Calcium Imaging of Anopheles coluzzii Mosquito Antennae Expressing the Calcium Indicator GCaMP6f

Abstract

Calcium imaging is a technique used to measure functional neuronal activities in response to stimuli. It has been used for years to study odorant-induced responses in insects (i.e., honeybees, Drosophila, and moths) and was recently introduced into mosquitoes. Traditionally, calcium imaging in mosquitoes was performed using nonspecific calcium indicator dyes to examine neuronal responses in whole insect brain regions, but the development of genetically encoded calcium indicators (GECIs) has facilitated the ability to perform functional calcium imaging on specific tissues. For example, by specifically expressing a GECI in olfactory neurons, the odor-induced responses of these neurons in peripheral organs can be examined. Calcium imaging of mosquito antennae further provides an advantageous method for simultaneously visualizing the activity of several antennal neurons in a single experiment. In this protocol, we describe a calcium imaging method to study odor-evoked responses in Anopheles coluzzii antennae expressing the calcium indicator GCaMP6f. This method requires imaging equipment (compound microscope, light sources, and camera), an odorant delivery system, and image acquisition software. The mosquito preparation is straightforward but requires practice to minimize mosquito movement during imaging. Recorded videos can be analyzed using Fiji software to generate heatmaps and activity traces for odorant-evoked responses. This protocol can also be used, with some modifications, to study other peripheral organs (such as labella, palps, and tarsi).

MATERIALS

It is essential that you consult the appropriate Material Safety Data Sheets and your institution‘s Environmental Health and Safety Office for proper handling of equipment and hazardous materials used in this protocol.

Reagents

Anopheles coluzzii mosquitoes (3- to 10-d-old non-blood-fed females) that express GCaMP6f in Orco⁺ neurons

• We generated these mosquitoes by crossing AgOrco-QF2 (BEI Resources MRA-1300) and QUAS-GCaMP6f mosquitoes (available upon request to C. Potter) (Afify et al. 2019).

• This protocol can also be adapted to other species of mosquitoes in which olfactory neurons are similarly labeled by a calcium sensor. However, it is possible that a darker cuticle, such as that typically found in Aedes mosquito antennae, might interfere with calcium imaging.

Pure odorant

• In this example, we use 1-octen-3-ol (SAFC W280518).

Solvent for odorant

• Choose a solvent that has a weak or no odor and that completely dissolves the test odorants. Mineral and paraffin oils are good solvents, but odorants in solid forms might not dissolve in them and require a solvent like ethanol. In this example, we use paraffin oil (Sigma-Aldrich 18512).

Equipment

Calcium imaging setup

10× objective (Zeiss EC Epiplan-Neofluar 10x/0.25)

• This objective is used to image the whole mosquito antenna.

50× objective (LD EC Epiplan-Neofluar 50×/0.55 DIC)

• This objective is used to image single segments of the mosquito antenna.

Air isolator table (TMC Gimbal Piston)

Compound microscope (Zeiss Axio Examiner D1)

EMCCD camera (Andor iXon Ultra)

• To capture movies/images, we use this camera mounted on top of the microscope and connected to a computer with image acquisition software.

Fluorescence light source (Zeiss Illuminator HXP 200C) with eGFP filter cube (FL Filter Set 38 HE GFP shift free)

NIS Elements Advanced Research software (Nikon instruments) or Andor Solis software

• This software is used for image acquisition.

Stimulus controller (Syntech CS-55)

• A calcium imaging setup consists of a compound microscope with bright-field and fluorescence light sources, a camera with image acquisition software, and an odorant delivery system. The microscope should be mounted on an air table to reduce movements during imaging. There are many products on the market that offer different solutions and can be chosen based on the type of planned experiments. In this protocol, we used the products listed above.

