Direct-Coupled Electroretinogram (DC-ERG) for Recording the Light-Evoked Electrical Responses of the Mouse Retinal Pigment Epithelium

The retinal pigment epithelium (RPE) is a monolayer of specialized cells that line the posterior segment of the eye and exert critical functions to maintain retinal homeostasis¹. The RPE supports photoreceptors by regenerating their photon-capturing visual pigment in a process called the visual cycle², by participating in the diurnal phagocytosis of shed outer segment tips³, and in the transport of nutrients and metabolic products between photoreceptors and the choriocapillaris^4,5. Abnormalities in RPE function underlie numerous human retinal diseases, such as age-related macular degeneration⁶, Leber’s congenital amaurosis^7,8 and Best vitelliform macular dystrophy⁹. As donor eye tissues are often difficult to obtain solely for research purposes, animal models with genetic modifications can provide an alternative way to study the development of retinal diseases^10,11. Additionally, the emergence and application of CRISPR cas9 technology now permits genomic introductions (knock-in) or deletions (knock-out) in a simple, one-step process surpassing limitations of prior gene targeting technologies¹². The boom in the availability of new mouse models¹³ necessitates a more efficient recording protocol to non-invasively evaluate RPE function.

Measurement of the light-evoked electrical responses of the RPE can be achieved using a direct-coupled electroretinogram (DC-ERG) technique. When used in combination with conventional ERG recordings that measure the photoreceptor (a-wave) and bipolar (b-wave) cell responses¹⁴, the DC-ERG can define how the response properties of the RPE change with retinal degeneration^15,16,17 or whether RPE dysfunction precedes photoreceptor loss. This protocol describes a method adapted from the work of Marmorstein, Peachey, and colleagues who first developed the DC-ERG technique^16,18,19,20 and improves upon the reproducibility and ease of use.

The DC-ERG recording is difficult to perform because of the long acquisition time (9 min) during which any interruption or introduction of noise can complicate the interpretation of the data. The advantage of this new method is that the baselines reach steady state within a shorter amount of time reducing the likelihood that the animal will awaken prematurely from anesthesia and is less prone to bubble formation in the capillary electrodes.

This protocol follows the animal care guidelines outlined in the animal study protocol approved by the Animal Care and Use Committee of the National Eye Institute.

1. Importing light stimulation protocols for DC-ERG

NOTE: Follow the directions below to import the light stimulation protocols for the DC-ERG into the ERG system software (Table of Materials). The protocol consists of a 0.5 min pre-stimulus interval, followed by a step of light (10 cd/m²) for 7 min, and ending with a 1.5 min post-stimulus interval. The light intensity of 10 cd/m² (1 log₁₀ cd/m²) was selected since it evokes approximately half the maximal response for all the components of the DC-ERG in WT mice^18,21. The c-wave and fast oscillation are of particular interest as the origins of these electrical responses are well characterized and can be isolated and studied further in vitro RPE models (e.g., iPSC-RPE). The application of other light intensities can extract additional information, for instance, the off response undergoes a reversal of polarity at brighter light stimuli and may show differences at the intensity at which this reversal takes place. The user is free to change the light intensity settings at their discretion.

1. Open the ERG system software.
2. Click on Database Center.
3. Click on New (provide a new database file name). Click on Save. The popup box will display: “Database created, do you want to connect to the new database file.” Click on Yes. The current database name should now reflect the new file name.
4. Click on Transfer In in the Database Control Center Window.
5. Select Supplementary File 1: LightProtocols - TRANSFER.EXP. Click on Open.
6. Click on Close (Database Control Center) upon completion of the progress bar.
7. Click on the green Start button.
8. Click on New on the Select Patient Window. Create a new patient family name describing the mouse model, enter the date of birth (DOB, mm/dd/yyyy), toggle the gender (M/F) button to the appropriate description. Click on Close to save the experimental details.
9. Click on Protocols. The dark-adapted ERG and DC-ERG protocols should now be visible.
10. Lastly, place the Long Flash.col file into the following folder C:\ERG User Files\Long Flash.col.

