Visualizing Monocarboxylates and Other Relevant Metabolites in the Ex Vivo Drosophila Larval Brain Using Genetically Encoded Sensors

Summary

Here we present a protocol to visualize the transport of monocarboxylates, glucose, and ATP in glial cells and neurons using genetically encoded Förster resonance energy transfer-based sensors in an ex-vivo Drosophila larval brain preparation.

Abstract

The high energy requirements of brains due to electrical activity are one of their most distinguishing features. These requirements are met by the production of ATP from glucose and its metabolites, such as the monocarboxylates lactate and pyruvate. It is still unclear how this process is regulated or who the key players are, particularly in Drosophila.

Using genetically encoded Förster resonance energy transfer-based sensors, we present a simple method for measuring the transport of monocarboxylates and glucose in glial cells and neurons in an ex-vivo Drosophila larval brain preparation. The protocol describes how to dissect and adhere a larval brain expressing one of the sensors to a glass coverslip.

We present the results of an entire experiment in which lactate transport was measured in larval brains by knocking down previously identified monocarboxylate transporters in glial cells. Furthermore, we demonstrate how to rapidly increase neuronal activity and track metabolite changes in the active brain. The described method, which provides all necessary information, can be used to analyze other Drosophila living tissues.

Introduction

The brain has high energy requirements due to the high cost of restoring ion gradients in neurons caused by neuronal electric signal generation and transmission, as well as synaptic transmission^1,2. This high energy demand has long been thought to be met by the continuous oxidation of glucose to produce ATP³. Specific transporters at the blood-brain barrier transfer the glucose in the blood to the brain. Constant glycemic levels ensure that the brain receives a steady supply of glucose⁴. Interestingly, growing experimental evidence suggests that molecules derived from glucose metabolism, such as lactate and pyruvate, play an important role in the brain cells' energy production^5,6. However, there is still some debate about how important these molecules are for energy production and which cells in the brain produce or use them^7,8. The lack of appropriate molecular tools with the high temporal and spatial resolution required for this task is a significant issue that has prevented this controversy from being completely resolved.

The development and application of several engineered fluorescent metabolic sensors have resulted in a remarkable increase in our understanding of where and how metabolites are produced and used, as well as how the metabolic fluxes occur during basal and high neuronal activity⁹. Genetically encoded metabolic sensors based on Förster resonance energy transfer (FRET) microscopy, such as ATeam (ATP), FLII¹²Pglu700µδ6 (glucose), Laconic (lactate), and Pyronic (pyruvate), have contributed to our understanding of brain energy metabolism^10,11,12,13. However, due to the high costs and sophisticated equipment required to conduct experiments on live animals or tissues, results in vertebrate models are still primarily limited to cell cultures (glial cells and neurons).

The emerging use of the Drosophila model to express these sensors has revealed that key metabolic features are conserved across species and their function can be easily addressed with this tool. More importantly, the Drosophila model has shed light on how glucose and lactate/pyruvate are transported and metabolized in the fly brain, the link between monocarboxylate consumption and memory formation, and the remarkable demonstration of how increases in neural activity and metabolic flux overlap^14,15,16,17. The method presented here for measuring monocarboxylate, glucose, and ATP levels using genetically encoded FRET sensors expressed in the larval brain allows researchers to learn more about how the brain of Drosophila uses energy, which can be applied to the brains of other animals.

We show that this method is effective for detecting lactate and glucose in glial cells and neurons, and that a monocarboxylate transporter (Chaski) is involved in lactate import into glial cells. We also demonstrate a simple method for studying metabolite changes during increased neuronal activity, which can be easily induced by bath application of a GABA_A receptor antagonist. Finally, we show that this methodology can be used to measure monocarboxylate and glucose transport in other metabolically significant tissues, such as fat bodies.

