We introduce a reproducible and stable optical recording method for brain slices using voltage-sensitive dye. The article describes voltage-sensitive dye staining and recording of optical signals using conventional hippocampal slice preparations.
Wide-field single photon voltage-sensitive dye (VSD) imaging of brain slice preparations is a useful tool to assess the functional connectivity in neural circuits. Due to the fractional change in the light signal, it has been difficult to use this method as a quantitative assay. This article describes special optics and slice handling systems, which render this technique stable and reliable. The present article demonstrates the slice handling, staining, and recording of the VSD-stained hippocampal slices in detail. The system maintains physiological conditions for a long time, with good staining, and prevents mechanical movements of the slice during the recordings. Moreover, it enables staining of slices with a small amount of the dye. The optics achieve high numerical aperture at low magnification, which allows recording of the VSD signal at the maximum frame rate of 10 kHz, with 100 pixel x 100-pixel spatial resolution. Due to the high frame rate and spatial resolution, this technique allows application of the post-recording filters that provide sufficient signal-to-noise ratio to assess the changes in neural circuits.
Wide-field single photon voltage-sensitive dye (VSD) imaging of bulk-stained brain slice preparations has become a useful quantitative tool to assess the dynamics of neural circuits1,2,3,4. After the analysis of the changes in optical properties due to membrane excitation5,6,7, VSD imaging was first described in the early 1970s by Cohen and others6,8,9.; it is a suitable method to monitor the brain functions in real-time as the dye directly probes the membrane potential changes (i.e., the primary signal of the neurons).
The earliest VSDs possessed the desirable characteristics to understand the brain system, such as a fast time-constant to follow the rapid kinetics of neuronal membrane potential events, and linearity with the change in membrane potential9,10,11,12,13,14,15. Similar to other imaging experiments, this technique requires a wide range of specific tunings, such as the cameras, optics, software, and slice physiology, to accomplish the desired results. Because of these technical pitfalls, the expected benefits during initial efforts did not necessarily materialize for most of the laboratories that did not specialize in this technique.
The primal cause of the technical difficulty was the low sensitivity of the VSD toward the membrane potential change when applied to bulk staining of slice preparations. The magnitude of the optical signal (i.e., the fractional change in fluorescence) is usually 10-4-10-3 of the control (F0) signal under physiological conditions. The time scale of membrane potential change in a neuron is approximately milliseconds to few hundreds of milliseconds. To measure the changes in the membrane potential of the neuron, the camera being used for the recording should be able to acquire images with high speed (10 kHz to 100 Hz). The low sensitivity of VSD and the speed needed to follow the neural signal requires a large amount of light to be collected at the camera at a high speed, with a high signal-to-noise ratio (S/N)2,16.
The optics of the recording system are also a critical element to ensure collection of sufficient light and to improve S/N. The magnification achieved by the optics is often excessively low, such as 1X to 10X, to visualize a local functional neural circuit. For example, to visualize the dynamics of the hippocampal circuit, a magnification of approximately 5 would be suitable. Such low magnification has low fluorescence efficiency; therefore, advanced optics would be beneficial for such recording.
In addition, the slice physiology is also essential. Since the imaging analysis requires the slices to be intact, careful slice handling is needed17. Furthermore, measures taken to maintain the slice viability for a longer time are important18.
The present article describes the protocol for preparation of slices, VSD staining, and measurements. The article also outlines the improvements to the VSDs, imaging device, and optics, and other additional refinements to the experimental system that have enabled this method to be used as a straightforward, powerful, and quantitative assay for visualizing the modification of the brain functions19,20,21,22,23,24,25. The technique can also be widely used for long-term potentiation in the CA1 area of hippocampal slices1. Moreover, this technique is also useful in optical recording of membrane potentials in a single nerve cell26.
All animal experiments were performed according to protocols approved by the Animal Care and Use Committee of Tokushima Bunri University. The following protocol for slice preparation is almost a standard procedure27 , but the modifications have been the protocols of staining and recording with VSD.