2. Capillary electrode preparation

1. Cut the 1.5 mm glass capillaries in half by using a ceramic tile (Table of Materials) to score the glass and break them cleanly using a table to provide physical counterforce and to stabilize the glass.
   NOTE: Blunt the cut ends as necessary.
2. Using a Bunsen burner (Table of Materials) allow the heat to make a small bend in the capillary while holding it over the flame with forceps.

3. Filling capillary electrodes

1. Connect the vacuum desiccator to the laboratory’s vacuum line through an in-line filter (Table of Materials).
2. Pour 30 mL of Hanks’ Balanced Salt Solution (HBSS) (Table of Materials) into an open 50 mL conical tube and place it (with the cap removed) into the vacuum chamber.
3. Turn on the vacuum and degas the HBSS while turning on the computer and the recording equipment.
4. After 5‒10 min turn off the vacuum and use the degassed HBSS to fill a 12 mL syringe through an attached 25 G needle. Use it to fill the bases of the electrode holders taking extra precaution not to introduce bubbles.
   1. To do this, remove the threaded screw cap and carefully slide the syringe needle through the silicone rubber gasket to reach the back wall (silver/silver chloride pellet) of the holder.
      NOTE: The silver/silver chloride pellets are advantageous over holders that use silver wire as they provide more surface area resulting in a stable low-noise baseline. However, silver/silver chloride pellets require a liquid interface free from air bubbles to achieve a good connection. Therefore, take great care not to introduce air bubbles during this process.
   2. Gradually inject HBSS to fill the entire microelectrode while slowly retracting the syringe needle. Reattach the threaded cap but do not tighten. Reinsert the syringe needle to fill the empty space within the cap with HBSS. Then fill the glass capillary while holding it horizontal to prevent solution from escaping from the other end.
   3. Hold the glass capillary from the bent end and slowly insert the opposite end through the loosened cap and then tighten the screw cap into place.
      NOTE: Glass electrodes filled with HBSS maintain lubrication of the mouse’s eye and prevent corneal dehydration that would occur with the use of standard gold loop electrodes.
5. Position the electrode holders with the capillary electrodes tilted upwards to allow bubbles to flow out. Run the vacuum for 5‒10 min to degas. Gas escaping from the glass and plastic surfaces will push HBSS out from the electrode holders.
6. Stop and then slowly release the vacuum. Refill the electrode holders and glass capillaries as described previously.
   NOTE: Bubbles tend to collect on or near the silicone rubber gasket and can also hide in the grooves of the threaded cap, therefore special attention must be paid to keep these areas bubble-free.
7. Install and secure the microelectrode holder into the custom-made T-clip/Magnetic ball joint stand (Figure 1A, inset). To make the custom stand for the microelectrode holder, modify a T-clip (5/16”-11/32” OD Tubing) #8 (Table of Materials) by removing the black polyacetal clips on one side. Have the cylinder base of the magnetic ball joints machined in half to adjust the height (Table of Materials). Secure the modified T-clips to the magnetic ball mounting screws with M3 sized nuts.
8. Place the microelectrode holder into the modified T-clip and hold it tightly in place by sliding in approximately a 1-inch tapered wooden handle made from breaking a cotton tipped cleaning stick (Table of Materials) at an angle. Use the rare earth magnet cylinder base to securely position the customized electrode holder stand onto the metal plate of the stage enabling 360° rotation on a 180° axis.

4. Test electrodes

1. Pour HBSS into a small container (e.g., the cap of the 50 mL conical tube).
2. Gently lower the fully assembled, HBSS-filled capillary microelectrodes into the cap containing HBSS to pre-equilibrate the electrodes and place the needle ground electrode (tail/rear leg) and the Ag/AgCl sintered pellet reference electrode (mouth) in the same HBSS to complete the circuit (Figure 1A).
   NOTE: Perform all subsequent steps under dim red light. Use a red-light flashlight to position the mouse and capillary electrodes. Remember to completely turn off all light sources prior to beginning the recording.
3. Select or create an appropriate identifier (family name) to describe the mouse to be tested and select the DC-ERG protocol to be performed by completing the registration according to the following order.
   1. Click on Protocols. Select DC-ERG. Click on Run. A dialog box will pop-up: “Current patient is XXX [DOB: XX/XX/XX) Is this correct for the test being performed?” Click on Yes. Then proceed to “Step 1/6.”
4. Close the doors to the faraday cage.
5. Display impedance mode by clicking on Impedance and verify that the values for the mouth reference, tail ground, and recording electrodes are acceptable (see Figure 1B).
6. Test the baseline stability (noise and drift) by clicking on Step (forward arrow) to select Step 4/6 “Long Flash No Light.”
   NOTE: The amount of drift observed when the electrodes are placed in the HBSS bath is generally less than 500 µV per 80 s once they have stabilized and is equivalent to the drift observed when the electrodes are connected to the mouse. Thus, the electrical readout of the electrodes in the HBSS bath are an important indicator of the status of the electrodes. The noise, measured as peak-to-peak, is generally ~ 10‒15% greater in the mouse than in the HBSS bath. This is probably due to the addition of motion artifacts from breathing.
7. Begin viewing the traces by clicking on Preview. The traces should be low noise with a peak-to-peak amplitude <200 µV. A slight drift (<500 µV/80 s) that gradually fades to baseline is acceptable (Figure 1C).

