist PD-168077-maleate (PD, 500 nM; solubilized in DMSO Waltham, Massachusetts, USA) was diluted in extracellular solution and applied to the recording bath by a gravity-driven superfusion system (Cell Microcontrols, Norfolk, VA, USA; ∼1 mL/min). The perfusate had a final DMSO concentration of less than 0.0025%. Chemicals were purchased from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise indicated.

Responses were recorded in a control dark-adapted state, agonist was applied for 5 minutes, and then responses were recorded in the continuous presence of agonist. Retinal slices on glass coverslips or whole-mount preps were placed in a custom chamber and heated to 32° (TC-324 temperature controller with SH-27B inline heater; Warner Instruments, Hamden, CT, USA). Whole-cell voltage-clamp recordings of light-evoked (L-) and spontaneous (s) excitatory postsynaptic currents (EPSCs) were made from ON-s ganglion cells voltage clamped at –60 mV, the reversal potential for Cl− currents. Series resistance was uncompensated. Electrodes with resistances of 3 to 7 MΩ were pulled (borosilicate glass; World Precision Instruments, Sarasota, FL, USA) using a P97 Flaming/Brown puller (Sutter Instruments, Novato, CA, USA). Calculated liquid junction potentials of 20 mV (Clampex; Molecular Devices, Sunnyvale, CA, USA), were corrected before recording. Recordings were sampled at 10 kHz, filtered at 6 kHz (Bessel filter, Multi-Clamp 700B amplifier; Molecular Devices), and digitized with a Digidata 1140 and Clampex software (Molecular Devices).

During whole-cell recordings, cells were passively filled with sulforhodamine-B included in the intracellular solution. ON-s ganglion cells were targeted by their large soma size (>15 µM diameter). Similar to a previous report, confirmation of ganglion cell morphology and presence of an axon was done at the end of each recording using an Intensilight fluorescence lamp and Digitalsight camera controlled by Elements software (Nikon Instruments, Tokyo, Japan). Ganglion cells were further characterized by their light-evoked EPSC resulting from a 500-ms duration 9.5·105 photons·µm−2·s−1 flash of light. They were classified as ON-sustained if a light-evoked EPSC coincided with the onset of light, did not return to baseline until after light offset, and did not possess a distinct OFF response as well.

Light Stimuli

Full-field light stimuli were evoked with a light-emitting diode (LED; HLMP-3950, λpeak = 525 nm; Agilent, Palo Alto, CA, USA) that was calibrated with an S471 optometer (Gamma Scientific, San Diego, CA, USA) and projected through the camera port of the microscope via a 4× objective. Stimuli frequency (1/30 seconds), intensities, and duration (30 ms) were controlled by varying the current through the LED. The stimulus intensities were chosen to cover the mesopic range of light intensities (9.5, 95.0, 950.0, 9.5·103, 9.5·104, and 9.5·105 photons·µm−2·s−1). These intensities were calculated to be equivalent to 4.75, 47.50, 475.00, 4.75·103, 4.75·104, and 4.75·105 R*·rod−1·s−1, respectively.

Data Analysis and Statistics

2 to 4 L-EPSC traces for each condition were averaged using Clampfit (Molecular Devices). The peak amplitude, charge transfer (Q), time to peak, and decay to 37% of the peak (D37) were determined. For each cell, the timing of the longest duration response (from the beginning to return to baseline, typically 1-2 seconds) was measured, and these times were used to calculate Q for all responses from that cell. Time to peak and D37 were calculated as the temporal differences between response peak amplitude and stimulus onset or decline to 37% of peak amplitude, respectively. L-EPSC data from each cell were normalized to the response recorded for that cell at the maximum intensity in dark-adapted conditions. If there was no response for a given light intensity after averaging, the peak amplitude was recorded as 0, and it was excluded from analysis of response kinetics. Comparisons between experimental conditions and luminance intensities were made with two-way ANOVA tests using the Student–Newman–Keuls (SNK) method for pairwise comparisons in SigmaPlot (Systat Software, San Jose, CA, USA). If any data were shown to have a nonnormal distribution or unequal variance, tests were repeated on the log10 values (or square root values for peak amplitudes) of data.