Protocol

1. Fly strain maintenance and larval synchronization

1. To perform these experiments, use fly cultures raised at 25 °C on standard Drosophila food composed of 10% yeast, 8% glucose, 5% wheat flour, 1.1% agar, 0.6% propionic acid, and 1.5% methylparaben.
2. To follow this protocol, use the following lines: w¹¹¹⁸ (experimental control background), OK6-GAL4 (driver for motor neurons), repo-GAL4 (driver for all glial cells), CG-GAL4 (driver for fat bodies), UAS-Pyronic (pyruvate sensor), UAS-FLII¹²Pglu700µδ6 (glucose sensor), UAS-Laconic (lactate sensor), UAS-GCaMP6f (calcium sensor), UAS-AT1.03NL (ATP sensor) and UAS-Chk RNAi GD1829. All the lines expressing sensors or RNAi are in the w¹¹¹⁸ genetic background.
3. To obtain synchronized third instar wandering larvae, place 300 flies (3-5 days old, 100 males, 200 virgin females) of the desired genetic cross in the egg-laying chamber, which contains a 60 mm Petri dish covered with 1% agarose in phosphate-buffered saline solution (PBS). Place a drop of liquid/creamy yeast (1.5 cm diameter) in the center of the plaque to encourage flies to lay eggs (Figure 1).
4. Keep the flies at 25 °C for 3 days replacing the agarose Petri dish with a fresh one and freshly dissolved yeast twice daily.
5. Allow the flies to lay eggs for 4 h on the fourth day before changing the Petri dish. Remove this plaque. Then, for 3 h, allow the flies to lay eggs in a new agarose plaque with freshly dissolved yeast. Use these larvae in the experiments.
6. After 3 h, remove the plaque containing the eggs and place it in a 25 °C incubator for 24 h.
7. Collect 50 to 100 newly hatched larvae from this plaque (0 h after larval hatching) and place them in a plastic vial containing standard food. To allow the larvae to feed properly, ensure that the food is ground and soft. Use the larvae 96 h after the transfer.

2. Make the glass coverslips with poly-L-lysine

1. Perform this step 1 h before the larval dissection. In a 6-well cell culture plate, place 25 mm glass coverslips (the diameter of the covers to be used depends on the recording chamber available in the microscope).
2. Place a drop (300 µL) of poly-L-lysine in the center of each coverslip for 30 min at room temperature.
3. Wash each coverslip 3x with distilled water, then 2x with a saline solution devoid of Ca²⁺ (the same solution used to dissect the larvae, see step 3.2). Install the covers in the recording chamber and fill it with Ca²⁺-free saline solution.
   ​NOTE: Although the larval brain can adhere directly to the glass coverslips, the adhesion is weak, and the brain occasionally moves or displaces during the experiment due to the continuous flow of saline solution. The addition of the poly-L-lysine step reduces the risk of movements or even brain loss as a result of the washes.

3. Dissect the ventral nerve cord (VNC) and fat bodies (FBs)

1. Gather the wandering third-instar larvae from the desired genetic cross (from step 1.7) and thoroughly wash them 3x with distilled water.
2. Place the larvae in a glass dissection dish well containing 750 µL of nominally zero Ca²⁺ ice-cold saline solution composed of 128 mM NaCl, 2 mM KCl, 4 mM MgCl₂, 5 mM trehalose, 5 mM HEPES, and 35 mM sucrose (pH = 6.7, measured with a pH meter and adjusted with 1 M NaOH and HCl 37% v/v).
   NOTE: Setting the pH to 6.7 is critical because, as previously observed, monocarboxylate transport is highly dependent on the pH of the solution. In addition, 6.7 corresponds to the pH in the hemolymph of a third instar larva^17,18.
3. Place the larva under a stereomicroscope and make a transverse cut across the back of the abdomen with a pair of forceps.
4. Push the jaw with the forceps while turning the larvae inside out.
5. Observe the ventral nerve cord (VNC) next to the jaw. Carefully remove the imaginal discs and remaining brain-caudal tissues.
6. Separate the VNC with the central brain and optic lobes from the rest of the tissue by cutting the nerves. Transfer the VNC, picking it up with forceps from the remaining nerves, and placing it in the recording chamber containing the Ca²⁺-free recording solution (same solution from step 3.2). The VNCs will immediately adhere to the coverslips.
7. To avoid interference from the remaining nerves, attach them at the bottom of the coverslip using forceps (the nerves will also be fluorescent depending on the driver line used) (Figure 2).
8. To perform experiments in fat bodies (FBs) measuring glucose or monocarboxylate transport in another set of experiments, proceed to isolate the tissues following steps 3.1-3.4. Once the larvae are turned inside out, observe that the FB is a white, bilateral, flat tissue.
9. Place the isolated FBs expressing the FRET sensors (obtained from a genetic cross using the appropriate drivers) in the recording chamber containing the recording solution without Ca²⁺.
   NOTE: The FB is highly hydrophobic (it tends to float on the surface of the solution) and extremely vulnerable to manipulation with forceps, it must be handled with care. Any rough handling of this tissue will result in dead cells later (with a decreased fluorescence of the sensors).
10. Place the recording chamber containing VNCs/FBs on the microscope stage.