5. Mouse and electrode positioning

1. Keep the mice overnight in a well-ventilated light-tight box for dark adaptation.
2. Anesthetize the animals by intraperitoneal injection of ketamine (80 mg/kg) and xylazine (8 mg/kg).
3. Apply a drop of 0.5% proparacaine HCl topically to anaesthetize the cornea as well as a drop of 2.5% phenylephrine HCl and 0.5% tropicamide to dilate the pupils.
4. Trim the mouse whiskers with scissors to prevent inadvertent twitching from disturbing the glass capillary electrodes during the recording.
   NOTE: The DC-ERG stimulus protocol within the ERG system has several built-in stimulus routines that can be selected by clicking on the Step (forward arrow) or Step (backward arrow). Only Steps 1, 4, and 5 in the software are required to prepare and perform the DC-ERG recording.
5. In the ERG system software verify that the correct patient is selected. Click on the green Protocols button. Under Protocol Description select DC-ERG. Then click on Run. Verify that this is the correct test being performed by clicking on Yes.
6. Use Step 1/6 designated Red Light stimulus to turn on the red light inside the dome to help position the mouse and electrodes while observing the changes in impedance.
7. Place the mouse on a heated recording table and carefully tent the skin of the rear leg using forceps. Hold the needle electrode firmly in one hand and insert it subcutaneously into the rear leg to secure it into place.
8. Place the reference Ag/AgCl electrode (Table of Materials) inside the mouth so that the sintered pellet rests along the back cheek and is held in place behind the teeth.
   NOTE: Gold wire electrodes should not be used as the mouth reference electrode as they have different impedance characteristics and increase the peak-to-peak noise.
9. Prior to placing the capillary electrodes to the mouse’s eye, hold the electrode holder with the glass capillaries vertical, flick the electrode holder with the index finger to remove any bubbles that may have been introduced. Fill the tip with HBSS using a 25 G needle attached to a syringe and inspect to ensure that there is no air bubble trapped at the tip. Position the electrode holder stand so that the open tip of the HBSS-filled capillaries are in gentle contact with the cornea.
10. Use special precaution to avoid introducing air bubbles by holding the lubricant eye gel dispenser inverted and discard the initial drops. Place a drop of lubricant eye gel on each eye to maintain conductivity and prevent desiccation during the recording.

6. DC-ERG recording

1. Click on Step (forward arrow) to select Step 4/6 “Long Flash No Li.”
2. Click on Impedance. Use the Impedance Checking screen to examine the resistances of the left and right eyes. The impedance values for the recording electrodes at each eye are expected to be similar (~39 kΩ). The impedance values for both the ground and reference electrodes are expected to be less than 10 kΩ).
3. Click on Preview to view the traces for the left and right eye. Wait for a stable baseline to be achieved (<10 min). Click on Stop to exit the trace preview.
   NOTE: Abrupt changes in drift direction or aberrant noise in the recording will not improve with time and will require identifying the capillary electrode that requires attention. The most likely cause is a bubble introduced to the tip of the electrode. Refill the tip with HBSS. Click on Preview again to check the baseline.
4. Click on Step (forward arrow) to select Step 5/6 “Long Flash 10 cd 7 min."
5. Click Run to start the recording (Figure 1D).