1. Preparation Before the day of Experiment
2. Preparation of VSD (di-4-ANEPPS) Stock Solution
3. Daily Preparation of ACSF (1 L) (Table 4)
4. Daily Preparation of the Staining VSD Ssolution
5. Preparation for Surgery
6. Surgery (Mice)
7. Staining and Rinsing of the Slices (Mice)
8. Daily Preparation of Experimental Apparatus
9. Starting a Recording Session
Figure 5 shows the representative optical signal upon electrical stimulation of the Schaffer collateral in area CA1 of a mouse hippocampal slice. The consecutive images in Figure 5A show the optical signal before any spatial and temporal filters were applied, while Figure 5B shows the same data after applying a 5 x 5 x 5 cubic filter (a Gaussian Kernel convolution, 5 x 5 spatial- and 5 to temporal-dimension) twice. Due to the high frame rate (0.1 ms/frame) and high spatial resolution (100 pixels x 100 pixels), the application of the filter did not change the signal but filtered out the noise, which can also be observed in the time course recorded in pixels in a single trial (Figure 5C, no filter; Figure 5D, with filter; and Figure 5E, superimposed).
Figure 6 compares the typical response in area CA1 of a hippocampal slice between mouse (Figure 6A) and rat (Figure 6B). As is evident in the figure, hyperpolarizing response due to inhibitory inputs is apparent in the rat hippocampal slice upon applying the same stimulus to the Schaffer collateral near the CA1/CA3 border. The small hyperpolarizing response is observed in the distal side of the CA1 after 24 ms in the mouse hippocampal slice, but more massive hyperpolarizing response overtook the depolarizing response in the rat hippocampal slice. The VSD imaging can clearly demonstrate the difference between mouse and rat hippocampal slices.
Figure 1: An illustration of the agar block used to mount brain tissue for slicing and a jig to make the block.(A) A schematic illustration of the brain block and the 4% agar block (B). (C, D) Photograph of an adjustable jig made by a Plexiglass plate (5-mm thick each) to make the agar block. When preparing the agar block, the upper and the lower plates should be stacked as shown in D. (E) By inserting a blade into the long thin slots (1) and (2), the triangular part is cut out, the slot (3) is used to trim the entire depth of the block. (4) Removal of the upper plate enables slicing out of the non-necessary parts. Using the slot 1 and 2, make only 5 mm deep cuts so that resulting the block is as shown in (A) and (B). Please click here to view a larger version of this figure.
Figure 2: An illustration of the interface type incubating chamber used to maintain slice physiology. (A) Overview of the system. (B) Interior details. (C) Illustration of the recovery chamber. Specimen should be placed on a filter paper positioned on an ACSF-filled 90 mm and 60 mm Petri dish. The latter dish is to support the filter paper. The dishes and ACSF bubbling container are kept in place with a Plexiglass plate. The filter paper should not touch the wall of the air tight box nor the container. The air is supplied through a bubbling bottle that incorporates moistened gases in the chamber. Please click here to view a larger version of this figure.
Figure 3: Hippocampal slice preparations from a mouse brain. (A) The isolated mouse brain should be cut first at the dashed line (a), then (b). Finally, the cut should be made along the line (c). The cut should be perpendicular to the bottom. (B) The brain block should be mounted on an agar block. (C) The agar block should be placed on the slicer. (D) Resulting slice. The excess tissue should be cut at the dashed line. (E) The slice should be placed at the center of the membrane filter with a Plexiglass holder (Outer diameter 15 mm, Inner diameter 11 mm, the PTFE membrane filter of 13 mm). Please click here to view a larger version of this figure.
Figure 4: Recording system for optical signals from slice preparations. (A) A photograph of the microscope used to image the slices in the current manuscript. The optics consist of an objective lens (5x NA0.60), a mirror box for dichroic filter (580 nm), and a projection (tube) lens (PLAN APO x1.0). A high-speed camera is attached on the top of the projection lens through a c-mount. There is another usual USB camera for observation. Excitation light is introduced using fiber optics. (B) Photograph of the lenses that are compatible with the mirror box. (C) Schematic diagram of the recording system. The imaging system and electrophysiological recording system are controlled by a PC. LED illumination system with a photo-diode feedback control system was used as a light source. Please click here to view a larger version of this figure.
Figure 5: Representative optical signal in area CA1 of a mouse hippocampal slice in a single trial. (A) The consecutive images show the propagation of the neuronal signal acquired at the frame rate of 0.1 ms/frame along the Schaffer collateral pathway before applying any spatial and temporal filters (every 0.2 ms). Excitation was 530 nm and emission was >590 nm. (B) The same data after an application of a three-dimensional Gaussian kernel of 5 x 5 x 5 twice. (C, D) The traces of optical signals in the representative pixels [each two pixels (36 µm) along a line in the middle of the CA1 is shown in the square in A, * denotes the pixel in the stratum pyramidale] from the data shown in A and B. (E) The superimposed signal from the optical signals in C and D. Please click here to view a larger version of this figure.