4. Live cell imaging

1. To acquire images of the tissue expressing sensors, use an upright fluorescence microscope coupled to an emission splitting system and a CCD camera. Turn on the illumination system of the microscope 30 min before starting any experiment.
   NOTE: Here we use a Spinning Disk fluorescence microscope equipped with a 20x/0.5 water immersion objective to observe both VNCs and FBs. Figure 2 also shows the use of a 40x/0.8 water immersion objective.
2. To visualize GCaMP6f fluorescence, set the excitation and emission wavelengths at 488 nm and 540 nm, respectively. For Laconic/Pyronic/ATP/glucose sensors, set the excitation wavelength at 440 nm and the emission wavelengths at 488 nm (mTFP, CFP) and 540 nm (Venus, YFP).
3. Acquire images of 512 x 512 pixels size every 2 s for GCaMP6f and every 10-30 s for Laconic/Pyronic/ATP/glucose sensors. Determine the optimal exposure to the light source observing the quality of the image obtained. To follow this protocol, ensure that the images are of 300 ms exposure for Laconic and Pyronic sensors and 100-150 ms for the glucose and ATP sensors.
   NOTE: The exposure can vary depending on the use of a confocal or normal fluorescence microscope.
4. Once the recording chamber is installed in the stage of the microscope, carefully place the water immersion objective and be sure that the objective remains submerged throughout the experiment. Keep the room temperature at 25 °C.
5. Connect the recording chamber to a perfusion system and keep the tissues bathed in the recording solution containing 128 mM NaCl, 2 mM KCl, 1.5 mM CaCl₂, 4 mM MgCl₂, 5 mM trehalose, 5 mM HEPES, and 35 mM sucrose, pH 6.7 (measured with a pH meter and adjusted with NaOH and HCl).
6. Maintain a constant flow of 3 mL/min of recording solution through the tissue using gravity, coupled to a low flux peristaltic pump to extract the liquid from the chamber while keeping the volume constant. To achieve the required flow, position the solution-containing tubes (50 mL plastic tubes) 25 cm above the microscope stage.
   NOTE: This gravity-driven flow is used to prevent tissue movement caused by peristaltic pumps, allowing for more accurate image analysis later on.
7. Before any stimulation, maintain the recording solution flowing for 5-10 min to obtain a stable baseline of fluorescence from the sensors. Change the duration as necessary to obtain this baseline.
8. Replace the flowing solution with the stimulation solution for 5 min to stimulate the VNC/FBs (with glucose, pyruvate, or lactate at the concentration described for each experiment dissolved in saline solution; see each figure).
   NOTE: The duration of stimulation can be varied according to the needs of the experiment (for example, glucose pulses can be increased to 10 min to reach a plateau in neurons, as previously reported¹⁶).
9. To maintain osmolarity in the stimulation solution, rectify the sucrose concentration (a carbohydrate nonmetabolizable by the Drosophila brain) according to the concentration of the monocarboxylate or glucose added.
10. Change the solutions using a simple system of valves. Be careful not to move the microscope stage.
11. Expose the VNC to 80 µM picrotoxin (PTX, continuously flowing) to stimulate neuronal activity. This procedure increases the frequency of Ca²⁺ oscillations, making it possible to observe changes in metabolically relevant molecules (e.g., intracellular ATP).
12. At the end of any experiment, thoroughly wash the complete flow system, including the tubes and the recording chamber, for at least 10 min with the saline solution and another 10 min with distilled water to eliminate any trace of the solutions used.