Conical tubes (50-mL)

Coverslips

Dissecting microscope (Zeiss Stemi 2000)

Dremel rotary tool (Dremel 8000–03)

Fiji software with the Image Stabilizer and HeatMap Histogram plug-ins (Schindelin et al. 2012)

Filter paper (Whatman 1001090)

Forceps (Dumont #3)

Handheld insect aspirator (John W. Hock Company PN 419)

Micromanipulator (Kite Manual Micromanipulator, World Precision Instrument)

Microscope slide

Modeling clay (Crayola)

Pasteur pipettes (14.6-cm; Fisher Scientific 13–678-6A)

Plastic serological pipette (10-mL; Denville Scientific Inc)

Pulled glass capillary tubes (1-mm×0.5-mm×100-mm [OD×ID×L]; Harvard Apparatus)

• The exact size of the pulled glass capillaries is not crucial. They can be pulled by a filament- or laser-based micropipette puller.

PVC tubing, clear (4.8-mm inner diameter, 15-cm-long) (two; see Step 2)

PVC tubing, clear (4.8-mm inner diameter, ∼1.5-m-long)

PVC tubing, clear (8-mm inner diameter, ∼2-m-long)

Razor blade

Y-shaped connection (e.g., Wye Connectors, Tube, or McMaster-Carr)

METHOD

Setup of Calcium Imaging Equipment

• 1Program the A and B output ports on the Syntech stimulus controller to deliver equal amounts of air. Connect a long clear PVC tubing (4.8-mm inner diameter, ∼1.5-m-long) to the A port (Fig. 1A). Next, connect the foot pedal (included with the stimulus controller) to the “Pedal Switch” port on stimulus controller and place that pedal under the air table.

  During imaging, the pedal will be used to divert air between the A and B ports (see the Imaging section below).

• 2Connect two short (15-cm-long) clear pieces of PVC tubing (4.8-mm inner diameter) to the B and C output ports on the Syntech stimulus controller. Use a Y-shaped connection to connect the two short lengths of tubing to one long piece of tubing (8-mm inner diameter, ∼2-m-long). Insert a 10-mL plastic serological pipette at the end of the long tubing (Fig. 1A).
• 3Make a small hole (3-mm diameter) using a Dremel rotary tool into the long plastic pipette 20 cm from the tip. Mount the pipette on a micromanipulator so it can be directed at the mosquito antenna.

  During imaging, this long pipette will deliver charcoal-filtered continuous air from the stimulus controller to the antenna, and the small hole will be used to inject an airstream containing odorants into the continuous airstream (Fig. 1A; see the Imaging section below).

  The stimulus controller can be adjusted to deliver air at a desired airflow. We use 8.3 mL/sec.

FIGURE 1.

Calcium imaging of antennal olfactory neurons in Anopheles coluzzii mosquitoes. (A) Schematic of the imaging setup. Arrows indicate the direction of airflow. A 1-sec air pulse leaves the A port of the stimulus controller (rectangle), passes through the Pasteur pipette that contains the odorant, and ends in the long pipette directed at the mosquito antenna. Continuous air starts from the B and C ports of the stimulus controller and ends in the long pipette. (B) Mosquito preparation for imaging. (1) The mosquito is inserted into the pipette tip using forceps. (2) Forceps are used to push the mosquito thorax and move the mosquito forward. (3) The pipette tip is cut from the front and back around the mosquito. (4) A small piece of clay is used to push the mosquito forward. (5) The mosquito is trapped in the pipette tip with only the antenna stretching out. (6) Two pieces of clay are used to attach the pipette tip and a coverslip to the slide. (C) The mosquito is ready for imaging. The mosquito is trapped in a pipette tip with only the antennae extended outside the pipette tip. Two pulled glass capillaries are used to hold the antenna down on a coverslip. (D) Example heatmap of the responses toward 1% of 1-octen-3-ol (a human odorant). Dashed red line indicates the borders of the 11th antennal segment where imaging was performed. (E) Traces from the calcium imaging recording in C. Colors represent different ROIs (neurons), and the y-axis shows ΔF/F values (odor-evoked change in fluorescence from baseline divided by baseline fluorescence).

Odorant Preparation

• 4Prepare odorant stocks by diluting pure odorants in a solvent.

  In this example, we used 1-octen-3-ol, a human skin odorant, diluted to a 1% concentration in paraffin oil.

  The solvent will also be used as the control odorant.