Figure 6: Comparison of optical signal between mouse and rat hippocampal slices. The representative consecutive images upon electrical stimulation of the Schaffer collateral pathway near the CA3/CA1 border of a mouse (A) and rat (B) hippocampal slice. Video 1 shows the mouse hippocampal slice and Video 2 shows the rat hippocampal slice. The time-course of the optical signal in the middle of the CA1 at the stratum pyramidale (st. pyr.) and stratum radiatum (st. rad.) is shown on the right of the figure. Please click here to view a larger version of this figure.
Final mM | Weight | |
NaH2PO4 • 2H2O | 1.25 mM | 3.90 g |
MgSO4•7H2O | 2 mM | 9.86 g |
KCl | 2.5 mM | 3.73 g |
Add H2O to make 50 mL |
Table 1: Stock A.
Final mM | Weight | |
MgSO4•7H2O | 2 mM | 12.32 g |
Add H2O to make 50 mL |
Table 2: Stock B.
Final mM | Weight | |
CaCl2•2H2O | 2 mM | 5.88 g |
Add H2O to make 50 mL |
Table 3: Stock C.
Final (mM) | weight | |
NaCl | 124 mM | 7.25 g |
NaHCO3 | 26 mM | 2.18 g |
Glucose | 10 mM | 1.8 g |
Stock A | 2.5 mL | |
Add about 950 mL of H2O and bubble with mixed gas (95 % O2/5 % CO2) (10 min) | ||
Stock C (CaCl2) | 2 mM | 2.5 mL |
Add H2O to make 1,000 mL |
Table 4: Daily preparation of ACSF.
Final (mM) | Weight | |
NaHCO3 | 26 mM | 2.18 g |
Sucrose | 205.35 mM | 70.29 g |
Glucose | 10 mM | 1.8 g |
Stock A | 2.5 mL | |
Stock B | 2.0 mL | |
Add about 900 mL of H2O and bubble with mixed gas (95 % O2/5% CO2) (10 min) | ||
Stock C | 0.4 mM | 0.5 mL |
Add H2O to make 1,000 mL |
Table 5: Modified ACSF (cutting solution).
The slice physiology is vital for collecting the right signal. The use of the ring-membrane filter system in this protocol ensures that the slice remains healthy and un-distorted throughout the procedure2,16,17. Other systems can be used to retain slice physiology during the recording, but the slice should not get deformed at any time as the imaging needs every part of the slice to be healthy. The ring-membrane filter system is also better for staining, as this helps us minimize the volume of the staining solution required. It is also important to control the intensity of excitation light, as it should be low with respect to the time-fraction. The continuous illumination can damage the specimen; therefore, appropriate use of the shutter is necessary.
The toxicity of the VSD has been often discussed29, but it is a result of non-linear multiplication of the dye concentration, excitation light intensity, and duration of exposure to the light. The staining procedure shown in this protocol did not cause any measurable changes in the physiological parameters of the slice such as the input-output relationships of the field-potential recordings and those on the long-term potentiation (LTP), paired-pulse facilitation (PPF). The deterioration of the slice physiology is sizable especially under continuous illumination16, but it can be managed by monitoring the field potential. By using these precautions, we can record LTP with continuous optical recording using VSD for more than 12 h24, which is comparable to the best conditioned in vitro experiments.
During the recording, the air table might be useful, but insulation from other devices should be allowed because of the low magnification of the optics. However, mechanical disturbances are one of the significant causes behind poor imaging. If the considerable amount of false signal in image consists of opposite signs at the edge of the object, thus the difference of the brightness, it is the most probably caused by the mechanical disturbances. The movements of the specimen and fluid are the most frequent causes of motion noise, and hence, should be avoided or minimalized.
The VSD signal (fractional change in the light intensity) is small (10-3 to 10-4 of the initial fluorescence). To detect such a small change, the fluorescence should exceed 105 to 106 photons at the detector in the fraction of time to overcome the effect of photon-shot noise. Furthermore, to follow the neuronal activity, the frame rate should be fast, close to the time constants needed to perform electrophysiological recording such as that around the kHz range. A combination of these two conditions requires the amount of fluorescence that is far more extensive than other kinds of fluorescent imaging. This requires a high numerical aperture in the whole optics, and the usual microscope is not the best option. Larger pupil and aperture are needed as shown in Figure 4.