Spontaneous events were analyzed in the baseline after a light response up to 1 second before the subsequent light stimulus. Events were identified via the MATLAB code and methodology outlined in Andor-Ardo et al. Frequency, amplitude, interevent interval (IEI), and decay τ (single exponent fit) for identified sEPSCs were calculated using custom-written MATLAB (MathWorks, Natick, MA, USA) scripts. Effects of treatment on sEPSCs were analyzed at the single cell level with Kolmogorov–Smirnov (K-S) tests. To allow for visualization of the impact of the D4R agonist across cells, amplitude, decay τ, and IEI cumulative distributions were normalized along the x-axis to the maximum value recorded for each cell. D4R agonist effects on average sEPSC parameters were normalized to the dark-adapted state for each cell and analyzed before and after agonist with paired t-tests and between different groups of cells (control versus diabetic) with unpaired t-tests. Cells were included in the analysis if they had 10 or more spontaneous events per treatment condition. Differences were considered significant when P ≤ 0.05 and data are reported as means ± 95% confidence intervals.

mRNA Imaging and Analysis

Mouse eyecups were prepared as described above. Retinas were fixed in the eyecup (24 hours, room temperature), rinsed in PBS, dissected out of the eyecup, stored in 70% EtOH at 4°C (<7 days), embedded in paraffin, sliced at 5 µm onto Fisher Superfrost slides (#12-550-17; Fisher, Waltham, Massachusetts, USA) and air-dried overnight (room temperature). Slides were baked (60°C, 1 hour), dewaxed, treated with alcohol, baked again (30 minutes, 60°C), and incubated with ER 2 (95°C, 5 minutes) and protease (1:15 dilution, 15 minutes). Slides were run on a Leica BOND automatic slide processor (Leica Biosystems, Wetzlar, Germany) using the RNAscope LS Multiplex Fluorescent Reagent Kit (#322800; ACD, USA) and the RNAscope probe 2.5LS Mm-Drd4 (#418178; ACD) following the manufacturer's instructions. Positive (Mm-Polr2a, #320888; ACD) and negative (against the bacteria-specific transcript DapB, #320878; ACD) control probes were used on control retina slices and run alongside experimental retinal slices.

Slides were imaged on a Zeiss LSM 880 inverted confocal microscope (Carl Zeiss Microscopy, Oberkochen, Germany) with a 40× objective (Plan-Apochromat 40×/1.3 oil) and a Z-stack step of 0.38 µm. Drd4 mRNA signal was detected using the 488-nm laser at 7% intensity with pinhole set at 1 Airy unit (AU) and was collected between 498 and 553 nm. DAPI nuclear staining was detected with the 405-nm laser at 0.5% intensity with pinhole at 1 AU and was collected between 410 and 480 nm. The laser detection intensity was set such that minimal signal was detected on negative control slides.

Images were analyzed with ImageJ (National Institutes of Health, Bethesda, MD, USA). Maximum intensity projections of five slices from each Z-stack were analyzed. After subtracting background, regions of interest (ROIs) were drawn around each retinal layer on the despeckled, thresholded image (threshold set at 20% of maximum intensity of Drd4 in the inner nuclear layer). Analyze Particles was used to select and measure Drd4-positive mRNA particles in each layer. The results were analyzed by a custom MATLAB program. Each retina measurement reported is from two independently run slices, which were averaged.

Results

D4R Activation Decreases the Magnitude of L-EPSCs in Control and Diabetic ON-s Ganglion Cells

To assess whether D4R modulation of ON-s ganglion cell L-EPSCs was affected after 6 weeks of diabetes, light responses were first recorded in dark-adapted ON-s ganglion cells from control mice (control, Fig. 1A) before and after application of a D4R agonist (PD, 500 nM). To adjust for any differences in connectivity or responsivity between retinas, normalized values were compared for all analyses (see Methods). D4R activation significantly decreased peak amplitude (Fig. 1B, Table) and charge transfer (Q, Fig. 1C) at all except the dimmest light intensity (SNK, P < 0.05). D4R activation also significantly delayed time to peak values (Fig. 1D), mainly due to differences at 95 photons·µm−2·s−1 (SNK, P = 0.025). No change in D37 values was recorded (Fig. 1E).

Figure 1.

