Running hot and cold: behavioral strategies, neural circuits, and the molecular machinery for thermotaxis in C. elegans and Drosophila

Abstract

Like other ectotherms, the roundworm Caenorhabditis elegans and the fruit fly Drosophila melanogaster rely on behavioral strategies to stabilize their body temperature. Both animals use specialized sensory neurons to detect small changes in temperature, and the activity of these thermosensors governs the neural circuits that control migration and accumulation at preferred temperatures. Despite these similarities, the underlying molecular, neuronal, and computational mechanisms responsible for thermotaxis are distinct in these organisms. Here, we discuss the role of thermosensation in the development and survival of C. elegans and Drosophila, and review the behavioral strategies, neuronal circuits, and molecular networks responsible for thermotaxis behavior.

Figure 1.

Stochasticity, determinism, and decision-making in C. elegans and Drosophila thermotaxis. (A,B) Biased random walks. Trajectories of individual C. elegans nematodes or D. melanogaster larvae are intrinsically stochastic, resembling biased random walks when viewed at low magnification. Each trajectory is marked by periods of forward movement (runs) interrupted by abrupt reorientation maneuvers (turns). (A) Trajectories of four worms grown at 15°C and placed on a temperature gradient above T_c. All worms exhibit negative thermotaxis, but no two trajectories are alike. (B) Trajectories of four Drosophila larvae exhibiting positive thermotaxis moving from below preferred temperature toward warmer temperatures. (Gray circles) Starting points. (C) Unbiased reorientation in C. elegans. The initial run, in which the worm is headed to the lower left, is terminated by a reversal. The reversal is terminated by an Ω turn. Following the Ω turn, the worm starts a new run in a new direction. The worm biases its random walk by regulating the frequency of these abrupt turns, depending on direction with respect to the surrounding temperature gradient. Runs are lengthened toward preferred temperatures and shortened away from preferred temperatures. (D) Biased reorientation in D. melanogaster. During each reorientation maneuver, a Drosophila larva will pause and deliberately swing its head back and forth. If the larva happens to be isothermally aligned, it will encounter falling temperatures if it moves its head in one direction (illustrated by blue dots that trace the trajectory of the thermosensory neurons at the larva's head) and rising temperatures if it moves its head in the other direction (the trace of red dots). The larva biases the probability of starting a new run based on the temperature it encounters during each head sweep. Thus, after being isothermally aligned, the larva tends to start new runs toward the larva's preferred temperature. (E) C. elegans steers in a deterministic manner to maintain isothermal alignment when near its T_c (isothermal tracking). During isothermal tracking, C. elegans continuously swings its head back and forth; thus, the thermosensory neurons at its head encounter regular and alternating phases of falling temperature (blue dots) and rising temperature (red dots). The worm uses this sinusoidal variation in temperature to correct the curvature of its undulation, veering its nose away from every temperature variation and stabilizing isothermal alignment to within ∼0.01°C. (F) When C. elegans is exposed to positive temperature pulses when it swings its head to one side (filled red dots) but not the other side (open red dots), it curves its overall trajectory away from the thermal stimulation. (G) Comparison of behavioral strategies that govern C. elegans thermotaxis, D. melagnoster larval thermotaxis, and Escherichia coli chemotaxis.

Figure 2.

C. elegans and Drosophila thermoreceptor neurons are activated by warming in vivo. (A) Warming above a threshold, T*, increases intracellular calcium in wild-type AFD neurons. T* is variable and is approximately equal to the cultivation temperature, T_c (Kimura et al. 2004; Clark et al. 2006). (B) Warming above ∼25°C–27°C increases the activity of wild-type AC neurons in Drosophila. This response requires an intact trpA1 gene (gray) (Hamada et al. 2008). The experimental basis for the schemas in A and B is in vivo calcium imaging using genetically encoded calcium indicators. For simplicity, increases in intracellular calcium are assumed to indicate increases in cell activity.

Figure 3.

A working model of the circuit underlying thermotactic navigation behaviors in C. elegans. (A,B) Proposed circuits regulating isothermal tracking behaviors at temperatures around T_c (A) and negative thermotaxis at temperatures above T_c (B). Note that the circuit generating deterministic steering during isothermal tracking remains unknown. At T < T_c, positive thermotaxis is elicited under specific conditions via mechanisms that are not fully defined. Sensory neurons and interneurons are indicated by triangles and hexagons, respectively. Neurons experimentally implicated in the indicated behaviors are indicated in color. The AIB interneurons have been suggested previously to be a component of this circuit (Mori and Ohshima 1995). Known excitatory chemical synapses are indicated by arrows, and known inhibitory synapses are indicated by bars. Lines with diamonds on each end indicate gap junctions. Lines with filled balls indicate chemical synapses of unknown sign. See the text for additional details and references.