This protocol describes extracellular recording of the action potential responses fired by labellar taste neurons in Drosophila.
The peripheral taste response of insects can be powerfully investigated with electrophysiological techniques. The method described here allows the researcher to measure gustatory responses directly and quantitatively, reflecting the sensory input that the insect nervous system receives from taste stimuli in its environment. This protocol outlines all key steps in performing this technique. The critical steps in assembling an electrophysiology rig, such as selection of necessary equipment and a suitable environment for recording, are delineated. We also describe how to prepare for recording by making appropriate reference and recording electrodes, and tastant solutions. We describe in detail the method used for preparing the insect by insertion of a glass reference electrode into the fly in order to immobilize the proboscis. We show traces of the electrical impulses fired by taste neurons in response to a sugar and a bitter compound. Aspects of the protocol are technically challenging and we include an extensive description of some common technical challenges that may be encountered, such as lack of signal or excessive noise in the system, and potential solutions. The technique has limitations, such as the inability to deliver temporally complex stimuli, observe background firing immediately prior to stimulus delivery, or use water-insoluble taste compounds conveniently. Despite these limitations, this technique (including minor variations referenced in the protocol) is a standard, broadly accepted procedure for recording Drosophila neuronal responses to taste compounds.
The sense of taste allows an insect to detect a vast range of soluble chemicals and plays an important role in the acceptance of a nutritious substance, or the rejection of a noxious or toxic one. Taste is also thought to play a role in mate selection, through the detection of pheromones1-5. These important and diverse functions have made the insect taste system a compelling target of investigation into how sensory systems translate environmental cues into relevant behavioral outputs.
The primary unit of the Drosophila melanogaster taste system is the taste hair, or sensillum. Molecules enter the sensillum via a pore at its tip2,6. Sensilla are found on the labellum, the legs, the wing margin, and the pharynx6. On the labellum, the number and location of sensilla is stereotyped. There are three morphological classes of sensilla based on length: the long (L), intermediate (I), and short (S) sensilla7,8. Each sensillum contains either two (I-type) or four (L- and S- type) gustatory receptor neurons (GRNs)9. Different GRNs respond to different categories of taste stimuli: bitter, sugar, salt and osmolarity7,10 and express different subsets of gustatory receptors8,11-13. Only I and S-type sensilla contain bitter-responsive GRNs8,10. The GRNs project to the subesophageal ganglion (SOG) and their activation by taste molecules is relayed to the higher central nervous system for decoding, resulting in a behavioral response6. The relatively small number of neurons and the amenability to molecular and behavioral analysis make the Drosophila taste system an excellent model for the investigation of gustatory systems in general. The relative ease with which the system can be manipulated via genetic mutation or the GAL4-UAS expression system also serves as a valuable tool14,15.
Because these sensilla protrude from the surface of the labellum, they make excellent targets for electrophysiology. The firing of the GRNs can be monitored using extracellular recording. Historically, the side-wall recording method, which uses a glass electrode inserted into the sensillum to record neuronal activity,26 has been used. However, this method is technically challenging to perform, and it is difficult to record for long from each preparation. The tip-recording method, which measures the response of the neurons with an electrode that simultaneously delivers a tastant, has since become the method of choice9,16. It has been utilized to investigate the taste system of Drosophila melanogaster8,10,17,18 as well as a number of other insect species19-23. It has been greatly facilitated by the development of the tastePROBE amplifier, which overcame one of the major drawbacks of the tip-recording method by compensating for the large potential difference between the reference electrode and the insect sensillum, allowing the GRN action potentials to be recorded without excessive amplification or filtering24. Another important development was the use of tricholine citrate as the recording electrolyte25. TCC suppresses responses from the osmolarity-sensitive GRN and does not stimulate the salt-sensitive GRN, making responses generated by bitter and sugar tastants much easier to analyze25.
Here we describe how tip recording of Drosophila labellar sensilla is currently performed in the Carlson laboratory. This protocol will explain how to establish a suitable electrophysiology rig, how to prepare the fly, and how to perform taste recordings. We also present some representative data obtained by recording from subsets of Drosophila sensilla, as well as some common issues and potential solutions that may be encountered when using this technique.