D4R activation decreases the magnitude and increases the times to peak of L-EPSCs in control ON-s ganglion cells. (A) Example L-EPSC traces from the same cell at increasing light intensities before (dark blue) and after (light blue) application of the D4R agonist PD-168077 maleate (500 nM). The gold bars represent 30-ms light stimuli. (B–E) Comparison of average peak amplitude (B), Q (C), time to peak (D), and D37 (E) between dark-adapted and D4R-agonized conditions. All values are normalized on a cell-by-cell basis to the response recorded at the maximum intensity in dark-adapted conditions. Main-effects p values for 2-way ANOVAs between treatment conditions are shown. *Significantly different main effect between dark-adapted and agonist-treated states. #Significantly different pairwise comparison between dark-adapted and D4R activated states at specific light intensity. n = 13 cells from eight animals.

Table.

Statistics for Comparisons of Values from Control and Diabetic L-EPSCs Before (DA) and After (D4R) the Addition of the D4R Agonist PD

Mouse	Comparison	Measurement	P Value (Two-Way ANOVA)
Control	DA vs. D4R	Peak amplitude	<0.001*
Control	DA vs. D4R	Q	<0.001*
Control	DA vs. D4R	Time to peak	0.013*
Control	DA vs. D4R	D37	0.149
Diabetic	DA vs. D4R	Peak amplitude	0.002*
Diabetic	DA vs. D4R	Q	<0.001*
Diabetic	DA vs. D4R	Time to peak	0.743
Diabetic	DA vs. D4R	D37	0.334
Control vs. diabetic	D4R vs. D4R	Peak amplitude	0.003*
Control vs. diabetic	D4R vs. D4R	Q	0.002*
Control vs. diabetic	D4R vs. D4R	Time to peak	0.138
Control vs. diabetic	D4R vs. D4R	D37	0.328

Indicates significance.

D4R activation also modulated ON-s ganglion cells from diabetic mice (Fig. 2A). D4R activation significantly decreased peak amplitude (Fig. 2B, Table) with a pairwise difference at 950 photons·µm−2·s−1 (SNK, P < 0.05) and significantly decreased Q (Fig. 2C) with specific differences at 950 and 9.5·104 photons·µm−2·s−1 (SNK, P < 0.05). No difference was found for time to peak (Fig. 2D) or D37 (Fig. 2E) values. These findings suggest that even after 6 weeks of diabetes, activation of D4Rs significantly diminishes ON-s ganglion cell L-EPSCs and can continue to contribute to light adaptation.

Figure 2.

D4R activation decreases the magnitude but not kinetics of L-EPSCs in diabetic ON-s ganglion cells. (A) Example L-EPSC traces from the same cell at increasing light intensities before (dark red) and after (light red) application of the D4R agonist PD-168077 maleate (500 nM). The gold bars represent 30-ms light stimuli. (B–E) Comparison of average peak amplitude (B), Q (C), time to peak (D), and D37 (E) between dark-adapted and D4R-agonized conditions. All values are normalized on a cell-by-cell basis to the response recorded at the maximum intensity in dark-adapted conditions. Main-effects p values for 2-way ANOVAs between treatment conditions are shown. *Significantly different main effect between dark-adapted and agonist-treated states. #Significantly different pairwise comparison between dark-adapted and D4R activated states at specific light intensity. n = 12 cells from 10 animals.

D4R Modulation Is Reduced in Diabetic ON-s Ganglion Cells

To assess whether D4R activation was equivalent in control and diabetic ganglion cells, the values from the D4R agonist responses in Figures 1 and 2 were compared (Fig. 3). These values are normalized to the response to maximum light intensity in the dark-adapted retina, so if one group has larger values, this shows a decreased effect of the D4R agonist. Peak amplitude values were significantly higher in diabetic than control cells (Fig. 3A, Table), with specific differences at 9.5·105 photons·µm−2·s−1 and 9.5·106 photons·µm−2·s−1 (SNK P < 0.05). Q values were also significantly larger in diabetic than control cells (Fig. 3B), with a significant pairwise difference at 9.5·105 photons·µm−2·s−1 (SNK, P = 0.011). No significant difference was found between the two groups in time to peak (Fig. 3C) or D37 (Fig. 3D). These data suggest that D4R activation has a smaller effect on L-EPSCs from diabetic ON-s ganglion cells.

Figure 3.