  Diluted odorants should not be kept for extended periods (e.g., >3 mo) because they change over time (degrade or are oxidized). These stocks should be kept in tightly closed and dark containers at the same temperature as recommended for the pure odorant, away from direct sun light, and discarded/reprepared on a regular basis.

• 5On the day of the experiment, place a piece of filter paper (1-cm× 2-cm) in a glass Pasteur pipette and then pipette the desired amount of diluted odorant (∼20 μL) onto the filter paper. Cover the wide end of the Pasteur pipette with a 1000-μL pipette tip. Prepare pipettes with solvent only (e.g., paraffin oil or mineral oil) as a control.

  Pasteur pipettes containing odorants should be prepared fresh every day, preferably before preparing mosquitoes for imaging and discarded at the end of the experiment or after approximately five odorant deliveries. More volatile odorants should be reprepared more frequently.

Mosquito Preparation

• 6Use a handheld aspirator to insert a single mosquito into a 50-mL conical tube. Place the tube on ice for ∼1 min to immobilize the mosquito.
• 7Under a dissecting microscope, use forceps to carefully hold the mosquito from the tip of its abdomen and insert the mosquito into a 200-μL pipette tip (Fig. 1B1), with the mosquito head facing the tip of the pipette tip. Use forceps to push the mosquito thorax and move the mosquito toward the tip of the pipette tip (Fig. 1B2), while making sure the two antennae are always extended forward, until the mosquito is halfway through the pipette tip. Then, use a sharp razor to cut the front of the pipette tip so the front opening is around the same diameter as the mosquito head (∼1-mm in diameter, Fig. 1B3).

  When pushing the mosquito forward, the antennae might bend to the back of the pipette tip. This will lead the fragile antennae to be stuck between the mosquito thorax and the wall of the pipette tip. This can be resolved by pulling the mosquito back and pushing it forward again. Alternatively, carefully insert one branch of the fine forceps from behind the mosquito to push the antennae forward until it is freed from the thorax.

• 8Push the mosquito again using forceps, while making sure the antennae are always extended forward, until the antennae reach the front opening of the pipette tip. Cut the back of the pipette tip right behind the mosquito abdomen (Fig. 1B3) and use small pieces of modeling clay to push the mosquito forward until only the antennae are extended outside the pipette tip (Fig. 1B4,5).
• 9Attach a small piece of modeling clay to a glass slide and press the pipette tip down into the clay (Fig. 1B6). Use another piece of modeling clay to attach a glass coverslip to the slide in front of the pipette tip, in front of the mosquito antennae (Fig. 1B6). Now, use a finger to push the coverslip toward the pipette tip and place the antennae on the coverslip (Fig. 1C).
• 10Put a piece of modeling clay on the slide on each side of the pipette tip, and mount two pulled glass capillaries on the two pieces of clay. Place the pulled tip of one of the capillaries by hand on the third to fourth segment of one antenna to press it down on the coverslip. Place the other capillary on the 12th to 13th antennal segment (the two most distal segments) and make sure the antenna is flattened on the slide (Fig. 1C).

  Flattening the antenna on the coverslip is the most crucial step in mosquito preparation and requires practice. The fragile antenna can be easily damaged by the capillaries. However, failing to adequately flatten the antenna allows it to move while imaging; this will also cause the antenna to slightly bend which will require imaging at different Z planes.

  Now the mosquito is ready for imaging.

Imaging

• 11Place the prepared mosquito on the stage of the compound microscope. Turn on the bright-field light source, the fluorescence light source, the stimulus controller, and the camera.
• 12Start the image acquisition software, which has many settings to control the length, quality, and format of the output videos. We record 20-sec videos, at a resolution of 512 × 512 pixels, and an exposure time of 200 msec (5 Hz).

  Although the length of videos can be freely adjusted, a lower resolution or slower frame rate could make the analysis of the imaging videos challenging and potentially miss important data.

  The image acquisition software will give an error message if it is started before the camera is turned on.