The following protocol complies with all the animal care guidelines of Yale University.
1. Reagents and Equipment Preparation
2. Drosophila Preparation
Figure 3. Preparation of fly for recording. (A) Insertion position of reference electrode into dorsal thorax of fly. The white arrow indicates the reference electrode. (B) Intermediate position of reference electrode: advanced through neck and head, proboscis not yet extended. (C,D) Fly with reference electrode in final position with tip of electrode inside labellum, and proboscis fully extended. Click here to view larger image.
3. Recording from Labellar Sensilla
Figure 4. Recording from fly. (A) labellum of fly preparation on left with recording electrode aligned for contact on right, under high magnification. (B) recording electrode and single sensillum on labellum in contact, under high magnification. Click here to view larger image.
Figure 5A shows the response of an L sensillum to a sugar, sucrose. The same sensillum does not respond to a bitter compound, berberine. Figure 5B shows that an I type sensillum, which contains a bitter responsive neuron, displays larger amplitude spikes in response to berberine, and smaller amplitude spikes in response to sucrose. L sensilla display a minimal background response to the solvent control, TCC, while I sensilla display virtually no response to TCC (Figure 5). For more information on salt and water responses of labellar GRNs, please refer to Hiroi10.
Figure 5. Representative traces of wild-type Drosophila labellar responses (A) L sensillum response to 100 mM sucrose (SUC), 1 mM berberine (BER), and 30 mM TCC. (B) I sensillar response to SUC, BER, and TCC. The arrowhead indicates the contact artifact that occurs at the beginning of each recording. Click here to view larger image.
Figure 6. Representative suboptimal electrophysiological results. (A) complete lack of signal (B) 50/60 Hz “noise” (C) stochastic noise (D) mechanosensory neuron firing alone (E) bitter GRN (open triangles) and mechanosensory neuron (filled triangles) both firing. Click here to view larger image.
Labellar sensilla vary in the ease of recording due to differences in morphology and anatomical organization. Sometimes a sensillum does not respond to any tastants, even one that is known to elicit a positive response. The frequency with which this occurs varies depending on sensillum type. L sensilla are most consistently responsive and are relatively easy to access due to their length. In general, S sensilla are consistently responsive, but their short length and position on the labellum make good contact challenging. I sensilla can be accessed more readily, depending on the angle of the preparation; however, they are more frequently unresponsive. On any given fly preparation, a greater proportion of I sensilla may be unresponsive than L or S sensilla. Genetic background can affect the consistency of taste responses as well. For example, some transgenic flies may display less consistent responses than wild-type, presumably because the transgenes affect the general health of the fly. We have observed that w– mutant flies are particularly challenging to record from.
One common technical problem is a lack of signal, i.e. no spikes are observed (Figure 6A). First, sometimes one particular sensillum may be unresponsive, while others of that same class on the same fly may respond. Second, there may be an air bubble in the recording electrode or the reference electrode. If the recording electrode is suspected, this can be fixed by simply removing and refilling the glass electrode, tapping gently and inspecting under magnification to ensure no bubbles are present. If the reference electrode is suspected to contain an air bubble, remaking the prep with a new fly is the easiest way to resolve this issue. Third, sometimes the wires carrying the electrical signal may not be securely connected. Fourth, occasionally the voltage signal being received may be either higher or lower than the range the amplifier can measure. If using the tastePROBE amplifier, check to see if either the clip up or clip down indicator light is on. If the clip up indicator light is on, often removing and refilling the glass reference electrode, while taking care to fill not more than halfway and wiping down the outside to remove moisture will resolve the problem. Moisture on the outside of the glass electrode can make an electrical connection between the metal case of the electrode and the wire, sending the signal out of range of the amplifier. If that fails to solve the issue, or the clip down indicator light is on, consider suggestions in the following paragraph to combat electrical noise in the system. Fifth, sometimes a fly may die during preparation or is otherwise unresponsive despite the preparation’s healthy appearance. Growth conditions, such as humidity, temperature, age, food quality, and microbiota, as well as a less healthy genetic background could contribute to a higher proportion of “unresponsive” flies. Lastly, rarely, a piece of equipment may be nonfunctional. If signal is consistently not being achieved and all other possibilities have been exhausted, it may be necessary to investigate the functionality of each piece of equipment: headstage, amplifier, and digitizer. The easiest way to do this is to replace a piece of equipment with another from a rig that is known to be functional. If only one rig is present in a lab, a signal generator can be used to test functionality of the electronic components.
