Annual Report 2004: FlyWorld

Transcription

1 167 Appendix A Annual Report 2004: FlyWorld A.1 Introduction of FlyWorld Observation rather than experimentation dominates the study of animal behavior, limiting our understanding. We require the ability to study behavior in which aspects of an animals environment can be controlled. To meet this goal, we built a multi-chambered, biosphere in which we can control parameters to replicate and examine pertinent aspects of an animals natural environment, while precisely quantifying its behavior. FlyWorld allows both controlled input manipulation and precise behavioral quantification, while capable of parallel, high-throughput analysis essential for neuro-genetic studies in the fruit fly Drosophila melanogaster. Whereas we built this tool for studies of the elementary decision processes in Drosophila, its design being both general and modular anticipates a variety of future behavioral studies. A.2 Modular, experimental chambers FlyWorld is designed to give a researcher the control and quantitative means to compare the behavior of Drosophila both within parallel experimental runs on a single day

2 168 as well between trials run over separate days. This tool incorporates commercially available regulated environmental incubators (capable of creating various ambient temperatures, photoperiods, and monitoring humidity) to house isolated chambers of our own design connected by tubing that may be combined in various conformations (see Fig. A.1). Automated, solenoid-driven gates and infrared sensors together control and quantify locomotor behavior between the chambers described above (see Fig. A.2). The general, modular design allows the experimental flexibility for use in a variety of behavioral studies that include parallel configurations necessary for high-throughput experimentation for anticipated neuro-genetics studies. A.3 Dedicated circuit boards Dedicated, programmable microprocessors incorporated in circuit boards of our own design keep computation local and thus allow a single computer the capacity to run over 900 modules (see Fig. A.3 and Fig. A.4). The Atmel ATMega8 microprocessors we use are easily programmed with any compatible Atmel programmer. Activation of the gates between chambers is controlled through H-bridges (L293DNE). And finally, additional connections are incorporated into our circuit design to anticipate future experiments. Some examples of possible inputs are surface temperature readings through a thermocouple and humidity, gas, or vibration sensors. Examples of outputs are solenoid-driven doors to release animals, uncover food, or power to drive countless other devises.

3 169 A.4 User-friendly software The final component of FlyWorld is a home-born computer software package to query and analyze the bi-directional counts from the microprocessors as well as drive gates between chambers to a variety of control networks. For example, we can specify that either all gates or a subset may be opened or closed by a specified time, or a directional count in a specified counter block. A user-friendly GUI (graphical user interface) aids in the setup of experiments (see Fig. A.5). A.5 High-throughput, quantitative behavioral studies Abstract: Ethological studies on resource-emigration in the fruit fly, Drosophila melanogaster A new direction for the Dickinson laboratory is to understand how the simple nervous system of the fruit fly Drosophila melanogaster supports sophisticated behaviors such as decision making. Specifically, we aim to study decisions involved in resource assessment, an element of habitat selection. We propose a quantitative analysis of specific external, internal, and intrinsic factors that appear to influence emigration behavior from an established food resource. The objective is to establish a research program to assist in the defining of genes and neural hierarchies involved in elementary decision processes (see Fig. A.6 for an example of preliminary results).

4 170 AA B C Figure 1. Prototype components of the FlyWorld apparatus. (A) Top schematic view of simple configuration consisting of two canisters, connector tubing, detector, and gate. (B) CAD-rendered view of configuration drawn in A. (C) Figure A.1: Prototype components of the FlyWorld apparatus. (A) Top schematic Photograph of prototype detector block fabrication from stereo lithography. view of simple configuration consisting of two canisters, connector tubing, detector, and gate. (B) CAD-rendered view of configuration drawn in A. (C) Photograph of prototype detector block fabrication from stereo lithography. 3 Figure 2. Photograph of current detector block and solenoid-driven gate. One of two infrared LED emitter and infrared photodiode detector pairs (arrows) that Figure A.2: Photograph of detector block and solenoid-driven gate. One of two in- make up the bi-directional counter has been removed and displayed along the frared LED emitterdetector and and infrared gate unit. photodiode detector pairs (arrows) that make up the bi-directional counter has been removed and displayed along the detector and gate unit. 4

5 171 E D F B A C G H I Figure 3. Photograph of circuit board with key components identified. (A) Atmel Figure A.3: Photograph microchip, of(b) theh-bridge circuit to board drive with gates, key (C) components inductor/capacitor identified. act to filter (A) Atmel microchip, (B) H-bridge oscillations toin drive power gates, required (C) for the inductor/capacitor analog digital converter act toneeded filter oscillations to read in power required for signals thefrom analog infrared to digital detectors, converter (D) detector needed jumpers, to(e) read emitter signals lines, (F) from gate infrared detectors, (D) detector activation jumpers, lines, (G) extra (E) emitter connections lines, for future (F) gate experiments, activation (H) 12V lines, and 5V (G) extra connections SWARM, for future power 2004 Annual input, experiments, and Report (I) serial (H) ports 12V to connect and 5V circuit power board input, to computer. and (I) serial ports to connect circuit board to computer. 5 Figure 4. Photograph of FlyWorlds in a room with temperature and photoperiod- Figure A.4: Photograph of the FlyWorlds in a room with temperature and photoperiodcontrol. The parallel, two chamber configuration is used for experiments described control. The parallel, two chamber is used for within this described report. within this report.

6 172 Figure 5. A computer screen shot of our software s application window. A reader can see pull-down menus and toggle buttons that enable the Figure A.5: A computer experimenter screen to modify shot experimental of parameters. our software s Here we test application the influence of window. A reader food deprivation and genetic variation on the baseline locomotor behavior of ~50 can see pull-down menus and toggle buttons that enable the experimenter to modify flies per experiment between two chambers ( blue and red traces are a natural experimental parameters. Here we test the influence of food deprivation and genetic fly isolate collected from Chicago, IL deprived of food for 12 or 36 hours variation on the baseline respectively; locomotor the green and behavior yellow traces are oftwo 50 distinct flies isolates per collected experiment by between two others from Toronto, Canada and deprived of food for 12 hours). chambers ( blue and red traces are a natural fly isolate collected from Chicago, IL deprived of food for 12 or 36 hours respectively; the green and yellow traces are two distinct isolates collected by others from Toronto, Canada and deprived of food for 12 hours). 7

7 173 Canton-S Index of Locomotion (counts) raw data w/ means+/-se TIME (hours) Figure 6. An example of results where we establish emigration baselines by manipulating various experimental parameters. Shown above is the baseline Figure A.6: An example locomotor of behavior the results from a first to where a second we chamber establish ~50 12-hour emigration food-deprived baselines by ma- flies (blue) as compared to the locomotor behavior of ~50 12-hour food-deprived nipulating various experimental parameters. Shown above is the baseline locomotor flies introduced to a first chamber that contains a small food resource (red). Here behavior from a first to a second chamber hour food-deprived flies (blue) as we examine Canton-S the widely-used fly stock from which many moleculargenetic tools have behavior been derived. of We 50 manipulate 12-hour the amount food-deprived of resource to flies introduced to compared to the locomotor be consumed over the time course of the experiment. While initially the flies stay a first chamber that contains a small food resource (red). Here we examine Canton-S in the first chamber, eventually they become food-deprived and emigrate to the the widely-used fly stock second chamber frompresumably whichin many search of molecular-genetic food. tools have been derived. We manipulate the amount of food resource to be consumed over the time course of the experiment. While initially the flies stay in the first chamber, eventually they become food-deprived and emigrate to the second chamber presumably in search of food. 8