• 13Check the antenna through the microscope eyepieces at low (10×) and high (50×) magnification to make sure it is flattened on the coverslip. Locate the shutter on the microscope, which allows switching between viewing through the eyepieces and viewing with the image acquisition software on the computer screen. Open this shutter to view live images on the image acquisition software.
• 14Use fluorescent light to verify that antennal neurons are visible.

  For viewing GCaMP6-labeled neurons, use a GFP filter cube (stimulate with 470-nm light, visualize the 525-nm emitted light). If the antenna is not adequately flattened, only a few neurons will be visible in a single plane under fluorescent light.

  Close the shutter on the fluorescence light source when not using the fluorescent light. This will protect neurons from continuous exposure to fluorescent light and prevent photobleaching of the GCaMP fluorescent protein.

• 15Use the micromanipulator to direct the long pipette (the one connected to the B and C ports of the stimulus controller) at the mosquito antenna (∼1 cm away from the antenna).

  This will deliver a continuous charcoal-filtered airstream to the antenna (Fig. 1A).

• 16To deliver odorants to the antenna, connect the back of the Pasteur pipette that contains odorant-soaked filter paper to the long tubing that is connected to the A port of the stimulus controller. Then, insert the tip of the Pasteur pipette into the hole in the long pipette (Fig. 1A).
• 17Start recording the video on the software. Allow a few seconds of recording before puffing the odorant.

  We deliver the odorants on the 10th sec of a 20-sec video. The 9-sec segment at the beginning of each video serves as the “prestimulus” and represents the neurons’ baseline activity.

• 18To deliver the odorant (on the 10th sec), press the foot pedal connected to the stimulus controller.

  This will divert a 1-sec pulse of air into the Pasteur pipette and deliver the odorant from the filter paper into the continuous airstream, reaching the mosquito antenna at the end.

  Program the stimulus controller to deliver equal amounts of air through the A and B ports at 5 mL/sec. Airstreams from the B and C ports merge at the Y-shaped part of the hose to deliver 8.3-mL/sec continuous airstream to the antenna. When the foot pedal is pressed, the stimulus controller will stop air from the B port and will deliver air through the A port for 1 sec. This will ensure that a continuous undisturbed airstream reaches the mosquito antenna at all times. To identify potential issues caused by odor-induced changes to neuronal activity, randomize the stimulus order of odorants and controls.

• 19Save the video in a suitable format.

  In Solis software, save videos in AVI format using the “Export As” function. This function uses different codecs to compress videos. Use the “raw video” codec which allows the videos to be imported by Fiji for data analysis. The “raw” (uncompressed format) is also recommended to avoid potential compression artifacts being introduced into the images.

Analysis of Calcium Imaging Recordings

There are different ways to extract data from calcium imaging videos. We use Fiji (version 1.53c), an open-source software, to produce two types of data outputs: (1) antennal heatmaps—these heatmaps assign colors to different parts of the mosquito antenna based on their response level (Fig. 1D) and (2) intensity time traces—these are the mean intensity values of different regions of interests (ROIs) on the antenna across time (Fig. 1E).

Antennal Heatmaps

• 20The first time Fiji is needed, install the latest version of Fiji and the Image Stabilizer plug-in into Fiji (https://imagej.net/plugins/image-stabilizer) (Li and Kang 2008).
• 21Import the calcium imaging video “x.avi” in Fiji. File → Open, then choose the video to be imported. Be sure to uncheck the “Use Virtual Stack” box.

  Note that the imported video will show some information like the number of slices (frames) and duration of the video. The number of frames will depend on the recording speed and the duration of the video. If the recording was at 5 Hz and for 20 sec, it will be a 100-frame video.