7. Data export

1. Select the “Patient” (Family Name) describing the mouse recording to be exported.
2. Click on Old Tests.
3. Under Protocol Description select DC-ERG. Click on the green Load button to load the previously acquired data.
4. Click on Step (forward arrow) to advance to Step 5/6 “Long Flash 10 cd 7 min.”
5. Click on Export.
6. Provide the filename (e.g., filename.csv). A valid filename must begin with a letter, followed by letters, numbers, or underscores. Do not use special characters or hyphens. The data analysis program (DCERG_Analysis.exe) requires that table entries meet the requirements for variable names.
7. Place a checkmark next to Data Table. Next to separator, select Tab. Place checkmarks next to Options (Titles, Vertical), Include All (Steps, Chans, Results), Data Columns (Contents, Results, Sweeps), and Format (File).
8. Then click on Export (Figure 1E). This saves the *.csv file to the C:\Multifocal folder.

8. Data analysis

1. Download and install the appropriate runtime installer (Table of Materials)
2. Download and install the DCERG_Analysis.exe installer.
   NOTE: This installs the script that will perform the analysis of the DC-ERG components and creates a shortcut to run the program in the Start Menu folder.
3. Click on the shortcut created in the Start Menu > Programs folder.
4. Select the exported data file or files (*.csv) for analysis. Use Ctrl + left click on the mouse to select more than one file.
   NOTE: The executable file generates two types of plots: 1) the raw data is plotted with a best fit line indicating the measured drift; 2) the drift corrected response is plotted after being smoothed with a moving average (with a span of ~5 s). From this plot the amplitudes and time-to-peak of the DC-ERG components are identified: c-wave, fast oscillation, light peak, and off response. The data is then exported in table format to excel where each sheet corresponds to a different mouse recording. These sheets are followed by two summary sheets: (i) compiled DC-ERG amplitudes (mV); (ii) compiled DC-ERG time-to-peaks (light onset, t = 0 min).

Figure 2 is a sample dataset from miR-204 ko/ko cre/+ (conditional KO) and wild type (WT) mice. MiR-204 ko/ko cre/+ are mice with a conditional knockout of microRNA 204 in the retinal pigment epithelium. These mice are generated by crossing floxed miR-204 mice (produced by NEIGEF)²² with VMD2-CRE mice²³. MiR-204 is highly expressed in the RPE where it regulates the expression of proteins critical for epithelial function that maintain tight junction integrity (e.g., claudins), the maintenance of potassium homeostasis through the expression of Kir 7.1 potassium channels, and the expression of several visual cycle genes (e.g., LRAT, RPE65)²⁴.

Since abnormal RPE morphology was reported in several RPE-specific Cre expressing mouse lines²⁵, we monitored for normal RPE morphology in Cre expressing mice with the WT phenotype. The structural and functional abnormalities of the miR-204 ko/ko cre+ (conditional KO) mouse retina resemble the features found in miR-204 null mice¹⁵ characterized by hyper autofluorescence (lipofuscin-like deposits) and increased microglia localized to the RPE apical surface. In null mice these changes were accompanied with decreased light-evoked electrical responses of the RPE, with minimal alteration to photoreceptor responses (assessed by retinal ERG). Thus, perturbation of miR-204 expression in miR-204 ko/ko cre/+ mice is also expected to alter the electrical response of the RPE.

In the example presented, a mouse is placed on the heated platform and the electrodes are positioned appropriately prior to lowering the dome. Impedance and drift are checked as previously described using the bath solution. Representative “negative” results are shown in Figure 2A. In Figure 2A (top panel), the trace suffers from minute bubbles in the electrode that increase the peak-to-peak noise in the trace (shaded in blue). In another example (Figure 2A, lower panel), when bubbles detach from the surface of the glass and move along the length of the electrode this causes abrupt changes in the direction of the baseline drift that cannot be compensated by drift subtraction. Figure 2B shows representative “positive” recordings of WT and miR-204 ko/ko cre/+ mice where the bubbles have been eliminated using the vacuum chamber prior to assembling the microelectrodes into the electrode holder stands.

The best fit line to the initial 25 s (green) is calculated and shown in blue (Figure 2B). The drift corrected responses are replotted in Figure 2C along with the identification of the amplitudes of the DC-ERG components. Using the DC-ERG technique described in this protocol animals from both WT and miR-204 ko/ko cre/+ strains can quickly be recorded and analyzed.