The recoding system should match with the larger photon well depth, fast frame rate, and low noise. The choice mostly depends on the speed of neural system. The faster signal such as the hippocampal signal transduction needs specialized ultrafast, low-noise system. However, the slow signal such as the slow spread of activity in the cortices might be detected using the usual but scientific grade cameras.
The selection of the light source is also critical. The choice of the light depends on its intensity, stability, and the area of illumination. In the case of low-magnification wide-field imaging, point light source such as arcs and lasers need to expand, which makes it difficult to use these sources. Arcs, such as mercury and Xenon lamps, are the bright light source but usually are not stable. However, the recent development of Xenon light might overcome the problem. The halogen lamp is stable and has a larger area of filament that can easily match with wide-field imaging, but is limited in the strength especially at 530 nm. The recent development of power LED has enabled us to use it as the potential light source, but it must have the feedback stabilizer because of the temperature dependency. Lasers can be used but the high coherency results in a speckled noise, which is usually unacceptable for wide-field imaging.
The VSD imaging protocol presented in this article measures a value relative to the resting condition. Absolute measurement of the membrane potential cannot be performed using the current technique. Ratiometric imaging and fluorescence lifetime measurements can be used to assess the absolute membrane potentials.
The imaging of brain slices bulk-stained with VSD at low magnification can demonstrate the sub-threshold membrane potential changes in the micro-circuitry interactions of the brain. Such functional scope regarding the connection between the micro-circuitry at real-time resolution will be useful in many areas of brain research, especially to analyze the pathological aspects most likely caused by such excitatory and inhibitory functional connections between different brain areas. This application will be critical to investigate the changes in neural circuits related to certain types of neuropsychiatric diseases30,31,32.
The development of the genetically encoded voltage indicator33,34 is the future direction for optical membrane potential recordings that will pave the way for the attractive applications of cell type-specific analysis of neural circuit-level events.
There is much room for improvement in the optics, especially for visualizing the wide-field functional connections. Our novel confocal optics35 will enable high-speed and high-S/N ratio recording of the VSD signal.
The authors have nothing to disclose.
TT received the JSPS KAKENHI Grant (JP16H06532, JP16K21743, JP16H06524, JP16K0038, and JP15K00413) from MEXT and grants from the Ministry of Health, Labour and Welfare (MHLW-kagaku-ippan-H27 [15570760] and H30 [18062156]). We would like to thank Editage (www.editage.jp) for English language editing.
High speed image acquisition system | Brainvision co. Ltd. | MiCAM – Ultima | Imaging system |
High speed image acquisition system | Brainvision co. Ltd. | MiCAM 02 | Imaging system |
Macroscepe for wide field imaging | Brainvision co. Ltd. | THT macroscope | macroscope |
High powere LED illumination system with photo-diodode stablilizer | Brainvision co. Ltd. | LEX-2G | LED illumination |
Image acquisition software | Brainvision co. Ltd. | BV-ana | image acquisition software |
Multifunctional electric stimulator | Brainvision co. Ltd. | ESTM-8 | Stimulus isolator+AD/DA converter |
Slicer | Leica | VT-1200S | slicer |
Slicer | Leica | VT-1000 | slicer |
Blade for slicer | Feather Safety Razor Co., Ltd. | #99027 | carbon steel razor blade |
Membrane filter for slice support | Merk Millipore Ltd., MA, USA | Omnipore, JHWP01300, 0.45 µm pores, | membrane filter/ 0.45 13 |
Numerical analysis software | Wavemetrics Inc., OR, USA | IgorPro | analysing software |
Stimulation isolator | WPI Inc. | A395 | Stimulus isolator |
AD/DA converter | Instrutech | ITC-18 | AD/DA converter |
Voltage sensitive dye Di-4-ANEPPS | Invitrogen, Thermo-Fisher Scientific, Waltham, MA, USA | catalog number: D-1199 | VSD: Di-4-ANEPPS |
poloxamer | Invitrogen, Thermo-Fisher Scientific, Waltham, MA, USA | Pluronic F-127 P30000MP | poloxamer / Pluronic F-127 (20% solution in DMSO) |
polyethoxylated castor oil | Sigma-Aldrich | Cremophor EL C5135 | polyethoxylated castor oil |