• 22Use the Image Stabilizer plug-in to correct for any movements that occurred during recording. Go to Plugins → Image Stabilizer. Check the “Output to a New Stack” box. Be aware that this will produce a new corrected or stabilized video, which will be automatically named “Stabilized x,” while keeping the original video “x.” Close the original video file to avoid any confusion.
• 23Produce the ΔF image, which shows the change in fluorescence intensity after odorant delivery. To do this, first choose the frame with the maximum fluorescence intensity “F_i” by going through the stabilized video, frame by frame, and manually choosing the frame that shows the maximum intensity after odorant delivery. Try to choose the frame that contains the brightest spots; sometimes multiple olfactory neurons reach their maximum responses at different times. In this case, use the frame in which the majority of the neurons have reached their maximum responses. Next, click Image → Stacks → Tools → Slice Keeper and write the number that corresponds to the “F_i” frame in the “First Slice” and in the “Last Slice.” Input “1” in the “Increment” window, which will extract the “F_i” image from the stabilized video and automatically name it “Stabilized x kept stack.”
• 24Use the stabilized video to produce an F₀ image—that is, the image with the mean baseline fluorescence (“prestimulus” fluorescence), which will be representative of the 9 sec before odorant delivery (45 frames). To do this, click Image → Stacks → Z project, and write 1 in the “Start slice,” 45 in the “Stop slice,” and choose Average Intensity from the “Projection type” drop-down menu.

  The output image will be automatically named “AVG_ Stabilized x.”

• 25Produce the ΔF image by subtracting the mean baseline fluorescence F₀ from the image of maximum fluorescence F_i. To do this, click Process → Image calculator, and choose “Stabilized x kept stack” in the “Image1” drop-down menu, “subtract” in the “Operation” drop-down menu, and “AVG_ Stabilized x” in the “Image2” drop-down menu. Check the “Create new window” box to produce the ΔF image in a new window automatically named “Result of Stabilized x kept stack.”
• 26Produce a heatmap from the ΔF image using a plug-in called “HeatMap Histogram.” Install the plug-in into Fiji (https://www.researchgate.net/publication/322420028_HeatMap-Histogram_from_Samuel_Pean) (Pean 2018), then click on the ΔF image (Result of Stabilized x kept stack) to select it. Click Plugins → HeatMap Histogram to produce two identical heatmap images, one of which will contain a calibration bar for the color code of the heatmap (Fig. 1D). Save those heatmaps using the “Save As” function from the “File” drop-down menu. Close all windows except the “Stabilized x” video, which will be used to produce intensity time traces.

  Available methods for generating heatmaps differ in the colors they assign to different intensity levels.

Intensity Time Traces

• 27Click Analyze → Tools → ROI Manager, which will open the ROI Manager window that can be used to draw ROIs.
• 28Select the “Oval” selection tool from the main Fiji window and use it to draw circles or oval outlines around the regions of interest (responding neurons) in the “Stabilized x” video. To do this, go through the video, frame by frame, and draw outlines around the neurons that show increased fluorescence after odorant delivery. The heatmap can be used as a reference to choose the responding neurons.
• 29After drawing an outline around each neuron/ROI, click “Add [t]” in the ROI manager window to save it. Otherwise, it will be lost if a new ROI is drawn before saving.

  Check the “Show All” box in the ROI manager to show the ROIs on the video. If the “Labels” box is checked, it will show the number assigned to each ROI in the middle of that ROI.

• 30After drawing all ROIs, click “More” in the ROI manager → Multi Measure, and check the “Measure all 100 slices” and the “One row per slice” boxes. This will produce a “Results” window that contains one column numbered 1–100 (the number of each frame), and one column for each ROI (Mean1, Mean2, Mean3, …). Each of the ROI columns will contain mean intensity values from all pixels in that ROI across time (all frames).

  Before clicking “Multi Measure,” make sure not to select any particular ROI in the video. Otherwise, the produced “Results” window will contain intensity values for only that particular ROI.

• 31Click “File” in the “Results” window → Save as, to save the intensity values as an Excel spreadsheet.
• 32
  Finally, calculate ΔF/F values in an Excel spreadsheet using the formula

  $$\Delta F/F=F_i-F_0/F_0,$$

  where F_i is the fluorescence intensity value at frame i (the frame with maximum intensity) and F₀ is the mean fluorescence intensity before odorant delivery (first 9 sec, 45 frames).

  ΔF/F values can be used to plot time traces (Fig. 1E) in R (R Core Team 2018). Odorant responses or further analysis can be compared using R or other statistical analysis software.