The c-wave is composed of two components: a hyperpolarization of the RPE apical membrane due to increased potassium conductance in response to a decrease in potassium in the subretinal space due to photoreceptor activity and a separate contribution originating from inner retinal cells (slow P3 component – reflecting the activity of Müller cells). The fast oscillation provides information regarding the hyperpolarization of the RPE basolateral membrane²⁶, primarily due to changes in the conductance of a Cl transporter called cystic fibrosis transmembrane conductance regulator (CFTR)²⁷. The light peak is thought to originate from a change in the concentration of a photoreceptor driven substance²⁸ that through a second messenger system depolarizes the RPE’s basolateral membrane by modulating the activity of Ca²⁺ dependent Cl channels²¹. Lastly, the off-response is a complex interaction of responses that differ in polarity and vary with light intensity¹⁸.

As expected, reduced expression of K_ir 7.1 K⁺ channels greatly attenuates the c-wave²⁹ and fast oscillation as shown in the averaged responses in Figure 2D, indicating a significant impairment of the RPE’s electrical properties. A summary of the changes in the components of the DC-ERG are provided in Figure 2E. The relative amplitudes of the DC-ERG components (normalized to WT) are plotted against the relative two largest light-evoked a-wave amplitudes (1 cd·s/m²; 10 cd·s/m²) (normalized to WT) and shown in Figure 2F‒H. The reduction in the a-wave response to the brightest light intensity (10cd·s/m²) (Figure 2F‒H, Supplementary Figure S1A,B) suggests a delay in the recovery of sensitivity due to visual cycle impairment (e.g., resulting from reduced LRAT or RPE65 expression as a result of genomic knockout of miR-204^24,30).

Figure 1: The diagram highlights key steps in the DC-ERG protocol. (A) Image of the completed circuit accomplished by lowering the recording (glass capillary microelectrodes), reference, and ground electrodes into the same bath solution. This configuration enables preliminary tests to be run (prior to anesthetizing the mouse) to evaluate the characteristic impedance, noise, and drift. Inset (upper left) showing a side-view schematic of the custom microelectrode holder stand. (B) Representative image of the Impedance Checking Mode showing the appropriate values for electrode impedances. The impedance in the Left and Right eye electrodes should be comparable, within 5 KΩ of each other (e.g., Left eye: 38.7 KΩ vs. Right eye: 40.36 KΩ). The impedance of the mouth reference electrode should be less than 1 KΩ, whereas the tail electrode should be around 2.5 KΩ. (C) Representative image of the preview trace (Step 4/6) is shown. Step 4 (Long Flash No Light) is selected as no light is delivered during the preview of this step. The traces should be low noise and may have a slight drift that gradually fades with time to baseline. Once the traces have achieved a constant drift in both channels and become relatively flat, the actual recording can begin. (D) Using Step 5/6 (Long Flash 10 cd 7 min) after 0.5 min of darkness, a light step of 10 cd/m² is delivered to the mouse for 7 min followed by a return to darkness for 1.5 min. (E) Image of the export parameters used to convert the data to a *.csv file. This precise format is required to run the DC-ERG analysis software. Please click here to view a larger version of this figure.

Figure 2: Representative traces and workflow of DC-ERG analysis. Image of a negative DC-ERG result displaying excessive (A, top panel) peak-to-peak noise and (A, bottom panel) drift. (B) Images of positive DC-ERG recording results from a WT and miR-204 ko/ko cre/+ mouse. Generated plots of the raw traces showing the best fit lines (blue) to the initial 25 s (green) prior to light onset. (C) Plots of the drift corrected DC-ERG responses for the WT and miR-204 ko/ko cre/+ mice shown in B. The amplitudes of the components of the DC-ERG are indicated in the legend. (D) Averaged DC-ERG responses of 3–8 months old WT (n = 6) and miR-204 ko/ko cre/+ (n = 6) mice. The DC-ERG components are labeled on the WT trace and the light stimulation parameters are defined below. (E) Summary of DC-ERG components taken from recordings of WT and miR-204 ko/ko cre/+ mice. Bar plots represent mean, error bars indicate standard error. The relative amplitudes of the (F) c-wave, (G) fast oscillation, and (H) off response are plotted against the relative two largest light-evoked a-wave amplitudes (1 cd·s/m²; 10 cd·s/m²) (normalized to WT). Significance is indicated by asterisks: (Student’s t-test; * = p < 0.05, ** = p < 0.01, *** = p < 0.001). Please click here to view a larger version of this figure.

