Inexpensive Methods for Live Imaging of Central Pattern Generator Activity in the Drosophila Larval Locomotor System

Abstract

Central pattern generators (CPGs) are neural networks that produce rhythmic motor activity in the absence of sensory input. CPGs produce 'fictive' behaviours in vitro which parallel activity seen in intact animals. CPG networks have been identified in a wide variety of model organisms and have been shown to be critical for generating rhythmic behaviours such as swimming, walking, chewing and breathing. Work with CPG preparations has led to fundamental advances in neuroscience; however, most CPG preparations involve intensive dissections and require sophisticated electrophysiology equipment, making export to teaching laboratories problematic. Here we present an integrated approach for bringing the study of locomotor CPGs in Drosophila larvae into teaching laboratories. First, we present freely available genetic constructs that enable educators to express genetically encoded calcium indicators in cells of interest in the larval central nervous system. Next, we describe how to isolate the larval central nervous system and prepare it for live imaging. We then show how to modify standard compound microscopes to enable fluorescent imaging using 3D printed materials and inexpensive optical components. Finally, we show how to use the free image analysis programme ImageJ and freely available features in the signal analysis programme DataView to analyse rhythmic CPG activity in the larval CNS. Comparison of results to those obtained on research equipment shows that signal-to-noise levels are comparable and core features of larval CPG activity can be observed. Overall, this work shows the viability of exporting live imaging experiments to low cost environments and paves the way for new teaching laboratory exercises revolving around optical imaging of CPG activity.

Keywords: Drosophila; GcamP; calcium imaging; central nervous system; epifluorescence; larval locomotion; motor systems; open source; teaching equipment.

Appendix Figure 1

Printing and assembling 3D printed fluorescence module. a) 3D components laid out for a single print run in the slicing program Autodesk Cura. b) 3D components printing in a standard Fused Deposition Modelling printer. Note the rough Yotino adhesion mat on the buildplate (recommended for printing to prevent components from being knocked over mid printing). c) Supports are removed after printing manually from the filter tray (i), emissions filter holder (ii), main body (iii) and light source holder (iv), if there is any difficulty with this consider reducing support density. d) Assembly instructions: Carefully place the dichroic mirror into the filter tray (i), followed by the excitation filter (ii) and emissions filter (iii), then friction fit the C-mount adapter ring to the main body (iv) as well as the light source holder (v). After aligning the LED as much as possible, insert the filter tray (vi) and trinocular adapter (vii) then insert into microscope camera port with camera threaded (viii).

Figure 1

CPG activity is sculpted by sensory feedback and neuromodulation to produce rhythmic behaviours across a wide range of vertebrate and invertebrate species.

Figure 2

Conversion of a trinocular compound microscopes into an epifluorescence live imaging system. a) A simple episcopic illuminator module was designed and 3D printed allowing it to fit to the top of any trinocular brightfield microscope. b) Customized module based on work by Stewart and Giannini (2016), allowing for a simple customizable filter cube which spectrally filters and redirects light from a 30W bicycle light onto the sample through the objective. c) Photograph of a working setup mounted on a standard teaching microscope. d) Wandering third instar larvae expressing GCAMP6S in all neurons were selected. e) Animals were then carefully dissected, pinned and all visceral organs were removed, leaving behind the central nervous system. f) The larval CNS was then isolated from the body and pinned by the imaginal discs and posterior nerve roots. g) Snapshot of CNS expressing GCAMP6S in all neurons. scalebar = 100 μm h) Subtracting the average background allowed for imaging of rhythmically active neurons exclusively; scalebar = 100 μm. i) Fictive locomotor patterns can be seen via a kymograph by showing the change of fluorescent activity across a line defined in x-y space over time (line shown in g,h) scale bar = 10 seconds.

Figure 3

DataView GUI and analysis tools for analyzing calcium dynamics. a) Average projection of maximum intensity over time in a CNS expressing GamP6s. Circles indicate ROIs in which pixel intensity is averaged on every time step. b) Raw pixel intensities from ROIs visualized in DataView. c) dF/F calculations (baseline corrections) done for a selected trace and previewed as a new trace. i) Shows baseline correction function ii) Shows baseline F₀ measurement function iii) baseline normalization options d) Burst/peak detection for a given a trace (blue). i) Shows trace selection option. ii) Shows options for setting burst detection thresholds.

Figure 4

Comparison of data produced on research grade and teaching grade imaging systems. a, b) dF/F traces showing segmentally coordinated activity generated using research grade and teaching grade imaging systems, respectively. Green shows activity on left side, blue shows activity on right side; body segments shown at left. Highlighted areas denote different fictive motor patterns: i) Fictive backward waves, ii) fictive head sweeps and iii) fictive forward waves. c, d) Expanded time scale views of activity patterns shown in a, b. Examples shown are from two separate preparations.