D4R activation causes larger decreases in L-EPSCs from control ON-s ganglion cells. (A–D) Comparison of normalized L-EPSC peak amplitude (A), Q (B), time to peak (C), and D37 (D) values between control and diabetic cells after D4R activation. D4R activation reduced L-EPSC Q and peak amplitude to a greater degree in control ON-s ganglion cells than diabetic ones. p values for 2-way ANOVAs between conditions are shown. *Significantly different main effect between control and diabetic groups. #Significantly different pairwise comparison between control and diabetic groups at a specific light intensity. Values shown here are the same as the D4R-activated values from Figures 1 and 2.

Presynaptic Effects of D4R Activation on ON-s Ganglion Cell sEPSCs Remain Unperturbed After 6 Weeks of Diabetes

We previously showed that D4R activation decreases sEPSC frequency and amplitude in ON-s ganglion cells. This suggests that D4R activation reduces ON bipolar cell release, shown by frequency reduction, and possibly causes postsynaptic changes in ON-s ganglion cells, shown by amplitude reduction. To determine if the impaired D4R function seen in diabetic L-EPSCs was attributed to changes in ON-s ganglion cells or presynaptic circuits, sEPSCs were analyzed from the same control and diabetic cells from Figures 1 and 2 (see Methods). In control ON-s ganglion cells, D4R activation decreased average sEPSC amplitude, decay τ, and frequency (Figs. 4A, 4B). When the distributions of sEPSC values in individual cells were compared (Fig. 4C), D4R activation significantly shifted sEPSC distributions toward smaller amplitudes (7/8 cells; K-S, p < 0.05), shorter decay τs (6/8 cells; K-S, p < 0.05), and longer interevent intervals (6/8 cells; K-S, P < 0.05). This suggests that for most control cells, D4R activation results in presynaptic changes to ON bipolar cell release and possibly to postsynaptic changes in ON-s ganglion cells.

Figure 4.

D4R activation decreases the amplitude, decay τ, and frequency of sEPSCs in control ON-s ganglion cells. (A) Example sEPSC traces from the same cell before (dark blue, left) and after (blue, right) application of the D4R agonist PD-168077 maleate (500 nM). (B) Average amplitude (left), decay τ (middle), and frequency (right) values of sEPSCs before (dark blue) and after (light blue) D4R activation, normalized to the dark-adapted values. Average values for individual cells are plotted as white circles. Error bars indicate average values + 95% confidence interval. (C) Normalized cumulative histograms (cumulative distribution function [CDF]) for sEPSC amplitudes (left), decay τs (middle), and interevent intervals (right). Individual CDFs were averaged for dark-adapted (dark blue) and D4R-activated (light blue) conditions. For each CDF, the y-axis was normalized to the total number of events recorded for each condition (either dark adapted or D4R activated), while the x-axis was normalized to the maximum value recorded on a cell-by-cell basis. Shaded areas represent 95% confidence intervals. n = 8 cells from five animals for all panels.

For diabetic ganglion cells, D4R activation significantly decreased average sEPSC amplitude and frequency but not decay τ (Figs. 5A, 5B). When the distributions of sEPSC values in individual cells were compared (Fig. 5C), D4R activation significantly shifted sEPSC distributions toward smaller amplitudes (7/9 cells; K-S, p < 0.05), shorter decay τs (6/9 cells; K-S, P < 0.05), and longer interevent intervals (7/9 cells; K-S, P < 0.05). Thus, sEPSCs from diabetic ON-s ganglion cells responded to a D4R agonist in a similar fashion as controls.

Figure 5.

D4R activation decreases the amplitude and frequency of sEPSCs in diabetic ON-s ganglion cells. (A) Example sEPSC traces from the same cell before (dark red, left) and after (light red, right) application of the D4R agonist PD-168077 maleate (500 nM). (B) Average amplitude (left), decay τ (middle), and frequency (right) values of sEPSCs before (red) and after (light red) D4R activation, normalized to the dark-adapted values. Average values for individual cells are plotted as white circles. Error bars indicate average values + 95% confidence interval. (C) Normalized cumulative histograms for sEPSC amplitudes (left), decay τs (middle), and interevent intervals (right). Individual CDFs were averaged for dark-adapted (red) and D4R-activated (light red) conditions. For each CDF, the y-axis was normalized to the total number of events recorded for each condition, while the x-axis was normalized to the maximum value recorded on a cell-by-cell basis. Shaded areas represent 95% confidence intervals. n = 9 cells from seven animals for all panels.