Another common technical issue is that of “noise,” which is an observed signal that does not appear to represent neuronal action potentials fired in response to a gustatory stimulus (Figures 6B-E). First, the signal may result from 50/60 Hz electrical noise from recording equipment or other equipment nearby (Figure 6B). With no fly on the reference electrode, directly connect the recording and reference electrodes through a drop of Ringer’s solution and enter the passthrough mode on the amplifier by pressing the up button. If noise is observable on the passthrough signal, this likely means that the noise is external to the fly preparation. Ensure that all rig equipment is properly grounded and that tin foil shields are in place. Try unplugging nearby equipment to see if the noise is eliminated, or shield additional components. Second, the noise may appear stochastic (Figure 6C). In this case, the steps detailed for 50/60 Hz noise should still be undertaken. Additionally, try unplugging or replacing different components of the recording equipment, particularly the headstage and/or amplifier. If no noise is observed when the electrodes are directly connected, the source is likely the fly preparation itself. It is usually simplest to prepare a new fly for recording, taking care to minimize damage to the fly. Third, activation of the mechanosensory neuron contained within the sensillum (Figures 6D and 6E) may be observed. The mechanosensory neuron can be activated if the sensillum is deflected or bent upon application of the recording electrode, or bumped during contact. The spikes are usually distinguishable from chemosensory spikes by their irregular pattern, which usually appears coordinated with the mechanical disruption, not the application of a gustatory stimulus. Mechanosensory firing can be minimized by aligning the recording electrode with the sensillum and advancing gently only as far as is necessary to make contact with the tip of the sensillum. Fourth, stochastic spike “bursting” may be observed; this appears similar to neuronal firing, but is of high frequency and amplitude, not coordinated in response to a stimulus. This usually results from the fly prep itself, not from the equipment, and may be due to a nerve disrupted by the reference electrode.
A third common technical issue is that the preparation is mobile, causing the labellum to move, which makes connection with a sensillum difficult. First, the fly preparation may be unstable. Check that the reference electrode is correctly positioned, and readjust if necessary. Second, the reference electrode may be too thin at the tip to hold the proboscis and labellum immobile. Try breaking off a longer amount of the tip before preparing the fly. If that is not sufficient, readjust the pipette puller settings as needed to change the shape of reference electrode such that the taper is more gradual and the diameter is slightly increased. Third, the fly may be unusually active. Remake the preparation with a new fly.
For general electrophysiology information and more troubleshooting guidance, refer to Axon Guide30.
There are a few limitations to the tip-recording method outlined in this publication. One limitation is that the tastant must be water soluble, as it is delivered in the recording electrode along with the electrolyte. This increases the difficulty of recording with hydrocarbon compounds, though use of a solvent like DMSO has made some recording with pheromones possible4. Alternative approaches are to use a sharpened tungsten electrode to perform the recordings from the socket base of the sensillum, or use a glass electrode to perform recordings from the side wall of the sensillum, in both of which the tastant is delivered independently of the recording electrode26,27. However, these techniques are challenging and side-wall recordings are more injurious to the taste organ. Another limitation is the amount of time required to exchange the tastant solution (Protocol step 3.3), which reduces throughput, and limits the use of complicated stimulus paradigms often seen in olfactory recordings. Gustatory receptor neurons exhibit some variability in amplitude that is dependent on spike frequency. This feature can complicate assessment of neuronal identity and make advanced spike sorting more difficult25,31-33. In addition, because of the nature of the tip-recording method one cannot record the basal firing immediately prior to the delivery of a stimulus, as is commonly done in olfactory recordings. Despite these drawbacks, the tip-recording method has been successfully used to elucidate many of the principles of taste coding in Drosophila and other species8,10,17,19,21-23.