5. Image processing and data analysis

1. To process the images obtained, proceed with the steps shown in Figure 3.
2. Import the image (Laconic) to the ImageJ software. The image has two views of the sample: the left one is the mTFP signal and the one on the right is the Venus signal. Select a region that includes the cells to be analyzed and separate these images (mTFP and Venus) in a new window. Ensure that these images contain exactly the same area.
3. Open the registration plugin and correct the small drift of the tissue that is normally observed in the experiment with the function rigid body. Perform this in both images (mTFP and Venus).
4. Select at least 10 regions of interest (ROIs) in the cytosol of the corresponding 10 cells in the mTFP signal (488 nm) and transfer the selections to the Venus (540 nm) image.
5. Select two or three ROIs close to the cells being analyzed but with no discernible signal (always within the VNC or tissue analyzed, see Figure 3). These are the background ROIs.
6. With the ROIs selected in both images, obtain the mean grey value of each signal using the function measure. Transfer the data to a data sheet and subtract the average background value for each signal.
7. Determine the mTFP fluorescence ratio over Venus (for Laconic and Pyronic). This value is computed differently than the other FRET sensors described here (where the ratio is usually YFP/CFP) because when bound to their respective substrates, both Laconic and Pyronic reduce their FRET efficiencies.
8. Normalize the data dividing each recorded value by the baseline. In a separate data sheet, calculate the baseline by taking the mean value of the mTFP/Venus ratio during the 2 min preceding the stimulus.
9. For the glucose and ATP measurements using FRET-based sensors, calculate the YFP/CFP ratio, and normalize the data as described in step 5.8.

Results

For up to 1 h, this procedure allows for easy measurement of intracellular changes in the fluorescence of monocarboxylate and glucose sensors. As shown in Figure 4, Laconic sensors in both glial cells and motor neurons respond to 1 mM lactate at a similar rate at the start of the pulse, but motor neurons reach a higher increase over the baseline during the 5 min pulse, as previously demonstrated¹⁷. This lactate concentration was chosen because it is comparable to the ...

Discussion

The use of the Drosophila model for the study of brain metabolism is relatively new²⁶, and it has been shown to share more characteristics with mammalian metabolism than expected, which has primarily been studied in vitro in primary neuron cultures or brain slices. Drosophila excels at in vivo experiments thanks to the battery of genetic tools and genetically encoded sensors available that allows researchers to visualize in real time the metabolic changes caused...

Materials

Name	Company	Catalog Number	Comments
Agarose	Sigma	A9539
CaCl2	Sigma	C3881
CCD Camera ORCA-R2	Hamamatsu	-
Cell-R Software	Olympus	-
CG-GAL4	Bloomington Drosophila Stock Center	7011	Fat body driver
Dumont # 5 Forceps	Fine Science Tools	11252-30
DV2-emission splitting system	Photometrics	-
Glass coverslips (25 mm diameter)	Marienfeld	111650	Germany
Glucose	Sigma	G8270
GraphPad Prism	GraphPad Software	Version 8,0,2
HEPES	Sigma	H3375
ImageJ software	National Institues of Health	Version 1,53t
KCl	Sigma	P9541
LUMPlanFl 40x/0.8 water immersion objective	Olympus	-
Methylparaben	Sigma	H5501
MgCl2	Sigma	M1028
NaCl	Sigma	S7653
OK6-GAL4	Bloomington Drosophila Stock Center		Motor neuron driver
Picrotoxin	Sigma	P1675S	CAUTION-Fatal if swallowed
Poly-L-lysine	Sigma	P4707
Propionic Acid	Sigma	P1386
Repo-GAL4	Bloomington Drosophila Stock Center	7415	Glial cell driver (all)
Sodium Lactate	Sigma	71718
Sodium pyruvate	Sigma	P2256
Spinning Disk fluorescence Microscope BX61WI	Olympus	-
Sucrose	Sigma	S0389
Trehalose	US Biological	T8270
UAS-AT1.03NL	Kyoto Drosophila Stock Center	117012	ATP sensor
UAS-Chk RNAi GD1829	Vienna Drosophila Resource Center	v37139	Chk RNAi line
UAS-FLII12Pglu700md6	Bloomington Drosophila Stock Center	93452	Glucose sensor
UAS-GCaMP6f	Bloomington Drosophila Stock Center	42747	Calcium sensor
UAS-Laconic	Sierralta Lab	-	Lactate sensor
UAS-Pyronic	Pierre Yves Placais/Thomas Preat	-	CNRS-Paris
UMPlanFl 20x/0.5 water immersion objective	Olympus	-