Supplementary Figure 1: ERG responses of WT and miR-204 ko/ko cre/+ mice. (A) Responses of WT (black) and miR-204 ko/ko cre/+ mice (magenta) to 4 ms flashes of light of increasing intensity: 0.0001 cd·s/m² (n = 5), 0.001 cd·s/m² (n = 5), 0.01 cd·s/m² (n = 3), 0.1 cd·s/m² (n = 3), 1 cd·s/m² (n = 3), 10 cd·s/m² (n = 2). (B) Averaged a-wave amplitude plotted against flash intensity. (C) Averaged b-wave amplitude plotted against flash intensity. (D) Averaged time-to-peak of a-wave responses plotted against flash intensity. (E) Averaged time-to-peak of b-wave responses plotted against flash intensity. For all plots shown error bars indicate SEM. Significance is indicated by asterisks: (Student’s t-test; * = p < 0.05). Please click here to download this figure.

Supplementary Figure 2: Example of a DC-offset in the power line that can be mitigated with the use of a voltage regulator/power conditioner. (A) In the absence of voltage regulation voltage spikes (caused by the use of equipment in an adjacent room e.g., OCT) generate a DC-offset that can interfere with the measurement of the DC-ERG components, especially the light peak. The disruptive offset is magnified on the right. (B) With the voltage regulator/power conditioner enabled the initial spike is still noticeable but the damaging DC-offset is removed. The effect of the voltage regulator/power conditioner is magnified and shown to the right. Please click here to download this figure.

Supplementary Files. Please click here to download these files.

Critical Steps

A good DC-ERG recording requires stable electrodes that are free from bubbles that create artifacts and unwanted drift as they are extremely sensitive to outgassing and temperature changes. It is essential that a stable baseline is achieved when the electrodes are placed in the HBSS bath solution before proceeding forward with the mouse recording. Small bubbles tend to collect at the base of the capillary electrode or around the silicone gasket and are difficult to see once the electrode holder is fully assembled. When few bubbles are present, lightly flicking the holder will free them for removal. If there are too many bubbles or the drift or noise cannot be removed, it is often better to disassemble the electrode and start over while carefully inspecting for bubbles at each step of the process.

Modifications and Troubleshooting

The following customizations can be made to the setup (Table of Materials) in order to improve the fidelity of the DC-ERG recordings. Low-noise cables for microelectrode holders can be used to extend the existing cables from the 32-bit amplifier to the recording table. The additional length enables the careful placement and adjustment of the electrode holder without disturbing their position once the Ganzfeld dome is closed. A voltage regulator/power conditioner can be used to eliminate in line noise and power surges generated from lights or equipment in adjacent rooms being turned on and off (Figure S2). Additionally, the tabletop Ganzfeld dome stimulator and the 32-bit amplifier can be placed inside a Faraday cage grounded to the building ground bar to shield against any additional electrical noise.

Limitations of the Method

The DC-ERG can only be recorded faithfully on dark adapted animals meaning that once the light stimulus is turned on there is little that can be done to eliminate undesirable potentials or drift. Another limitation is that the polarity of some of the components of the DC-ERG (light-peak, off-response) is subject to the light intensity used¹⁶. This means that the greatest deviations from WT may occur at intensities not inherently present at the light intensity that this protocol uses (10 cd/m²). To this point, the DC-ERG analysis software was designed assuming a negative off response (a response minimum). Brighter light intensities that result in the reversal of polarity of the off response will require the need to alter the included analysis script file.

Significance

The RPE is involved in the homeostatic maintenance of the retinal environment and plays a critical role in the pathology of several retinal diseases. This method explains in detail how to setup a DC-ERG system to record the RPE electrical response that when performed in conjunction with conventional ERG recordings provides an objective measure of outer retinal and RPE function. These measures of RPE functionality can be used to evaluate transgenic mouse lines displaying degenerative phenotypes or to test for drug-efficacy or drug-induced cytotoxicity to the RPE.