To compare the magnitude of D4R-induced sEPSC changes in control and diabetic ganglion cells, average sEPSC values after D4R activation for each cell were normalized to those recorded before D4R activation. There were no significant differences in the effects of D4R activation on sEPSC amplitude or frequency (Figs. 6A, 6C) between control and diabetic ON-s ganglion cells, suggesting similar degrees of modulation. Although a difference between control and diabetic cells in decay τ modulation might have been expected since D4R activation significantly reduced decay τs in control but not diabetic ganglion cells, there was no significant difference (Fig. 6B). Overall, these results suggest that D4R activation still affects the output of ON cone bipolar cells to ON-s ganglion cells in diabetic retinas, but its postsynaptic actions on ON-s ganglion cells may or may not be impaired.

Figure 6.

D4R activation decreases the amplitude and frequency of sEPSCs in control and diabetic ON-s ganglion cells by similar degrees. Average sEPSC amplitude (left), decay τ (middle), and frequency (right) in control (light blue) and diabetic (light red) ON-s ganglion cells after D4R activation. Data were normalized to dark-adapted values on a cell-by-cell basis before averaging, and individual data points are plotted as white circles. When directly comparing the degree to which D4R activation reduces these parameters in control and diabetic cells, no significant difference was found. Error bars represent 95% confidence intervals. Control n = 8 cells from five animals and diabetic n = 9 cells from seven animals. ‡,§Significant difference between dark-adapted and D4R-treated values in control and diabetic ganglion cells, respectively. For all tests, significance level was set to α = 0.05.

D4R mRNA Expression Is Unchanged After 6 Weeks of Diabetes

Changes in the response of light-evoked excitation to a D4R agonist could potentially be attributed to a decline in D4R expression. To asses this, D4R mRNA was fluorescently labeled in control and diabetic retinal slices using the RNAscope system (Figs. 7A–D). Both the number of D4R particles/ROI and the total D4R particle area/ROI were highest in the outer nuclear layer and photoreceptor inner segments (Fig. 7), as expected from previous results.– There were no significant differences in the number of D4R particles/ROI or in the D4R particle area/ROI between control and diabetic retinas (Fig. 7E; two-way repeated measures ANOVA-F; area P = 0.724, number P = 0.798). This shows that the reduction in D4R activity is not due to reductions in D4R expression.

Figure 7.

There was no difference in D4R mRNA levels between control and diabetic retinas. (A–D) Fluorescence in situ hybridization using the RNAscope system labeling D4R mRNA (green, Drd4) and DAPI (white) in control (A, B) and diabetic (C, D) retinal slices. (E, F) Drd4 mRNA particles were quantified by the area occupied by Drd4 particles per square micron of layer area (E) and as the number of Drd4 mRNA particles in each retinal layer shown, normalized to the area of that layer (F). No differences were found between control and diabetic retinas. n = 4 retinas from four mice for control, n = 3 retinas from three mice for diabetic. GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; IS, photoreceptor inner segments; ONL, outer nuclear layer; OPL, outer plexiform layer.

Discussion

Dopamine is a key neuromodulator for retinal light adaptation. Although there is evidence that dopaminergic signaling is disrupted early in diabetes in rodents,,,, no specific mechanisms of this disruption had been identified. Here we demonstrated reduced efficacy of a D4R agonist on retinal ganglion cell signaling in diabetic mice. Our data suggested that this was due to reduced dopamine sensitivity of outer retinal neurons and potentially the ON-s ganglion cells themselves. This impaired sensitivity is likely not due to a decline in D4R mRNA expression but could be caused by changes in the cellular machinery responsible for dopamine signal transduction.