The fly preparation technique outlined here is just one possible approach. In this preparation method the proboscis is fixed in an extended position to facilitate contact of the recording electrode with the sensillum of interest, and the reference electrode is inserted into the animal. Other preparation methods include the mounting of the animal to a ball of modeling clay and the use of thin strips of tape to fix the proboscis34. Indeed, as long as the basic parameters of tissue stabilization and reference electrode placement are met, sensilla in other locations or from different species can be recorded from in much the same way. For example, leg sensilla can be recorded from by fixing the body of a fly to a sylgard-coated microscope slide with fine insect pins, splaying the legs off the edge of the glass slightly35. It is possible to deliver pharmacological agents to the sensilla via the recording electrode to investigate signal transduction in the gustatory receptor neurons. It is simply a task of experimentation to determine which approach works best for the desired outcome.
The authors have nothing to disclose.
This work was supported by an NRSA predoctoral grant 1F31DC012985 (to R.D.) and by NIH grants to J.C.
We would like to thank Dr. Linnea Weiss for helpful comments on the manuscript, Dr. Ryan Joseph for help compiling figures, and Dr. Frederic Marion-Poll for helpful technical advice. We would also like to acknowledge the helpful comments of four reviewers.
Stereo Zoom Microscope | Olympus | SZX12 DFPLFL1.6x PF eyepieces: WHN10x-H/22 | capable of ~150x magnification with long working distance table mount stand |
Anti-vibration Table | Kinetic Systems | BenchMate2210 | |
Micromanipulators | Narishige | NMN-21 | |
Magnetic stands | ENCO | Model #625-0930 | |
Reference Electrode Holder | Harvard Apparatus | ESP/W-F10N | Can be mounted on 5ml serological pipette for extended range |
Silver Wire | World Precision Instruments | AGW1510 | 0.3-0.5mm diameter |
Retort Stand | generic | ||
Outlet Plastic Tube | generic, 1cm diameter | ||
Flexible Plastic Tubing | Nalgene | 8000-0060 | VI grade 1/4 in internal diameter |
500 ml Conical Flask | generic, with side arm | ||
Aquarium Pump | Aquatic Gardens | Airpump 2000 | |
Fiber Optic Light Source | Dolan-Jenner Industries | Fiber-Lite 2100 | |
White Card/Paper | Whatman | 1001-110 | |
Digital Acquisition System | Syntech | IDAC-4 | Alternative: National Instruments NI-6251 |
Headstage | Syntech | DTP-1 | Tasteprobe |
Tasteprobe Amplifier | Syntech | DTP-1 | Tasteprobe |
Alligator Clips | Grainger | 1XWN7 | Any brand is fine |
Insulated Electrical Wire | Generic | ||
Gold Connector Pins | World Precision Instruments | 5482 | |
Personal Computer | Dell | Vostro | Check for compatibility with digital acquisition system and software |
Acquisition Software | Syntech | Autospike | Autospike works with IDAC-4; alternatively, use Labview with NI-6251 |
Aluminum Foil and/or Faraday Cage | Electro-magnetic noise shielding | ||
Borosilicate Glass Capillaries | World Precision Instruments | 1B100F-4 | |
Pipette Puller | Sutter Instrument Company | Model P-87 Flaming/Brown Micropipette Puller | |
Beadle and Ephrussi Ringer Solution | See recipe in protocol section | ||
Tricholine citrate, 65% | Sigma | T0252-100G | |
Stereo Microscope | Olympus | VMZ 1x-4x | Capable of 10x-40x magnification |
Ice Bucket | Generic | ||
p200 Pipette Tips | Generic | ||
Spinal Needle | Terumo | SN*2590 | |
1ml Syringe | Beckton-Dickenson | 301025 | |
Fly Aspirator | Assembled from P1000 pipette tips, flexible plastic tubing, and mesh | ||
Modeling Clay | Generic | ||
Forceps | Fine Science Tools By Dumont | 11252-00 | #5SF (super-fine tips) |
10ml Syringe | Beckton-Dickinson | 301029 | |
Plastic Tubing | Tygon | R-3603 |