D4R Activation Reduces Light-Evoked Excitation to ON-s Ganglion Cells in Both Control and Diabetic Retinas

D4R agonist reduced the size of L-EPSCs in ON-s ganglion cells in both control and diabetic animals. In the rodent retina, D4Rs are primarily expressed by photoreceptors (Fig. 7),,, suggesting that D4Rs could reduce ganglion cell excitation by decreasing photoreceptor light sensitivity and/or decreasing photoreceptor output to bipolar cells. D4Rs on photoreceptors have been shown to reduce cAMP levels,,– phosphorylation of phosducin, and gap-junctional coupling between rod and cone photoreceptors in the mouse. All of these mechanisms have been shown to reduce photoreceptor, especially rod, sensitivity to light.– Because there is evidence that D4R mRNA expression is unchanged (Fig. 7) in diabetic retinas, it follows that D4R activation still had effects on L-EPSCs in both control and diabetic ON-s ganglion cells.

D4R-Mediated Reduction of Light-Evoked Currents Is Impaired in Diabetic Retinas

Total retinal dopamine content is decreased in rodent models of early diabetes.,,, We previously showed that light adaptation, which is in part due to retinal dopamine release, was also impaired in diabetic animals. The reduced response of light-evoked currents to D4R agonist treatment in diabetic ON-s ganglion cells found here (Fig. 3) suggests that a diminished capacity for diabetic cells to respond to dopamine could contribute to reduced light adaptation. Since there was no decrease in D4R mRNA levels, this suggests that photoreceptors from diabetic retinas are deficient in the cellular machinery necessary to respond to dopaminergic signaling. Knocking out D4Rs in mouse photoreceptors reduces the expression of adenylyl cyclase,, the enzyme responsible for cAMP production, while application of a D4R agonist in wild-type animals increases adenylyl cyclase expression. Interestingly, these results imply that the regular activation of D4Rs in photoreceptors is necessary for proper expression of the machinery that these cells require to respond to dopamine. Thus, since absolute retinal dopamine levels are decreased in early diabetes, a chronic hypoactivation of D4Rs could explain the impaired acute response to a D4R agonist.

D4R-Mediated Reduction in Tonic Glutamate Release by ON Cone Bipolar Cells Is Maintained in Diabetic Animals, but Postsynaptic Changes in ON-s Ganglion Cells May Be Impaired

In contrast to the reduced effect of D4Rs on light-evoked currents in ON-s ganglion cells from diabetic animals, there were no significant differences in D4R modulation of spontaneous current amplitudes, decay τs, and frequencies between control and diabetic groups. A decline in sEPSC frequency shows a decrease in tonic glutamate release by ON bipolar cells. This could potentially be due to a D4R-mediated reduction of photoreceptor calcium concentrations, that would cause sustained depolarization of ON bipolar cells. Sustained depolarization has been shown to cause activity-induced adaptation of bipolar cells that would decrease tonic glutamate release onto ON-s ganglion cells. Because miniature EPSCs were not isolated in this study, it is possible that the observed changes in average sEPSC amplitude and decay τ could also be explained by the decrease in event frequency. A reduced rate of vesicle release by bipolar cells could reduce the probability of coordinated vesicle release, which in turn could affect amplitude and decay τ distributions. Alternatively, D4R expression has been reported in some populations of ganglion cells.,, D2R/D4R agonists can modify potassium, calcium, and voltage-gated sodium currents in dissociated rat ganglion cells,, so direct modulation of ON-s ganglion cell glutamate currents could cause sEPSC amplitude and decay τ changes. However, since there were no differences in D4R effects on sEPSCs between diabetic and control ON-s ganglion cells, any direct D4R effects on ON-s ganglion cells are not affected by diabetes.

Conclusions

Proper activity of retinal D4Rs is important for visual contrast sensitivity,, and contrast sensitivity deficits in diabetic mice can be acutely resolved via injection of a D4R agonist. Given these results, as well as the findings reported here, it seems likely that a deficiency in dopaminergic signaling at least in part underlies reports of impaired contrast sensitivity in diabetic human populations that lack any clinical presentation of diabetic retinopathy.– In addition, there is growing evidence for D4Rs playing an important role in the circadian control of retinal metabolism,,– as well as evidence for the disruption of these retinal circadian rhythms in early diabetes.,– If this early impairment in dopaminergic signaling is the cause for dysregulated retinal metabolism, it could serve as a direct link between the visual deficits associated with early diabetes and the progression of this disease toward the more severe symptomology of diabetic retinopathy. This would support using dopamine restorative therapies as an early intervention in diabetic models and patient populations to prevent the serious retinal complications that arise upon disease progression.